\input texinfo @c -*-texinfo-*- @c %**start of header (This is for running Texinfo on a region.) @setfilename gawk.info @settitle The GNU Awk User's Guide @c %**end of header (This is for running Texinfo on a region.) @c inside ifinfo for older versions of texinfo.tex @ifinfo @c I hope this is the right category @dircategory Programming Languages @direntry * Gawk: (gawk.info). A Text Scanning and Processing Language. @end direntry @end ifinfo @c @set xref-automatic-section-title @c @set DRAFT @c The following information should be updated here only! @c This sets the edition of the document, the version of gawk it @c applies to, and when the document was updated. @set TITLE Effective AWK Programming @set SUBTITLE A User's Guide for GNU Awk @set PATCHLEVEL 4 @set EDITION 1.0.@value{PATCHLEVEL} @set VERSION 3.0 @set UPDATE-MONTH April, 1999 @iftex @set DOCUMENT book @end iftex @ifinfo @set DOCUMENT Info file @end ifinfo @ignore Some comments on the layout for TeX. 1. Use at least texinfo.tex 2.159. It contains fixes that are needed to get the footings for draft mode to not appear. 2. I have done A LOT of work to make this look good. There are `@page' commands and use of `@group ... @end group' in a number of places. If you muck with anything, it's your responsibility not to break the layout. @end ignore @c merge the function and variable indexes into the concept index @ifinfo @synindex fn cp @synindex vr cp @end ifinfo @iftex @syncodeindex fn cp @syncodeindex vr cp @end iftex @c If "finalout" is commented out, the printed output will show @c black boxes that mark lines that are too long. Thus, it is @c unwise to comment it out when running a master in case there are @c overfulls which are deemed okay. @ifclear DRAFT @iftex @finalout @end iftex @end ifclear @smallbook @iftex @c @cropmarks @end iftex @ifinfo This file documents @code{awk}, a program that you can use to select particular records in a file and perform operations upon them. This is Edition @value{EDITION} of @cite{@value{TITLE}}, for the @value{VERSION}.@value{PATCHLEVEL} version of the GNU implementation of AWK. Copyright (C) 1989, 1991, 92, 93, 96, 97, 98, 99 Free Software Foundation, Inc. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. @ignore Permission is granted to process this file through TeX and print the results, provided the printed document carries copying permission notice identical to this one except for the removal of this paragraph (this paragraph not being relevant to the printed manual). @end ignore Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Foundation. @end ifinfo @setchapternewpage odd @titlepage @title @value{TITLE} @subtitle @value{SUBTITLE} @subtitle Edition @value{EDITION} @subtitle @value{UPDATE-MONTH} @author Arnold D. Robbins @ignore @sp 1 @author Based on @cite{The GAWK Manual}, @author by Robbins, Close, Rubin, and Stallman @end ignore @c Include the Distribution inside the titlepage environment so @c that headings are turned off. Headings on and off do not work. @page @vskip 0pt plus 1filll @ifset LEGALJUNK The programs and applications presented in this book have been included for their instructional value. They have been tested with care, but are not guaranteed for any particular purpose. The publisher does not offer any warranties or representations, nor does it accept any liabilities with respect to the programs or applications. So there. @sp 2 UNIX is a registered trademark of X/Open, Ltd. @* Microsoft, MS, and MS-DOS are registered trademarks, and Windows is a trademark of Microsoft Corporation in the United States and other countries. @* Atari, 520ST, 1040ST, TT, STE, Mega, and Falcon are registered trademarks or trademarks of Atari Corporation. @* DEC, Digital, OpenVMS, ULTRIX, and VMS, are trademarks of Digital Equipment Corporation. @* @end ifset ``To boldly go where no man has gone before'' is a Registered Trademark of Paramount Pictures Corporation. @* @c sorry, i couldn't resist @sp 3 Copyright @copyright{} 1989, 1991, 92, 93, 96, 97, 98, 99 Free Software Foundation, Inc. @sp 2 This is Edition @value{EDITION} of @cite{@value{TITLE}}, @* for the @value{VERSION}.@value{PATCHLEVEL} (or later) version of the GNU implementation of AWK. @sp 2 @center Published jointly by: @multitable {Specialized Systems Consultants, Inc. (SSC)} {Boston, MA 02111-1307 USA} @item Specialized Systems Consultants, Inc. (SSC) @tab Free Software Foundation @item PO Box 55549 @tab 59 Temple Place --- Suite 330 @item Seattle, WA 98155 USA @tab Boston, MA 02111-1307 USA @item Phone: +1-206-782-7733 @tab Phone: +1-617-542-5942 @item Fax: +1-206-782-7191 @tab Fax: +1-617-542-2652 @item E-mail: @code{sales@@ssc.com} @tab E-mail: @code{gnu@@gnu.org} @item URL: @code{http://www.ssc.com/} @tab URL: @code{http://www.fsf.org/} @end multitable @sp 1 @c this ISBN can change! Check with SSC @c This one is correct for gawk 3.0 and edition 1.0 from the FSF ISBN 1-882114-26-4 @* @c This one is correct for gawk 3.0.3 and edition 1.0.3 from SSC @c ISBN 1-57831-000-8 @* Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that this permission notice may be stated in a translation approved by the Foundation. @sp 2 @c Cover art by Etienne Suvasa. Cover art by Amy Wells Wood. @end titlepage @c Thanks to Bob Chassell for directions on doing dedications. @iftex @headings off @page @w{ } @sp 9 @center @i{To Miriam, for making me complete.} @sp 1 @center @i{To Chana, for the joy you bring us.} @sp 1 @center @i{To Rivka, for the exponential increase.} @sp 1 @center @i{To Nachum, for the added dimension.} @page @w{ } @page @headings on @end iftex @iftex @headings off @evenheading @thispage@ @ @ @strong{@value{TITLE}} @| @| @oddheading @| @| @strong{@thischapter}@ @ @ @thispage @ifset DRAFT @evenfooting @today{} @| @emph{DRAFT!} @| Please Do Not Redistribute @oddfooting Please Do Not Redistribute @| @emph{DRAFT!} @| @today{} @end ifset @end iftex @ifinfo @node Top, Preface, (dir), (dir) @top General Introduction @c Preface or Licensing nodes should come right after the Top @c node, in `unnumbered' sections, then the chapter, `What is gawk'. This file documents @code{awk}, a program that you can use to select particular records in a file and perform operations upon them. This is Edition @value{EDITION} of @cite{@value{TITLE}}, @* for the @value{VERSION}.@value{PATCHLEVEL} version of the GNU implementation @* of AWK. @end ifinfo @menu * Preface:: What this @value{DOCUMENT} is about; brief history and acknowledgements. * What Is Awk:: What is the @code{awk} language; using this @value{DOCUMENT}. * Getting Started:: A basic introduction to using @code{awk}. How to run an @code{awk} program. Command line syntax. * One-liners:: Short, sample @code{awk} programs. * Regexp:: All about matching things using regular expressions. * Reading Files:: How to read files and manipulate fields. * Printing:: How to print using @code{awk}. Describes the @code{print} and @code{printf} statements. Also describes redirection of output. * Expressions:: Expressions are the basic building blocks of statements. * Patterns and Actions:: Overviews of patterns and actions. * Statements:: The various control statements are described in detail. * Built-in Variables:: Built-in Variables * Arrays:: The description and use of arrays. Also includes array-oriented control statements. * Built-in:: The built-in functions are summarized here. * User-defined:: User-defined functions are described in detail. * Invoking Gawk:: How to run @code{gawk}. * Library Functions:: A Library of @code{awk} Functions. * Sample Programs:: Many @code{awk} programs with complete explanations. * Language History:: The evolution of the @code{awk} language. * Gawk Summary:: @code{gawk} Options and Language Summary. * Installation:: Installing @code{gawk} under various operating systems. * Notes:: Something about the implementation of @code{gawk}. * Glossary:: An explanation of some unfamiliar terms. * Copying:: Your right to copy and distribute @code{gawk}. * Index:: Concept and Variable Index. * History:: The history of @code{gawk} and @code{awk}. * Manual History:: Brief history of the GNU project and this @value{DOCUMENT}. * Acknowledgements:: Acknowledgements. * This Manual:: Using this @value{DOCUMENT}. Includes sample input files that you can use. * Conventions:: Typographical Conventions. * Sample Data Files:: Sample data files for use in the @code{awk} programs illustrated in this @value{DOCUMENT}. * Names:: What name to use to find @code{awk}. * Running gawk:: How to run @code{gawk} programs; includes command line syntax. * One-shot:: Running a short throw-away @code{awk} program. * Read Terminal:: Using no input files (input from terminal instead). * Long:: Putting permanent @code{awk} programs in files. * Executable Scripts:: Making self-contained @code{awk} programs. * Comments:: Adding documentation to @code{gawk} programs. * Very Simple:: A very simple example. * Two Rules:: A less simple one-line example with two rules. * More Complex:: A more complex example. * Statements/Lines:: Subdividing or combining statements into lines. * Other Features:: Other Features of @code{awk}. * When:: When to use @code{gawk} and when to use other things. * Regexp Usage:: How to Use Regular Expressions. * Escape Sequences:: How to write non-printing characters. * Regexp Operators:: Regular Expression Operators. * GNU Regexp Operators:: Operators specific to GNU software. * Case-sensitivity:: How to do case-insensitive matching. * Leftmost Longest:: How much text matches. * Computed Regexps:: Using Dynamic Regexps. * Records:: Controlling how data is split into records. * Fields:: An introduction to fields. * Non-Constant Fields:: Non-constant Field Numbers. * Changing Fields:: Changing the Contents of a Field. * Field Separators:: The field separator and how to change it. * Basic Field Splitting:: How fields are split with single characters or simple strings. * Regexp Field Splitting:: Using regexps as the field separator. * Single Character Fields:: Making each character a separate field. * Command Line Field Separator:: Setting @code{FS} from the command line. * Field Splitting Summary:: Some final points and a summary table. * Constant Size:: Reading constant width data. * Multiple Line:: Reading multi-line records. * Getline:: Reading files under explicit program control using the @code{getline} function. * Getline Intro:: Introduction to the @code{getline} function. * Plain Getline:: Using @code{getline} with no arguments. * Getline/Variable:: Using @code{getline} into a variable. * Getline/File:: Using @code{getline} from a file. * Getline/Variable/File:: Using @code{getline} into a variable from a file. * Getline/Pipe:: Using @code{getline} from a pipe. * Getline/Variable/Pipe:: Using @code{getline} into a variable from a pipe. * Getline Summary:: Summary Of @code{getline} Variants. * Print:: The @code{print} statement. * Print Examples:: Simple examples of @code{print} statements. * Output Separators:: The output separators and how to change them. * OFMT:: Controlling Numeric Output With @code{print}. * Printf:: The @code{printf} statement. * Basic Printf:: Syntax of the @code{printf} statement. * Control Letters:: Format-control letters. * Format Modifiers:: Format-specification modifiers. * Printf Examples:: Several examples. * Redirection:: How to redirect output to multiple files and pipes. * Special Files:: File name interpretation in @code{gawk}. @code{gawk} allows access to inherited file descriptors. * Close Files And Pipes:: Closing Input and Output Files and Pipes. * Constants:: String, numeric, and regexp constants. * Scalar Constants:: Numeric and string constants. * Regexp Constants:: Regular Expression constants. * Using Constant Regexps:: When and how to use a regexp constant. * Variables:: Variables give names to values for later use. * Using Variables:: Using variables in your programs. * Assignment Options:: Setting variables on the command line and a summary of command line syntax. This is an advanced method of input. * Conversion:: The conversion of strings to numbers and vice versa. * Arithmetic Ops:: Arithmetic operations (@samp{+}, @samp{-}, etc.) * Concatenation:: Concatenating strings. * Assignment Ops:: Changing the value of a variable or a field. * Increment Ops:: Incrementing the numeric value of a variable. * Truth Values:: What is ``true'' and what is ``false''. * Typing and Comparison:: How variables acquire types, and how this affects comparison of numbers and strings with @samp{<}, etc. * Boolean Ops:: Combining comparison expressions using boolean operators @samp{||} (``or''), @samp{&&} (``and'') and @samp{!} (``not''). * Conditional Exp:: Conditional expressions select between two subexpressions under control of a third subexpression. * Function Calls:: A function call is an expression. * Precedence:: How various operators nest. * Pattern Overview:: What goes into a pattern. * Kinds of Patterns:: A list of all kinds of patterns. * Regexp Patterns:: Using regexps as patterns. * Expression Patterns:: Any expression can be used as a pattern. * Ranges:: Pairs of patterns specify record ranges. * BEGIN/END:: Specifying initialization and cleanup rules. * Using BEGIN/END:: How and why to use BEGIN/END rules. * I/O And BEGIN/END:: I/O issues in BEGIN/END rules. * Empty:: The empty pattern, which matches every record. * Action Overview:: What goes into an action. * If Statement:: Conditionally execute some @code{awk} statements. * While Statement:: Loop until some condition is satisfied. * Do Statement:: Do specified action while looping until some condition is satisfied. * For Statement:: Another looping statement, that provides initialization and increment clauses. * Break Statement:: Immediately exit the innermost enclosing loop. * Continue Statement:: Skip to the end of the innermost enclosing loop. * Next Statement:: Stop processing the current input record. * Nextfile Statement:: Stop processing the current file. * Exit Statement:: Stop execution of @code{awk}. * User-modified:: Built-in variables that you change to control @code{awk}. * Auto-set:: Built-in variables where @code{awk} gives you information. * ARGC and ARGV:: Ways to use @code{ARGC} and @code{ARGV}. * Array Intro:: Introduction to Arrays * Reference to Elements:: How to examine one element of an array. * Assigning Elements:: How to change an element of an array. * Array Example:: Basic Example of an Array * Scanning an Array:: A variation of the @code{for} statement. It loops through the indices of an array's existing elements. * Delete:: The @code{delete} statement removes an element from an array. * Numeric Array Subscripts:: How to use numbers as subscripts in @code{awk}. * Uninitialized Subscripts:: Using Uninitialized variables as subscripts. * Multi-dimensional:: Emulating multi-dimensional arrays in @code{awk}. * Multi-scanning:: Scanning multi-dimensional arrays. * Calling Built-in:: How to call built-in functions. * Numeric Functions:: Functions that work with numbers, including @code{int}, @code{sin} and @code{rand}. * String Functions:: Functions for string manipulation, such as @code{split}, @code{match}, and @code{sprintf}. * I/O Functions:: Functions for files and shell commands. * Time Functions:: Functions for dealing with time stamps. * Definition Syntax:: How to write definitions and what they mean. * Function Example:: An example function definition and what it does. * Function Caveats:: Things to watch out for. * Return Statement:: Specifying the value a function returns. * Options:: Command line options and their meanings. * Other Arguments:: Input file names and variable assignments. * AWKPATH Variable:: Searching directories for @code{awk} programs. * Obsolete:: Obsolete Options and/or features. * Undocumented:: Undocumented Options and Features. * Known Bugs:: Known Bugs in @code{gawk}. * Portability Notes:: What to do if you don't have @code{gawk}. * Nextfile Function:: Two implementations of a @code{nextfile} function. * Assert Function:: A function for assertions in @code{awk} programs. * Round Function:: A function for rounding if @code{sprintf} does not do it correctly. * Ordinal Functions:: Functions for using characters as numbers and vice versa. * Join Function:: A function to join an array into a string. * Mktime Function:: A function to turn a date into a timestamp. * Gettimeofday Function:: A function to get formatted times. * Filetrans Function:: A function for handling data file transitions. * Getopt Function:: A function for processing command line arguments. * Passwd Functions:: Functions for getting user information. * Group Functions:: Functions for getting group information. * Library Names:: How to best name private global variables in library functions. * Clones:: Clones of common utilities. * Cut Program:: The @code{cut} utility. * Egrep Program:: The @code{egrep} utility. * Id Program:: The @code{id} utility. * Split Program:: The @code{split} utility. * Tee Program:: The @code{tee} utility. * Uniq Program:: The @code{uniq} utility. * Wc Program:: The @code{wc} utility. * Miscellaneous Programs:: Some interesting @code{awk} programs. * Dupword Program:: Finding duplicated words in a document. * Alarm Program:: An alarm clock. * Translate Program:: A program similar to the @code{tr} utility. * Labels Program:: Printing mailing labels. * Word Sorting:: A program to produce a word usage count. * History Sorting:: Eliminating duplicate entries from a history file. * Extract Program:: Pulling out programs from Texinfo source files. * Simple Sed:: A Simple Stream Editor. * Igawk Program:: A wrapper for @code{awk} that includes files. * V7/SVR3.1:: The major changes between V7 and System V Release 3.1. * SVR4:: Minor changes between System V Releases 3.1 and 4. * POSIX:: New features from the POSIX standard. * BTL:: New features from the Bell Laboratories version of @code{awk}. * POSIX/GNU:: The extensions in @code{gawk} not in POSIX @code{awk}. * Command Line Summary:: Recapitulation of the command line. * Language Summary:: A terse review of the language. * Variables/Fields:: Variables, fields, and arrays. * Fields Summary:: Input field splitting. * Built-in Summary:: @code{awk}'s built-in variables. * Arrays Summary:: Using arrays. * Data Type Summary:: Values in @code{awk} are numbers or strings. * Rules Summary:: Patterns and Actions, and their component parts. * Pattern Summary:: Quick overview of patterns. * Regexp Summary:: Quick overview of regular expressions. * Actions Summary:: Quick overview of actions. * Operator Summary:: @code{awk} operators. * Control Flow Summary:: The control statements. * I/O Summary:: The I/O statements. * Printf Summary:: A summary of @code{printf}. * Special File Summary:: Special file names interpreted internally. * Built-in Functions Summary:: Built-in numeric and string functions. * Time Functions Summary:: Built-in time functions. * String Constants Summary:: Escape sequences in strings. * Functions Summary:: Defining and calling functions. * Historical Features:: Some undocumented but supported ``features''. * Gawk Distribution:: What is in the @code{gawk} distribution. * Getting:: How to get the distribution. * Extracting:: How to extract the distribution. * Distribution contents:: What is in the distribution. * Unix Installation:: Installing @code{gawk} under various versions of Unix. * Quick Installation:: Compiling @code{gawk} under Unix. * Configuration Philosophy:: How it's all supposed to work. * VMS Installation:: Installing @code{gawk} on VMS. * VMS Compilation:: How to compile @code{gawk} under VMS. * VMS Installation Details:: How to install @code{gawk} under VMS. * VMS Running:: How to run @code{gawk} under VMS. * VMS POSIX:: Alternate instructions for VMS POSIX. * PC Installation:: Installing and Compiling @code{gawk} on MS-DOS and OS/2 * Atari Installation:: Installing @code{gawk} on the Atari ST. * Atari Compiling:: Compiling @code{gawk} on Atari * Atari Using:: Running @code{gawk} on Atari * Amiga Installation:: Installing @code{gawk} on an Amiga. * Bugs:: Reporting Problems and Bugs. * Other Versions:: Other freely available @code{awk} implementations. * Compatibility Mode:: How to disable certain @code{gawk} extensions. * Additions:: Making Additions To @code{gawk}. * Adding Code:: Adding code to the main body of @code{gawk}. * New Ports:: Porting @code{gawk} to a new operating system. * Future Extensions:: New features that may be implemented one day. * Improvements:: Suggestions for improvements by volunteers. @end menu @c dedication for Info file @ifinfo @center To Miriam, for making me complete. @sp 1 @center To Chana, for the joy you bring us. @sp 1 @center To Rivka, for the exponential increase. @sp 1 @center To Nachum, for the added dimension. @end ifinfo @node Preface, What Is Awk, Top, Top @unnumbered Preface @c I saw a comment somewhere that the preface should describe the book itself, @c and the introduction should describe what the book covers. This @value{DOCUMENT} teaches you about the @code{awk} language and how you can use it effectively. You should already be familiar with basic system commands, such as @code{cat} and @code{ls},@footnote{These commands are available on POSIX compliant systems, as well as on traditional Unix based systems. If you are using some other operating system, you still need to be familiar with the ideas of I/O redirection and pipes.} and basic shell facilities, such as Input/Output (I/O) redirection and pipes. Implementations of the @code{awk} language are available for many different computing environments. This @value{DOCUMENT}, while describing the @code{awk} language in general, also describes a particular implementation of @code{awk} called @code{gawk} (which stands for ``GNU Awk''). @code{gawk} runs on a broad range of Unix systems, ranging from 80386 PC-based computers, up through large scale systems, such as Crays. @code{gawk} has also been ported to MS-DOS and OS/2 PC's, Atari and Amiga micro-computers, and VMS. @menu * History:: The history of @code{gawk} and @code{awk}. * Manual History:: Brief history of the GNU project and this @value{DOCUMENT}. * Acknowledgements:: Acknowledgements. @end menu @node History, Manual History, Preface, Preface @unnumberedsec History of @code{awk} and @code{gawk} @cindex acronym @cindex history of @code{awk} @cindex Aho, Alfred @cindex Weinberger, Peter @cindex Kernighan, Brian @cindex old @code{awk} @cindex new @code{awk} The name @code{awk} comes from the initials of its designers: Alfred V.@: Aho, Peter J.@: Weinberger, and Brian W.@: Kernighan. The original version of @code{awk} was written in 1977 at AT&T Bell Laboratories. In 1985 a new version made the programming language more powerful, introducing user-defined functions, multiple input streams, and computed regular expressions. This new version became generally available with Unix System V Release 3.1. The version in System V Release 4 added some new features and also cleaned up the behavior in some of the ``dark corners'' of the language. The specification for @code{awk} in the POSIX Command Language and Utilities standard further clarified the language based on feedback from both the @code{gawk} designers, and the original Bell Labs @code{awk} designers. The GNU implementation, @code{gawk}, was written in 1986 by Paul Rubin and Jay Fenlason, with advice from Richard Stallman. John Woods contributed parts of the code as well. In 1988 and 1989, David Trueman, with help from Arnold Robbins, thoroughly reworked @code{gawk} for compatibility with the newer @code{awk}. Current development focuses on bug fixes, performance improvements, standards compliance, and occasionally, new features. @node Manual History, Acknowledgements, History, Preface @unnumberedsec The GNU Project and This Book @cindex Free Software Foundation @cindex Stallman, Richard The Free Software Foundation (FSF) is a non-profit organization dedicated to the production and distribution of freely distributable software. It was founded by Richard M.@: Stallman, the author of the original Emacs editor. GNU Emacs is the most widely used version of Emacs today. @cindex GNU Project The GNU project is an on-going effort on the part of the Free Software Foundation to create a complete, freely distributable, POSIX compliant computing environment. (GNU stands for ``GNU's not Unix''.) The FSF uses the ``GNU General Public License'' (or GPL) to ensure that source code for their software is always available to the end user. A copy of the GPL is included for your reference (@pxref{Copying, ,GNU GENERAL PUBLIC LICENSE}). The GPL applies to the C language source code for @code{gawk}. A shell, an editor (Emacs), highly portable optimizing C, C++, and Objective-C compilers, a symbolic debugger, and dozens of large and small utilities (such as @code{gawk}), have all been completed and are freely available. As of this writing (early 1997), the GNU operating system kernel (the HURD), has been released, but is still in an early stage of development. @cindex Linux @cindex NetBSD @cindex FreeBSD Until the GNU operating system is more fully developed, you should consider using Linux, a freely distributable, Unix-like operating system for 80386, DEC Alpha, Sun SPARC and other systems. There are many books on Linux. One freely available one is @cite{Linux Installation and Getting Started}, by Matt Welsh. Many Linux distributions are available, often in computer stores or bundled on CD-ROM with books about Linux. (There are three other freely available, Unix-like operating systems for 80386 and other systems, NetBSD, FreeBSD,and OpenBSD. All are based on the 4.4-Lite Berkeley Software Distribution, and they use recent versions of @code{gawk} for their versions of @code{awk}.) @iftex This @value{DOCUMENT} you are reading now is actually free. The information in it is freely available to anyone, the machine readable source code for the @value{DOCUMENT} comes with @code{gawk}, and anyone may take this @value{DOCUMENT} to a copying machine and make as many copies of it as they like. (Take a moment to check the copying permissions on the Copyright page.) If you paid money for this @value{DOCUMENT}, what you actually paid for was the @value{DOCUMENT}'s nice printing and binding, and the publisher's associated costs to produce it. We have made an effort to keep these costs reasonable; most people would prefer a bound book to over 330 pages of photo-copied text that would then have to be held in a loose-leaf binder (not to mention the time and labor involved in doing the copying). The same is true of producing this @value{DOCUMENT} from the machine readable source; the retail price is only slightly more than the cost per page of printing it on a laser printer. @end iftex This @value{DOCUMENT} itself has gone through several previous, preliminary editions. I started working on a preliminary draft of @cite{The GAWK Manual}, by Diane Close, Paul Rubin, and Richard Stallman in the fall of 1988. It was around 90 pages long, and barely described the original, ``old'' version of @code{awk}. After substantial revision, the first version of the @cite{The GAWK Manual} to be released was Edition 0.11 Beta in October of 1989. The manual then underwent more substantial revision for Edition 0.13 of December 1991. David Trueman, Pat Rankin, and Michal Jaegermann contributed sections of the manual for Edition 0.13. That edition was published by the FSF as a bound book early in 1992. Since then there have been several minor revisions, notably Edition 0.14 of November 1992 that was published by the FSF in January of 1993, and Edition 0.16 of August 1993. Edition 1.0 of @cite{@value{TITLE}} represents a significant re-working of @cite{The GAWK Manual}, with much additional material. The FSF and I agree that I am now the primary author. I also felt that it needed a more descriptive title. @cite{@value{TITLE}} will undoubtedly continue to evolve. An electronic version comes with the @code{gawk} distribution from the FSF. If you find an error in this @value{DOCUMENT}, please report it! @xref{Bugs, ,Reporting Problems and Bugs}, for information on submitting problem reports electronically, or write to me in care of the FSF. @node Acknowledgements, , Manual History, Preface @unnumberedsec Acknowledgements @cindex Stallman, Richard I would like to acknowledge Richard M.@: Stallman, for his vision of a better world, and for his courage in founding the FSF and starting the GNU project. The initial draft of @cite{The GAWK Manual} had the following acknowledgements: @quotation Many people need to be thanked for their assistance in producing this manual. Jay Fenlason contributed many ideas and sample programs. Richard Mlynarik and Robert Chassell gave helpful comments on drafts of this manual. The paper @cite{A Supplemental Document for @code{awk}} by John W.@: Pierce of the Chemistry Department at UC San Diego, pinpointed several issues relevant both to @code{awk} implementation and to this manual, that would otherwise have escaped us. @end quotation The following people provided many helpful comments on Edition 0.13 of @cite{The GAWK Manual}: Rick Adams, Michael Brennan, Rich Burridge, Diane Close, Christopher (``Topher'') Eliot, Michael Lijewski, Pat Rankin, Miriam Robbins, and Michal Jaegermann. The following people provided many helpful comments for Edition 1.0 of @cite{@value{TITLE}}: Karl Berry, Michael Brennan, Darrel Hankerson, Michal Jaegermann, Michael Lijewski, and Miriam Robbins. Pat Rankin, Michal Jaegermann, Darrel Hankerson and Scott Deifik updated their respective sections for Edition 1.0. Robert J.@: Chassell provided much valuable advice on the use of Texinfo. He also deserves special thanks for convincing me @emph{not} to title this @value{DOCUMENT} @cite{How To Gawk Politely}. Karl Berry helped significantly with the @TeX{} part of Texinfo. @cindex Trueman, David David Trueman deserves special credit; he has done a yeoman job of evolving @code{gawk} so that it performs well, and without bugs. Although he is no longer involved with @code{gawk}, working with him on this project was a significant pleasure. @cindex Deifik, Scott @cindex Hankerson, Darrel @cindex Rommel, Kai Uwe @cindex Rankin, Pat @cindex Jaegermann, Michal Scott Deifik, Darrel Hankerson, Kai Uwe Rommel, Pat Rankin, and Michal Jaegermann (in no particular order) are long time members of the @code{gawk} ``crack portability team.'' Without their hard work and help, @code{gawk} would not be nearly the fine program it is today. It has been and continues to be a pleasure working with this team of fine people. @cindex Friedl, Jeffrey Jeffrey Friedl provided invaluable help in tracking down a number of last minute problems with regular expressions in @code{gawk} 3.0. @cindex Kernighan, Brian David and I would like to thank Brian Kernighan of Bell Labs for invaluable assistance during the testing and debugging of @code{gawk}, and for help in clarifying numerous points about the language. We could not have done nearly as good a job on either @code{gawk} or its documentation without his help. @cindex Hughes, Phil I would like to thank Marshall and Elaine Hartholz of Seattle, and Dr.@: Bert and Rita Schreiber of Detroit for large amounts of quiet vacation time in their homes, which allowed me to make significant progress on this @value{DOCUMENT} and on @code{gawk} itself. Phil Hughes of SSC contributed in a very important way by loaning me his laptop Linux system, not once, but twice, allowing me to do a lot of work while away from home. @cindex Robbins, Miriam Finally, I must thank my wonderful wife, Miriam, for her patience through the many versions of this project, for her proof-reading, and for sharing me with the computer. I would like to thank my parents for their love, and for the grace with which they raised and educated me. I also must acknowledge my gratitude to G-d, for the many opportunities He has sent my way, as well as for the gifts He has given me with which to take advantage of those opportunities. @sp 2 @noindent Arnold Robbins @* Atlanta, Georgia @* February, 1997 @ignore Stuff still not covered anywhere: BASICS: Integer vs. floating point Hex vs. octal vs. decimal Interpreter vs compiler input/output @end ignore @node What Is Awk, Getting Started, Preface, Top @chapter Introduction If you are like many computer users, you would frequently like to make changes in various text files wherever certain patterns appear, or extract data from parts of certain lines while discarding the rest. To write a program to do this in a language such as C or Pascal is a time-consuming inconvenience that may take many lines of code. The job may be easier with @code{awk}. The @code{awk} utility interprets a special-purpose programming language that makes it possible to handle simple data-reformatting jobs with just a few lines of code. The GNU implementation of @code{awk} is called @code{gawk}; it is fully upward compatible with the System V Release 4 version of @code{awk}. @code{gawk} is also upward compatible with the POSIX specification of the @code{awk} language. This means that all properly written @code{awk} programs should work with @code{gawk}. Thus, we usually don't distinguish between @code{gawk} and other @code{awk} implementations. @cindex uses of @code{awk} Using @code{awk} you can: @itemize @bullet @item manage small, personal databases @item generate reports @item validate data @item produce indexes, and perform other document preparation tasks @item even experiment with algorithms that can be adapted later to other computer languages @end itemize @menu * This Manual:: Using this @value{DOCUMENT}. Includes sample input files that you can use. * Conventions:: Typographical Conventions. * Sample Data Files:: Sample data files for use in the @code{awk} programs illustrated in this @value{DOCUMENT}. @end menu @node This Manual, Conventions, What Is Awk, What Is Awk @section Using This Book @cindex book, using this @cindex using this book @cindex language, @code{awk} @cindex program, @code{awk} @ignore @cindex @code{awk} language @cindex @code{awk} program @end ignore The term @code{awk} refers to a particular program, and to the language you use to tell this program what to do. When we need to be careful, we call the program ``the @code{awk} utility'' and the language ``the @code{awk} language.'' The term @code{gawk} refers to a version of @code{awk} developed as part the GNU project. The purpose of this @value{DOCUMENT} is to explain both the @code{awk} language and how to run the @code{awk} utility. The main purpose of the @value{DOCUMENT} is to explain the features of @code{awk}, as defined in the POSIX standard. It does so in the context of one particular implementation, @code{gawk}. While doing so, it will also attempt to describe important differences between @code{gawk} and other @code{awk} implementations. Finally, any @code{gawk} features that are not in the POSIX standard for @code{awk} will be noted. @iftex This @value{DOCUMENT} has the difficult task of being both tutorial and reference. If you are a novice, feel free to skip over details that seem too complex. You should also ignore the many cross references; they are for the expert user, and for the on-line Info version of the document. @end iftex The term @dfn{@code{awk} program} refers to a program written by you in the @code{awk} programming language. @xref{Getting Started, ,Getting Started with @code{awk}}, for the bare essentials you need to know to start using @code{awk}. Some useful ``one-liners'' are included to give you a feel for the @code{awk} language (@pxref{One-liners, ,Useful One Line Programs}). Many sample @code{awk} programs have been provided for you (@pxref{Library Functions, ,A Library of @code{awk} Functions}; also @pxref{Sample Programs, ,Practical @code{awk} Programs}). The entire @code{awk} language is summarized for quick reference in @ref{Gawk Summary, ,@code{gawk} Summary}. Look there if you just need to refresh your memory about a particular feature. If you find terms that you aren't familiar with, try looking them up in the glossary (@pxref{Glossary}). Most of the time complete @code{awk} programs are used as examples, but in some of the more advanced sections, only the part of the @code{awk} program that illustrates the concept being described is shown. While this @value{DOCUMENT} is aimed principally at people who have not been exposed to @code{awk}, there is a lot of information here that even the @code{awk} expert should find useful. In particular, the description of POSIX @code{awk}, and the example programs in @ref{Library Functions, ,A Library of @code{awk} Functions}, and @ref{Sample Programs, ,Practical @code{awk} Programs}, should be of interest. @c fakenode --- for prepinfo @unnumberedsubsec Dark Corners @display @i{Who opened that window shade?!?} Count Dracula @end display @sp 1 @cindex d.c., see ``dark corner'' @cindex dark corner Until the POSIX standard (and @cite{The Gawk Manual}), many features of @code{awk} were either poorly documented, or not documented at all. Descriptions of such features (often called ``dark corners'') are noted in this @value{DOCUMENT} with ``(d.c.)''. They also appear in the index under the heading ``dark corner.'' @node Conventions, Sample Data Files, This Manual, What Is Awk @section Typographical Conventions This @value{DOCUMENT} is written using Texinfo, the GNU documentation formatting language. A single Texinfo source file is used to produce both the printed and on-line versions of the documentation. @iftex Because of this, the typographical conventions are slightly different than in other books you may have read. @end iftex @ifinfo This section briefly documents the typographical conventions used in Texinfo. @end ifinfo Examples you would type at the command line are preceded by the common shell primary and secondary prompts, @samp{$} and @samp{>}. Output from the command is preceded by the glyph ``@print{}''. This typically represents the command's standard output. Error messages, and other output on the command's standard error, are preceded by the glyph ``@error{}''. For example: @example @group $ echo hi on stdout @print{} hi on stdout $ echo hello on stderr 1>&2 @error{} hello on stderr @end group @end example @iftex In the text, command names appear in @code{this font}, while code segments appear in the same font and quoted, @samp{like this}. Some things will be emphasized @emph{like this}, and if a point needs to be made strongly, it will be done @strong{like this}. The first occurrence of a new term is usually its @dfn{definition}, and appears in the same font as the previous occurrence of ``definition'' in this sentence. File names are indicated like this: @file{/path/to/ourfile}. @end iftex Characters that you type at the keyboard look @kbd{like this}. In particular, there are special characters called ``control characters.'' These are characters that you type by holding down both the @kbd{CONTROL} key and another key, at the same time. For example, a @kbd{Control-d} is typed by first pressing and holding the @kbd{CONTROL} key, next pressing the @kbd{d} key, and finally releasing both keys. @node Sample Data Files, , Conventions, What Is Awk @section Data Files for the Examples @cindex input file, sample @cindex sample input file @cindex @file{BBS-list} file Many of the examples in this @value{DOCUMENT} take their input from two sample data files. The first, called @file{BBS-list}, represents a list of computer bulletin board systems together with information about those systems. The second data file, called @file{inventory-shipped}, contains information about shipments on a monthly basis. In both files, each line is considered to be one @dfn{record}. In the file @file{BBS-list}, each record contains the name of a computer bulletin board, its phone number, the board's baud rate(s), and a code for the number of hours it is operational. An @samp{A} in the last column means the board operates 24 hours a day. A @samp{B} in the last column means the board operates evening and weekend hours, only. A @samp{C} means the board operates only on weekends. @c 2e: Update the baud rates to reflect today's faster modems @example @c system mkdir eg @c system mkdir eg/lib @c system mkdir eg/data @c system mkdir eg/prog @c system mkdir eg/misc @c file eg/data/BBS-list aardvark 555-5553 1200/300 B alpo-net 555-3412 2400/1200/300 A barfly 555-7685 1200/300 A bites 555-1675 2400/1200/300 A camelot 555-0542 300 C core 555-2912 1200/300 C fooey 555-1234 2400/1200/300 B foot 555-6699 1200/300 B macfoo 555-6480 1200/300 A sdace 555-3430 2400/1200/300 A sabafoo 555-2127 1200/300 C @c endfile @end example @cindex @file{inventory-shipped} file The second data file, called @file{inventory-shipped}, represents information about shipments during the year. Each record contains the month of the year, the number of green crates shipped, the number of red boxes shipped, the number of orange bags shipped, and the number of blue packages shipped, respectively. There are 16 entries, covering the 12 months of one year and four months of the next year. @example @c file eg/data/inventory-shipped Jan 13 25 15 115 Feb 15 32 24 226 Mar 15 24 34 228 Apr 31 52 63 420 May 16 34 29 208 Jun 31 42 75 492 Jul 24 34 67 436 Aug 15 34 47 316 Sep 13 55 37 277 Oct 29 54 68 525 Nov 20 87 82 577 Dec 17 35 61 401 Jan 21 36 64 620 Feb 26 58 80 652 Mar 24 75 70 495 Apr 21 70 74 514 @c endfile @end example @ifinfo If you are reading this in GNU Emacs using Info, you can copy the regions of text showing these sample files into your own test files. This way you can try out the examples shown in the remainder of this document. You do this by using the command @kbd{M-x write-region} to copy text from the Info file into a file for use with @code{awk} (@xref{Misc File Ops, , Miscellaneous File Operations, emacs, GNU Emacs Manual}, for more information). Using this information, create your own @file{BBS-list} and @file{inventory-shipped} files, and practice what you learn in this @value{DOCUMENT}. If you are using the stand-alone version of Info, see @ref{Extract Program, ,Extracting Programs from Texinfo Source Files}, for an @code{awk} program that will extract these data files from @file{gawk.texi}, the Texinfo source file for this Info file. @end ifinfo @node Getting Started, One-liners, What Is Awk, Top @chapter Getting Started with @code{awk} @cindex script, definition of @cindex rule, definition of @cindex program, definition of @cindex basic function of @code{awk} The basic function of @code{awk} is to search files for lines (or other units of text) that contain certain patterns. When a line matches one of the patterns, @code{awk} performs specified actions on that line. @code{awk} keeps processing input lines in this way until the end of the input files are reached. @cindex data-driven languages @cindex procedural languages @cindex language, data-driven @cindex language, procedural Programs in @code{awk} are different from programs in most other languages, because @code{awk} programs are @dfn{data-driven}; that is, you describe the data you wish to work with, and then what to do when you find it. Most other languages are @dfn{procedural}; you have to describe, in great detail, every step the program is to take. When working with procedural languages, it is usually much harder to clearly describe the data your program will process. For this reason, @code{awk} programs are often refreshingly easy to both write and read. @cindex program, definition of @cindex rule, definition of When you run @code{awk}, you specify an @code{awk} @dfn{program} that tells @code{awk} what to do. The program consists of a series of @dfn{rules}. (It may also contain @dfn{function definitions}, an advanced feature which we will ignore for now. @xref{User-defined, ,User-defined Functions}.) Each rule specifies one pattern to search for, and one action to perform when that pattern is found. Syntactically, a rule consists of a pattern followed by an action. The action is enclosed in curly braces to separate it from the pattern. Rules are usually separated by newlines. Therefore, an @code{awk} program looks like this: @example @var{pattern} @{ @var{action} @} @var{pattern} @{ @var{action} @} @dots{} @end example @menu * Names:: What name to use to find @code{awk}. * Running gawk:: How to run @code{gawk} programs; includes command line syntax. * Very Simple:: A very simple example. * Two Rules:: A less simple one-line example with two rules. * More Complex:: A more complex example. * Statements/Lines:: Subdividing or combining statements into lines. * Other Features:: Other Features of @code{awk}. * When:: When to use @code{gawk} and when to use other things. @end menu @node Names, Running gawk , Getting Started, Getting Started @section A Rose By Any Other Name @cindex old @code{awk} vs. new @code{awk} @cindex new @code{awk} vs. old @code{awk} The @code{awk} language has evolved over the years. Full details are provided in @ref{Language History, ,The Evolution of the @code{awk} Language}. The language described in this @value{DOCUMENT} is often referred to as ``new @code{awk}.'' Because of this, many systems have multiple versions of @code{awk}. Some systems have an @code{awk} utility that implements the original version of the @code{awk} language, and a @code{nawk} utility for the new version. Others have an @code{oawk} for the ``old @code{awk}'' language, and plain @code{awk} for the new one. Still others only have one version, usually the new one.@footnote{Often, these systems use @code{gawk} for their @code{awk} implementation!} All in all, this makes it difficult for you to know which version of @code{awk} you should run when writing your programs. The best advice we can give here is to check your local documentation. Look for @code{awk}, @code{oawk}, and @code{nawk}, as well as for @code{gawk}. Chances are, you will have some version of new @code{awk} on your system, and that is what you should use when running your programs. (Of course, if you're reading this @value{DOCUMENT}, chances are good that you have @code{gawk}!) Throughout this @value{DOCUMENT}, whenever we refer to a language feature that should be available in any complete implementation of POSIX @code{awk}, we simply use the term @code{awk}. When referring to a feature that is specific to the GNU implementation, we use the term @code{gawk}. @node Running gawk, Very Simple, Names, Getting Started @section How to Run @code{awk} Programs @cindex command line formats @cindex running @code{awk} programs There are several ways to run an @code{awk} program. If the program is short, it is easiest to include it in the command that runs @code{awk}, like this: @example awk '@var{program}' @var{input-file1} @var{input-file2} @dots{} @end example @noindent where @var{program} consists of a series of patterns and actions, as described earlier. (The reason for the single quotes is described below, in @ref{One-shot, ,One-shot Throw-away @code{awk} Programs}.) When the program is long, it is usually more convenient to put it in a file and run it with a command like this: @example awk -f @var{program-file} @var{input-file1} @var{input-file2} @dots{} @end example @menu * One-shot:: Running a short throw-away @code{awk} program. * Read Terminal:: Using no input files (input from terminal instead). * Long:: Putting permanent @code{awk} programs in files. * Executable Scripts:: Making self-contained @code{awk} programs. * Comments:: Adding documentation to @code{gawk} programs. @end menu @node One-shot, Read Terminal, Running gawk, Running gawk @subsection One-shot Throw-away @code{awk} Programs Once you are familiar with @code{awk}, you will often type in simple programs the moment you want to use them. Then you can write the program as the first argument of the @code{awk} command, like this: @example awk '@var{program}' @var{input-file1} @var{input-file2} @dots{} @end example @noindent where @var{program} consists of a series of @var{patterns} and @var{actions}, as described earlier. @cindex single quotes, why needed This command format instructs the @dfn{shell}, or command interpreter, to start @code{awk} and use the @var{program} to process records in the input file(s). There are single quotes around @var{program} so that the shell doesn't interpret any @code{awk} characters as special shell characters. They also cause the shell to treat all of @var{program} as a single argument for @code{awk} and allow @var{program} to be more than one line long. This format is also useful for running short or medium-sized @code{awk} programs from shell scripts, because it avoids the need for a separate file for the @code{awk} program. A self-contained shell script is more reliable since there are no other files to misplace. @ref{One-liners, , Useful One Line Programs}, presents several short, self-contained programs. As an interesting side point, the command @example awk '/foo/' @var{files} @dots{} @end example @noindent is essentially the same as @cindex @code{egrep} @example egrep foo @var{files} @dots{} @end example @node Read Terminal, Long, One-shot, Running gawk @subsection Running @code{awk} without Input Files @cindex standard input @cindex input, standard You can also run @code{awk} without any input files. If you type the command line: @example awk '@var{program}' @end example @noindent then @code{awk} applies the @var{program} to the @dfn{standard input}, which usually means whatever you type on the terminal. This continues until you indicate end-of-file by typing @kbd{Control-d}. (On other operating systems, the end-of-file character may be different. For example, on OS/2 and MS-DOS, it is @kbd{Control-z}.) For example, the following program prints a friendly piece of advice (from Douglas Adams' @cite{The Hitchhiker's Guide to the Galaxy}), to keep you from worrying about the complexities of computer programming (@samp{BEGIN} is a feature we haven't discussed yet). @example $ awk "BEGIN @{ print \"Don't Panic!\" @}" @print{} Don't Panic! @end example @cindex quoting, shell @cindex shell quoting This program does not read any input. The @samp{\} before each of the inner double quotes is necessary because of the shell's quoting rules, in particular because it mixes both single quotes and double quotes. This next simple @code{awk} program emulates the @code{cat} utility; it copies whatever you type at the keyboard to its standard output. (Why this works is explained shortly.) @example $ awk '@{ print @}' Now is the time for all good men @print{} Now is the time for all good men to come to the aid of their country. @print{} to come to the aid of their country. Four score and seven years ago, ... @print{} Four score and seven years ago, ... What, me worry? @print{} What, me worry? @kbd{Control-d} @end example @node Long, Executable Scripts, Read Terminal, Running gawk @subsection Running Long Programs @cindex running long programs @cindex @code{-f} option @cindex program file @cindex file, @code{awk} program Sometimes your @code{awk} programs can be very long. In this case it is more convenient to put the program into a separate file. To tell @code{awk} to use that file for its program, you type: @example awk -f @var{source-file} @var{input-file1} @var{input-file2} @dots{} @end example The @samp{-f} instructs the @code{awk} utility to get the @code{awk} program from the file @var{source-file}. Any file name can be used for @var{source-file}. For example, you could put the program: @example BEGIN @{ print "Don't Panic!" @} @end example @noindent into the file @file{advice}. Then this command: @example awk -f advice @end example @noindent does the same thing as this one: @example awk "BEGIN @{ print \"Don't Panic!\" @}" @end example @cindex quoting, shell @cindex shell quoting @noindent which was explained earlier (@pxref{Read Terminal, ,Running @code{awk} without Input Files}). Note that you don't usually need single quotes around the file name that you specify with @samp{-f}, because most file names don't contain any of the shell's special characters. Notice that in @file{advice}, the @code{awk} program did not have single quotes around it. The quotes are only needed for programs that are provided on the @code{awk} command line. If you want to identify your @code{awk} program files clearly as such, you can add the extension @file{.awk} to the file name. This doesn't affect the execution of the @code{awk} program, but it does make ``housekeeping'' easier. @node Executable Scripts, Comments, Long, Running gawk @subsection Executable @code{awk} Programs @cindex executable scripts @cindex scripts, executable @cindex self contained programs @cindex program, self contained @cindex @code{#!} (executable scripts) Once you have learned @code{awk}, you may want to write self-contained @code{awk} scripts, using the @samp{#!} script mechanism. You can do this on many Unix systems@footnote{The @samp{#!} mechanism works on Linux systems, Unix systems derived from Berkeley Unix, System V Release 4, and some System V Release 3 systems.} (and someday on the GNU system). For example, you could update the file @file{advice} to look like this: @example #! /bin/awk -f BEGIN @{ print "Don't Panic!" @} @end example @noindent After making this file executable (with the @code{chmod} utility), you can simply type @samp{advice} at the shell, and the system will arrange to run @code{awk}@footnote{The line beginning with @samp{#!} lists the full file name of an interpreter to be run, and an optional initial command line argument to pass to that interpreter. The operating system then runs the interpreter with the given argument and the full argument list of the executed program. The first argument in the list is the full file name of the @code{awk} program. The rest of the argument list will either be options to @code{awk}, or data files, or both.} as if you had typed @samp{awk -f advice}. @example @group $ advice @print{} Don't Panic! @end group @end example @noindent Self-contained @code{awk} scripts are useful when you want to write a program which users can invoke without their having to know that the program is written in @code{awk}. @strong{Caution:} You should not put more than one argument on the @samp{#!} line after the path to @code{awk}. This will not work. The operating system treats the rest of the line as a single agument, and passes it to @code{awk}. Doing this will lead to confusing behavior: most likely a usage diagnostic of some sort from @code{awk}. @cindex shell scripts @cindex scripts, shell Some older systems do not support the @samp{#!} mechanism. You can get a similar effect using a regular shell script. It would look something like this: @example : The colon ensures execution by the standard shell. awk '@var{program}' "$@@" @end example Using this technique, it is @emph{vital} to enclose the @var{program} in single quotes to protect it from interpretation by the shell. If you omit the quotes, only a shell wizard can predict the results. The @code{"$@@"} causes the shell to forward all the command line arguments to the @code{awk} program, without interpretation. The first line, which starts with a colon, is used so that this shell script will work even if invoked by a user who uses the C shell. (Not all older systems obey this convention, but many do.) @c 2e: @c Someday: (See @cite{The Bourne Again Shell}, by ??.) @node Comments, , Executable Scripts, Running gawk @subsection Comments in @code{awk} Programs @cindex @code{#} (comment) @cindex comments @cindex use of comments @cindex documenting @code{awk} programs @cindex programs, documenting A @dfn{comment} is some text that is included in a program for the sake of human readers; it is not really part of the program. Comments can explain what the program does, and how it works. Nearly all programming languages have provisions for comments, because programs are typically hard to understand without their extra help. In the @code{awk} language, a comment starts with the sharp sign character, @samp{#}, and continues to the end of the line. The @samp{#} does not have to be the first character on the line. The @code{awk} language ignores the rest of a line following a sharp sign. For example, we could have put the following into @file{advice}: @example # This program prints a nice friendly message. It helps # keep novice users from being afraid of the computer. BEGIN @{ print "Don't Panic!" @} @end example You can put comment lines into keyboard-composed throw-away @code{awk} programs also, but this usually isn't very useful; the purpose of a comment is to help you or another person understand the program at a later time. @strong{Caution:} As mentioned in @ref{One-shot, ,One-shot Throw-away @code{awk} Programs}, you can enclose small to medium programs in single quotes, in order to keep your shell scripts self-contained. When doing so, @emph{don't} put an apostrophe (i.e., a single quote) into a comment (or anywhere else in your program). The shell will interpret the quote as the closing quote for the entire program. As a result, usually the shell will print a message about mismatched quotes, and if @code{awk} actually runs, it will probably print strange messages about syntax errors. For example: @example awk 'BEGIN @{ print "hello" @} # let's be cute' @end example @node Very Simple, Two Rules, Running gawk, Getting Started @section A Very Simple Example The following command runs a simple @code{awk} program that searches the input file @file{BBS-list} for the string of characters: @samp{foo}. (A string of characters is usually called a @dfn{string}. The term @dfn{string} is perhaps based on similar usage in English, such as ``a string of pearls,'' or, ``a string of cars in a train.'') @example awk '/foo/ @{ print $0 @}' BBS-list @end example @noindent When lines containing @samp{foo} are found, they are printed, because @w{@samp{print $0}} means print the current line. (Just @samp{print} by itself means the same thing, so we could have written that instead.) You will notice that slashes, @samp{/}, surround the string @samp{foo} in the @code{awk} program. The slashes indicate that @samp{foo} is a pattern to search for. This type of pattern is called a @dfn{regular expression}, and is covered in more detail later (@pxref{Regexp, ,Regular Expressions}). The pattern is allowed to match parts of words. There are single-quotes around the @code{awk} program so that the shell won't interpret any of it as special shell characters. Here is what this program prints: @example @group $ awk '/foo/ @{ print $0 @}' BBS-list @print{} fooey 555-1234 2400/1200/300 B @print{} foot 555-6699 1200/300 B @print{} macfoo 555-6480 1200/300 A @print{} sabafoo 555-2127 1200/300 C @end group @end example @cindex action, default @cindex pattern, default @cindex default action @cindex default pattern In an @code{awk} rule, either the pattern or the action can be omitted, but not both. If the pattern is omitted, then the action is performed for @emph{every} input line. If the action is omitted, the default action is to print all lines that match the pattern. @cindex empty action @cindex action, empty Thus, we could leave out the action (the @code{print} statement and the curly braces) in the above example, and the result would be the same: all lines matching the pattern @samp{foo} would be printed. By comparison, omitting the @code{print} statement but retaining the curly braces makes an empty action that does nothing; then no lines would be printed. @node Two Rules, More Complex, Very Simple, Getting Started @section An Example with Two Rules @cindex how @code{awk} works The @code{awk} utility reads the input files one line at a time. For each line, @code{awk} tries the patterns of each of the rules. If several patterns match then several actions are run, in the order in which they appear in the @code{awk} program. If no patterns match, then no actions are run. After processing all the rules (perhaps none) that match the line, @code{awk} reads the next line (however, @pxref{Next Statement, ,The @code{next} Statement}, and also @pxref{Nextfile Statement, ,The @code{nextfile} Statement}). This continues until the end of the file is reached. For example, the @code{awk} program: @example /12/ @{ print $0 @} /21/ @{ print $0 @} @end example @noindent contains two rules. The first rule has the string @samp{12} as the pattern and @samp{print $0} as the action. The second rule has the string @samp{21} as the pattern and also has @samp{print $0} as the action. Each rule's action is enclosed in its own pair of braces. This @code{awk} program prints every line that contains the string @samp{12} @emph{or} the string @samp{21}. If a line contains both strings, it is printed twice, once by each rule. This is what happens if we run this program on our two sample data files, @file{BBS-list} and @file{inventory-shipped}, as shown here: @example $ awk '/12/ @{ print $0 @} > /21/ @{ print $0 @}' BBS-list inventory-shipped @print{} aardvark 555-5553 1200/300 B @print{} alpo-net 555-3412 2400/1200/300 A @print{} barfly 555-7685 1200/300 A @print{} bites 555-1675 2400/1200/300 A @print{} core 555-2912 1200/300 C @print{} fooey 555-1234 2400/1200/300 B @print{} foot 555-6699 1200/300 B @print{} macfoo 555-6480 1200/300 A @print{} sdace 555-3430 2400/1200/300 A @print{} sabafoo 555-2127 1200/300 C @print{} sabafoo 555-2127 1200/300 C @print{} Jan 21 36 64 620 @print{} Apr 21 70 74 514 @end example @noindent Note how the line in @file{BBS-list} beginning with @samp{sabafoo} was printed twice, once for each rule. @node More Complex, Statements/Lines, Two Rules, Getting Started @section A More Complex Example @ignore We have to use ls -lg here to get portable output across Unix systems. The POSIX ls matches this behavior too. Sigh. @end ignore Here is an example to give you an idea of what typical @code{awk} programs do. This example shows how @code{awk} can be used to summarize, select, and rearrange the output of another utility. It uses features that haven't been covered yet, so don't worry if you don't understand all the details. @example ls -lg | awk '$6 == "Nov" @{ sum += $5 @} END @{ print sum @}' @end example @cindex @code{csh}, backslash continuation @cindex backslash continuation in @code{csh} This command prints the total number of bytes in all the files in the current directory that were last modified in November (of any year). (In the C shell you would need to type a semicolon and then a backslash at the end of the first line; in a POSIX-compliant shell, such as the Bourne shell or Bash, the GNU Bourne-Again shell, you can type the example as shown.) @ignore FIXME: how can users tell what shell they are running? Need a footnote or something, but getting into this is a distraction. @end ignore The @w{@samp{ls -lg}} part of this example is a system command that gives you a listing of the files in a directory, including file size and the date the file was last modified. Its output looks like this: @example -rw-r--r-- 1 arnold user 1933 Nov 7 13:05 Makefile -rw-r--r-- 1 arnold user 10809 Nov 7 13:03 gawk.h -rw-r--r-- 1 arnold user 983 Apr 13 12:14 gawk.tab.h -rw-r--r-- 1 arnold user 31869 Jun 15 12:20 gawk.y -rw-r--r-- 1 arnold user 22414 Nov 7 13:03 gawk1.c -rw-r--r-- 1 arnold user 37455 Nov 7 13:03 gawk2.c -rw-r--r-- 1 arnold user 27511 Dec 9 13:07 gawk3.c -rw-r--r-- 1 arnold user 7989 Nov 7 13:03 gawk4.c @end example @noindent The first field contains read-write permissions, the second field contains the number of links to the file, and the third field identifies the owner of the file. The fourth field identifies the group of the file. The fifth field contains the size of the file in bytes. The sixth, seventh and eighth fields contain the month, day, and time, respectively, that the file was last modified. Finally, the ninth field contains the name of the file. @cindex automatic initialization @cindex initialization, automatic The @samp{$6 == "Nov"} in our @code{awk} program is an expression that tests whether the sixth field of the output from @w{@samp{ls -lg}} matches the string @samp{Nov}. Each time a line has the string @samp{Nov} for its sixth field, the action @samp{sum += $5} is performed. This adds the fifth field (the file size) to the variable @code{sum}. As a result, when @code{awk} has finished reading all the input lines, @code{sum} is the sum of the sizes of files whose lines matched the pattern. (This works because @code{awk} variables are automatically initialized to zero.) After the last line of output from @code{ls} has been processed, the @code{END} rule is executed, and the value of @code{sum} is printed. In this example, the value of @code{sum} would be 80600. These more advanced @code{awk} techniques are covered in later sections (@pxref{Action Overview, ,Overview of Actions}). Before you can move on to more advanced @code{awk} programming, you have to know how @code{awk} interprets your input and displays your output. By manipulating fields and using @code{print} statements, you can produce some very useful and impressive looking reports. @node Statements/Lines, Other Features, More Complex, Getting Started @section @code{awk} Statements Versus Lines @cindex line break @cindex newline Most often, each line in an @code{awk} program is a separate statement or separate rule, like this: @example awk '/12/ @{ print $0 @} /21/ @{ print $0 @}' BBS-list inventory-shipped @end example However, @code{gawk} will ignore newlines after any of the following: @example , @{ ? : || && do else @end example @noindent A newline at any other point is considered the end of the statement. (Splitting lines after @samp{?} and @samp{:} is a minor @code{gawk} extension. The @samp{?} and @samp{:} referred to here is the three operand conditional expression described in @ref{Conditional Exp, ,Conditional Expressions}.) @cindex backslash continuation @cindex continuation of lines @cindex line continuation If you would like to split a single statement into two lines at a point where a newline would terminate it, you can @dfn{continue} it by ending the first line with a backslash character, @samp{\}. The backslash must be the final character on the line to be recognized as a continuation character. This is allowed absolutely anywhere in the statement, even in the middle of a string or regular expression. For example: @example awk '/This regular expression is too long, so continue it\ on the next line/ @{ print $1 @}' @end example @noindent @cindex portability issues We have generally not used backslash continuation in the sample programs in this @value{DOCUMENT}. Since in @code{gawk} there is no limit on the length of a line, it is never strictly necessary; it just makes programs more readable. For this same reason, as well as for clarity, we have kept most statements short in the sample programs presented throughout the @value{DOCUMENT}. Backslash continuation is most useful when your @code{awk} program is in a separate source file, instead of typed in on the command line. You should also note that many @code{awk} implementations are more particular about where you may use backslash continuation. For example, they may not allow you to split a string constant using backslash continuation. Thus, for maximal portability of your @code{awk} programs, it is best not to split your lines in the middle of a regular expression or a string. @cindex @code{csh}, backslash continuation @cindex backslash continuation in @code{csh} @strong{Caution: backslash continuation does not work as described above with the C shell.} Continuation with backslash works for @code{awk} programs in files, and also for one-shot programs @emph{provided} you are using a POSIX-compliant shell, such as the Bourne shell or Bash, the GNU Bourne-Again shell. But the C shell (@code{csh}) behaves differently! There, you must use two backslashes in a row, followed by a newline. Note also that when using the C shell, @emph{every} newline in your awk program must be escaped with a backslash. To illustrate: @example % awk 'BEGIN @{ \ ? print \\ ? "hello, world" \ ? @}' @print{} hello, world @end example @noindent Here, the @samp{%} and @samp{?} are the C shell's primary and secondary prompts, analogous to the standard shell's @samp{$} and @samp{>}. @code{awk} is a line-oriented language. Each rule's action has to begin on the same line as the pattern. To have the pattern and action on separate lines, you @emph{must} use backslash continuation---there is no other way. @cindex backslash continuation and comments @cindex comments and backslash continuation Note that backslash continuation and comments do not mix. As soon as @code{awk} sees the @samp{#} that starts a comment, it ignores @emph{everything} on the rest of the line. For example: @example @group $ gawk 'BEGIN @{ print "dont panic" # a friendly \ > BEGIN rule > @}' @error{} gawk: cmd. line:2: BEGIN rule @error{} gawk: cmd. line:2: ^ parse error @end group @end example @noindent Here, it looks like the backslash would continue the comment onto the next line. However, the backslash-newline combination is never even noticed, since it is ``hidden'' inside the comment. Thus, the @samp{BEGIN} is noted as a syntax error. @cindex multiple statements on one line When @code{awk} statements within one rule are short, you might want to put more than one of them on a line. You do this by separating the statements with a semicolon, @samp{;}. This also applies to the rules themselves. Thus, the previous program could have been written: @example /12/ @{ print $0 @} ; /21/ @{ print $0 @} @end example @noindent @strong{Note:} the requirement that rules on the same line must be separated with a semicolon was not in the original @code{awk} language; it was added for consistency with the treatment of statements within an action. @node Other Features, When, Statements/Lines, Getting Started @section Other Features of @code{awk} The @code{awk} language provides a number of predefined, or built-in variables, which your programs can use to get information from @code{awk}. There are other variables your program can set to control how @code{awk} processes your data. In addition, @code{awk} provides a number of built-in functions for doing common computational and string related operations. As we develop our presentation of the @code{awk} language, we introduce most of the variables and many of the functions. They are defined systematically in @ref{Built-in Variables}, and @ref{Built-in, ,Built-in Functions}. @node When, , Other Features, Getting Started @section When to Use @code{awk} @cindex when to use @code{awk} @cindex applications of @code{awk} You might wonder how @code{awk} might be useful for you. Using utility programs, advanced patterns, field separators, arithmetic statements, and other selection criteria, you can produce much more complex output. The @code{awk} language is very useful for producing reports from large amounts of raw data, such as summarizing information from the output of other utility programs like @code{ls}. (@xref{More Complex, ,A More Complex Example}.) Programs written with @code{awk} are usually much smaller than they would be in other languages. This makes @code{awk} programs easy to compose and use. Often, @code{awk} programs can be quickly composed at your terminal, used once, and thrown away. Since @code{awk} programs are interpreted, you can avoid the (usually lengthy) compilation part of the typical edit-compile-test-debug cycle of software development. Complex programs have been written in @code{awk}, including a complete retargetable assembler for eight-bit microprocessors (@pxref{Glossary}, for more information) and a microcode assembler for a special purpose Prolog computer. However, @code{awk}'s capabilities are strained by tasks of such complexity. If you find yourself writing @code{awk} scripts of more than, say, a few hundred lines, you might consider using a different programming language. Emacs Lisp is a good choice if you need sophisticated string or pattern matching capabilities. The shell is also good at string and pattern matching; in addition, it allows powerful use of the system utilities. More conventional languages, such as C, C++, and Lisp, offer better facilities for system programming and for managing the complexity of large programs. Programs in these languages may require more lines of source code than the equivalent @code{awk} programs, but they are easier to maintain and usually run more efficiently. @node One-liners, Regexp, Getting Started, Top @chapter Useful One Line Programs @cindex one-liners Many useful @code{awk} programs are short, just a line or two. Here is a collection of useful, short programs to get you started. Some of these programs contain constructs that haven't been covered yet. The description of the program will give you a good idea of what is going on, but please read the rest of the @value{DOCUMENT} to become an @code{awk} expert! Most of the examples use a data file named @file{data}. This is just a placeholder; if you were to use these programs yourself, you would substitute your own file names for @file{data}. @ifinfo Since you are reading this in Info, each line of the example code is enclosed in quotes, to represent text that you would type literally. The examples themselves represent shell commands that use single quotes to keep the shell from interpreting the contents of the program. When reading the examples, focus on the text between the open and close quotes. @end ifinfo @table @code @item awk '@{ if (length($0) > max) max = length($0) @} @itemx @ @ @ @ @ END @{ print max @}' data This program prints the length of the longest input line. @item awk 'length($0) > 80' data This program prints every line that is longer than 80 characters. The sole rule has a relational expression as its pattern, and has no action (so the default action, printing the record, is used). @item expand@ data@ |@ awk@ '@{ if (x < length()) x = length() @} @itemx @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ END @{ print "maximum line length is " x @}' This program prints the length of the longest line in @file{data}. The input is processed by the @code{expand} program to change tabs into spaces, so the widths compared are actually the right-margin columns. @item awk 'NF > 0' data This program prints every line that has at least one field. This is an easy way to delete blank lines from a file (or rather, to create a new file similar to the old file but from which the blank lines have been deleted). @c Karl Berry points out that new users probably don't want to see @c multiple ways to do things, just the `best' way. He's probably @c right. At some point it might be worth adding something about there @c often being multiple ways to do things in awk, but for now we'll @c just take this one out. @ignore @item awk '@{ if (NF > 0) print @}' data This program also prints every line that has at least one field. Here we allow the rule to match every line, and then decide in the action whether to print. @end ignore @item awk@ 'BEGIN@ @{@ for (i = 1; i <= 7; i++) @itemx @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ print int(101 * rand()) @}' This program prints seven random numbers from zero to 100, inclusive. @item ls -lg @var{files} | awk '@{ x += $5 @} ; END @{ print "total bytes: " x @}' This program prints the total number of bytes used by @var{files}. @item ls -lg @var{files} | awk '@{ x += $5 @} @itemx @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ @ END @{ print "total K-bytes: " (x + 1023)/1024 @}' This program prints the total number of kilobytes used by @var{files}. @item awk -F: '@{ print $1 @}' /etc/passwd | sort This program prints a sorted list of the login names of all users. @item awk 'END @{ print NR @}' data This program counts lines in a file. @item awk 'NR % 2 == 0' data This program prints the even numbered lines in the data file. If you were to use the expression @samp{NR % 2 == 1} instead, it would print the odd numbered lines. @end table @node Regexp, Reading Files, One-liners, Top @chapter Regular Expressions @cindex pattern, regular expressions @cindex regexp @cindex regular expression @cindex regular expressions as patterns A @dfn{regular expression}, or @dfn{regexp}, is a way of describing a set of strings. Because regular expressions are such a fundamental part of @code{awk} programming, their format and use deserve a separate chapter. A regular expression enclosed in slashes (@samp{/}) is an @code{awk} pattern that matches every input record whose text belongs to that set. The simplest regular expression is a sequence of letters, numbers, or both. Such a regexp matches any string that contains that sequence. Thus, the regexp @samp{foo} matches any string containing @samp{foo}. Therefore, the pattern @code{/foo/} matches any input record containing the three characters @samp{foo}, @emph{anywhere} in the record. Other kinds of regexps let you specify more complicated classes of strings. @iftex Initially, the examples will be simple. As we explain more about how regular expressions work, we will present more complicated examples. @end iftex @menu * Regexp Usage:: How to Use Regular Expressions. * Escape Sequences:: How to write non-printing characters. * Regexp Operators:: Regular Expression Operators. * GNU Regexp Operators:: Operators specific to GNU software. * Case-sensitivity:: How to do case-insensitive matching. * Leftmost Longest:: How much text matches. * Computed Regexps:: Using Dynamic Regexps. @end menu @node Regexp Usage, Escape Sequences, Regexp, Regexp @section How to Use Regular Expressions A regular expression can be used as a pattern by enclosing it in slashes. Then the regular expression is tested against the entire text of each record. (Normally, it only needs to match some part of the text in order to succeed.) For example, this prints the second field of each record that contains the three characters @samp{foo} anywhere in it: @example @group $ awk '/foo/ @{ print $2 @}' BBS-list @print{} 555-1234 @print{} 555-6699 @print{} 555-6480 @print{} 555-2127 @end group @end example @cindex regexp matching operators @cindex string-matching operators @cindex operators, string-matching @cindex operators, regexp matching @cindex regexp match/non-match operators @cindex @code{~} operator @cindex @code{!~} operator Regular expressions can also be used in matching expressions. These expressions allow you to specify the string to match against; it need not be the entire current input record. The two operators, @samp{~} and @samp{!~}, perform regular expression comparisons. Expressions using these operators can be used as patterns or in @code{if}, @code{while}, @code{for}, and @code{do} statements. @ifinfo @c adding this xref in TeX screws up the formatting too much (@xref{Statements, ,Control Statements in Actions}.) @end ifinfo @table @code @item @var{exp} ~ /@var{regexp}/ This is true if the expression @var{exp} (taken as a string) is matched by @var{regexp}. The following example matches, or selects, all input records with the upper-case letter @samp{J} somewhere in the first field: @example @group $ awk '$1 ~ /J/' inventory-shipped @print{} Jan 13 25 15 115 @print{} Jun 31 42 75 492 @print{} Jul 24 34 67 436 @print{} Jan 21 36 64 620 @end group @end example So does this: @example awk '@{ if ($1 ~ /J/) print @}' inventory-shipped @end example @item @var{exp} !~ /@var{regexp}/ This is true if the expression @var{exp} (taken as a character string) is @emph{not} matched by @var{regexp}. The following example matches, or selects, all input records whose first field @emph{does not} contain the upper-case letter @samp{J}: @example @group $ awk '$1 !~ /J/' inventory-shipped @print{} Feb 15 32 24 226 @print{} Mar 15 24 34 228 @print{} Apr 31 52 63 420 @print{} May 16 34 29 208 @dots{} @end group @end example @end table @cindex regexp constant When a regexp is written enclosed in slashes, like @code{/foo/}, we call it a @dfn{regexp constant}, much like @code{5.27} is a numeric constant, and @code{"foo"} is a string constant. @node Escape Sequences, Regexp Operators, Regexp Usage, Regexp @section Escape Sequences @cindex escape sequence notation Some characters cannot be included literally in string constants (@code{"foo"}) or regexp constants (@code{/foo/}). You represent them instead with @dfn{escape sequences}, which are character sequences beginning with a backslash (@samp{\}). One use of an escape sequence is to include a double-quote character in a string constant. Since a plain double-quote would end the string, you must use @samp{\"} to represent an actual double-quote character as a part of the string. For example: @example $ awk 'BEGIN @{ print "He said \"hi!\" to her." @}' @print{} He said "hi!" to her. @end example The backslash character itself is another character that cannot be included normally; you write @samp{\\} to put one backslash in the string or regexp. Thus, the string whose contents are the two characters @samp{"} and @samp{\} must be written @code{"\"\\"}. Another use of backslash is to represent unprintable characters such as tab or newline. While there is nothing to stop you from entering most unprintable characters directly in a string constant or regexp constant, they may look ugly. Here is a table of all the escape sequences used in @code{awk}, and what they represent. Unless noted otherwise, all of these escape sequences apply to both string constants and regexp constants. @c @cartouche @table @code @item \\ A literal backslash, @samp{\}. @cindex @code{awk} language, V.4 version @item \a The ``alert'' character, @kbd{Control-g}, ASCII code 7 (BEL). @item \b Backspace, @kbd{Control-h}, ASCII code 8 (BS). @item \f Formfeed, @kbd{Control-l}, ASCII code 12 (FF). @item \n Newline, @kbd{Control-j}, ASCII code 10 (LF). @item \r Carriage return, @kbd{Control-m}, ASCII code 13 (CR). @item \t Horizontal tab, @kbd{Control-i}, ASCII code 9 (HT). @cindex @code{awk} language, V.4 version @item \v Vertical tab, @kbd{Control-k}, ASCII code 11 (VT). @item \@var{nnn} The octal value @var{nnn}, where @var{nnn} are one to three digits between @samp{0} and @samp{7}. For example, the code for the ASCII ESC (escape) character is @samp{\033}. @cindex @code{awk} language, V.4 version @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @item \x@var{hh}@dots{} The hexadecimal value @var{hh}, where @var{hh} are hexadecimal digits (@samp{0} through @samp{9} and either @samp{A} through @samp{F} or @samp{a} through @samp{f}). Like the same construct in ANSI C, the escape sequence continues until the first non-hexadecimal digit is seen. However, using more than two hexadecimal digits produces undefined results. (The @samp{\x} escape sequence is not allowed in POSIX @code{awk}.) @item \/ A literal slash (necessary for regexp constants only). You use this when you wish to write a regexp constant that contains a slash. Since the regexp is delimited by slashes, you need to escape the slash that is part of the pattern, in order to tell @code{awk} to keep processing the rest of the regexp. @item \" A literal double-quote (necessary for string constants only). You use this when you wish to write a string constant that contains a double-quote. Since the string is delimited by double-quotes, you need to escape the quote that is part of the string, in order to tell @code{awk} to keep processing the rest of the string. @end table @c @end cartouche In @code{gawk}, there are additional two character sequences that begin with backslash that have special meaning in regexps. @xref{GNU Regexp Operators, ,Additional Regexp Operators Only in @code{gawk}}. In a string constant, what happens if you place a backslash before something that is not one of the characters listed above? POSIX @code{awk} purposely leaves this case undefined. There are two choices. @itemize @bullet @item Strip the backslash out. This is what Unix @code{awk} and @code{gawk} both do. For example, @code{"a\qc"} is the same as @code{"aqc"}. @item Leave the backslash alone. Some other @code{awk} implementations do this. In such implementations, @code{"a\qc"} is the same as if you had typed @code{"a\\qc"}. @end itemize In a regexp, a backslash before any character that is not in the above table, and not listed in @ref{GNU Regexp Operators, ,Additional Regexp Operators Only in @code{gawk}}, means that the next character should be taken literally, even if it would normally be a regexp operator. E.g., @code{/a\+b/} matches the three characters @samp{a+b}. @cindex portability issues For complete portability, do not use a backslash before any character not listed in the table above. Another interesting question arises. Suppose you use an octal or hexadecimal escape to represent a regexp metacharacter (@pxref{Regexp Operators, , Regular Expression Operators}). Does @code{awk} treat the character as a literal character, or as a regexp operator? @cindex dark corner It turns out that historically, such characters were taken literally (d.c.). However, the POSIX standard indicates that they should be treated as real metacharacters, and this is what @code{gawk} does. However, in compatibility mode (@pxref{Options, ,Command Line Options}), @code{gawk} treats the characters represented by octal and hexadecimal escape sequences literally when used in regexp constants. Thus, @code{/a\52b/} is equivalent to @code{/a\*b/}. To summarize: @enumerate 1 @item The escape sequences in the table above are always processed first, for both string constants and regexp constants. This happens very early, as soon as @code{awk} reads your program. @item @code{gawk} processes both regexp constants and dynamic regexps (@pxref{Computed Regexps, ,Using Dynamic Regexps}), for the special operators listed in @ref{GNU Regexp Operators, ,Additional Regexp Operators Only in @code{gawk}}. @item A backslash before any other character means to treat that character literally. @end enumerate @node Regexp Operators, GNU Regexp Operators, Escape Sequences, Regexp @section Regular Expression Operators @cindex metacharacters @cindex regular expression metacharacters @cindex regexp operators You can combine regular expressions with the following characters, called @dfn{regular expression operators}, or @dfn{metacharacters}, to increase the power and versatility of regular expressions. The escape sequences described @iftex above @end iftex in @ref{Escape Sequences}, are valid inside a regexp. They are introduced by a @samp{\}. They are recognized and converted into the corresponding real characters as the very first step in processing regexps. Here is a table of metacharacters. All characters that are not escape sequences and that are not listed in the table stand for themselves. @table @code @item \ This is used to suppress the special meaning of a character when matching. For example: @example \$ @end example @noindent matches the character @samp{$}. @c NEEDED @page @cindex anchors in regexps @cindex regexp, anchors @item ^ This matches the beginning of a string. For example: @example ^@@chapter @end example @noindent matches the @samp{@@chapter} at the beginning of a string, and can be used to identify chapter beginnings in Texinfo source files. The @samp{^} is known as an @dfn{anchor}, since it anchors the pattern to matching only at the beginning of the string. It is important to realize that @samp{^} does not match the beginning of a line embedded in a string. In this example the condition is not true: @example if ("line1\nLINE 2" ~ /^L/) @dots{} @end example @item $ This is similar to @samp{^}, but it matches only at the end of a string. For example: @example p$ @end example @noindent matches a record that ends with a @samp{p}. The @samp{$} is also an anchor, and also does not match the end of a line embedded in a string. In this example the condition is not true: @example if ("line1\nLINE 2" ~ /1$/) @dots{} @end example @item . The period, or dot, matches any single character, @emph{including} the newline character. For example: @example .P @end example @noindent matches any single character followed by a @samp{P} in a string. Using concatenation we can make a regular expression like @samp{U.A}, which matches any three-character sequence that begins with @samp{U} and ends with @samp{A}. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} In strict POSIX mode (@pxref{Options, ,Command Line Options}), @samp{.} does not match the @sc{nul} character, which is a character with all bits equal to zero. Otherwise, @sc{nul} is just another character. Other versions of @code{awk} may not be able to match the @sc{nul} character. @ignore 2e: Add stuff that character list is the POSIX terminology. In other literature known as character set or character class. @end ignore @cindex character list @item [@dots{}] This is called a @dfn{character list}. It matches any @emph{one} of the characters that are enclosed in the square brackets. For example: @example [MVX] @end example @noindent matches any one of the characters @samp{M}, @samp{V}, or @samp{X} in a string. Ranges of characters are indicated by using a hyphen between the beginning and ending characters, and enclosing the whole thing in brackets. For example: @example [0-9] @end example @noindent matches any digit. Multiple ranges are allowed. E.g., the list @code{@w{[A-Za-z0-9]}} is a common way to express the idea of ``all alphanumeric characters.'' To include one of the characters @samp{\}, @samp{]}, @samp{-} or @samp{^} in a character list, put a @samp{\} in front of it. For example: @example [d\]] @end example @noindent matches either @samp{d}, or @samp{]}. @cindex @code{egrep} This treatment of @samp{\} in character lists is compatible with other @code{awk} implementations, and is also mandated by POSIX. The regular expressions in @code{awk} are a superset of the POSIX specification for Extended Regular Expressions (EREs). POSIX EREs are based on the regular expressions accepted by the traditional @code{egrep} utility. @cindex character classes @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @dfn{Character classes} are a new feature introduced in the POSIX standard. A character class is a special notation for describing lists of characters that have a specific attribute, but where the actual characters themselves can vary from country to country and/or from character set to character set. For example, the notion of what is an alphabetic character differs in the USA and in France. A character class is only valid in a regexp @emph{inside} the brackets of a character list. Character classes consist of @samp{[:}, a keyword denoting the class, and @samp{:]}. Here are the character classes defined by the POSIX standard. @table @code @item [:alnum:] Alphanumeric characters. @item [:alpha:] Alphabetic characters. @item [:blank:] Space and tab characters. @item [:cntrl:] Control characters. @item [:digit:] Numeric characters. @item [:graph:] Characters that are printable and are also visible. (A space is printable, but not visible, while an @samp{a} is both.) @item [:lower:] Lower-case alphabetic characters. @item [:print:] Printable characters (characters that are not control characters.) @item [:punct:] Punctuation characters (characters that are not letter, digits, control characters, or space characters). @item [:space:] Space characters (such as space, tab, and formfeed, to name a few). @item [:upper:] Upper-case alphabetic characters. @item [:xdigit:] Characters that are hexadecimal digits. @end table For example, before the POSIX standard, to match alphanumeric characters, you had to write @code{/[A-Za-z0-9]/}. If your character set had other alphabetic characters in it, this would not match them. With the POSIX character classes, you can write @code{/[[:alnum:]]/}, and this will match @emph{all} the alphabetic and numeric characters in your character set. @cindex collating elements Two additional special sequences can appear in character lists. These apply to non-ASCII character sets, which can have single symbols (called @dfn{collating elements}) that are represented with more than one character, as well as several characters that are equivalent for @dfn{collating}, or sorting, purposes. (E.g., in French, a plain ``e'' and a grave-accented ``@`e'' are equivalent.) @table @asis @cindex collating symbols @item Collating Symbols A @dfn{collating symbol} is a multi-character collating element enclosed in @samp{[.} and @samp{.]}. For example, if @samp{ch} is a collating element, then @code{[[.ch.]]} is a regexp that matches this collating element, while @code{[ch]} is a regexp that matches either @samp{c} or @samp{h}. @cindex equivalence classes @item Equivalence Classes An @dfn{equivalence class} is a locale-specific name for a list of characters that are equivalent. The name is enclosed in @samp{[=} and @samp{=]}. For example, the name @samp{e} might be used to represent all of ``e,'' ``@`e,'' and ``@'e.'' In this case, @code{[[=e]]} is a regexp that matches any of @samp{e}, @samp{@'e}, or @samp{@`e}. @end table These features are very valuable in non-English speaking locales. @strong{Caution:} The library functions that @code{gawk} uses for regular expression matching currently only recognize POSIX character classes; they do not recognize collating symbols or equivalence classes. @c maybe one day ... @cindex complemented character list @cindex character list, complemented @item [^ @dots{}] This is a @dfn{complemented character list}. The first character after the @samp{[} @emph{must} be a @samp{^}. It matches any characters @emph{except} those in the square brackets. For example: @example [^0-9] @end example @noindent matches any character that is not a digit. @item | This is the @dfn{alternation operator}, and it is used to specify alternatives. For example: @example ^P|[0-9] @end example @noindent matches any string that matches either @samp{^P} or @samp{[0-9]}. This means it matches any string that starts with @samp{P} or contains a digit. The alternation applies to the largest possible regexps on either side. In other words, @samp{|} has the lowest precedence of all the regular expression operators. @item (@dots{}) Parentheses are used for grouping in regular expressions as in arithmetic. They can be used to concatenate regular expressions containing the alternation operator, @samp{|}. For example, @samp{@@(samp|code)\@{[^@}]+\@}} matches both @samp{@@code@{foo@}} and @samp{@@samp@{bar@}}. (These are Texinfo formatting control sequences.) @item * This symbol means that the preceding regular expression is to be repeated as many times as necessary to find a match. For example: @example ph* @end example @noindent applies the @samp{*} symbol to the preceding @samp{h} and looks for matches of one @samp{p} followed by any number of @samp{h}s. This will also match just @samp{p} if no @samp{h}s are present. The @samp{*} repeats the @emph{smallest} possible preceding expression. (Use parentheses if you wish to repeat a larger expression.) It finds as many repetitions as possible. For example: @example awk '/\(c[ad][ad]*r x\)/ @{ print @}' sample @end example @noindent prints every record in @file{sample} containing a string of the form @samp{(car x)}, @samp{(cdr x)}, @samp{(cadr x)}, and so on. Notice the escaping of the parentheses by preceding them with backslashes. @item + This symbol is similar to @samp{*}, but the preceding expression must be matched at least once. This means that: @example wh+y @end example @noindent would match @samp{why} and @samp{whhy} but not @samp{wy}, whereas @samp{wh*y} would match all three of these strings. This is a simpler way of writing the last @samp{*} example: @example awk '/\(c[ad]+r x\)/ @{ print @}' sample @end example @item ? This symbol is similar to @samp{*}, but the preceding expression can be matched either once or not at all. For example: @example fe?d @end example @noindent will match @samp{fed} and @samp{fd}, but nothing else. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @cindex interval expressions @item @{@var{n}@} @itemx @{@var{n},@} @itemx @{@var{n},@var{m}@} One or two numbers inside braces denote an @dfn{interval expression}. If there is one number in the braces, the preceding regexp is repeated @var{n} times. If there are two numbers separated by a comma, the preceding regexp is repeated @var{n} to @var{m} times. If there is one number followed by a comma, then the preceding regexp is repeated at least @var{n} times. @table @code @item wh@{3@}y matches @samp{whhhy} but not @samp{why} or @samp{whhhhy}. @item wh@{3,5@}y matches @samp{whhhy} or @samp{whhhhy} or @samp{whhhhhy}, only. @item wh@{2,@}y matches @samp{whhy} or @samp{whhhy}, and so on. @end table Interval expressions were not traditionally available in @code{awk}. As part of the POSIX standard they were added, to make @code{awk} and @code{egrep} consistent with each other. However, since old programs may use @samp{@{} and @samp{@}} in regexp constants, by default @code{gawk} does @emph{not} match interval expressions in regexps. If either @samp{--posix} or @samp{--re-interval} are specified (@pxref{Options, , Command Line Options}), then interval expressions are allowed in regexps. @end table @cindex precedence, regexp operators @cindex regexp operators, precedence of In regular expressions, the @samp{*}, @samp{+}, and @samp{?} operators, as well as the braces @samp{@{} and @samp{@}}, have the highest precedence, followed by concatenation, and finally by @samp{|}. As in arithmetic, parentheses can change how operators are grouped. If @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}), character classes and interval expressions are not available in regular expressions. The next @ifinfo node @end ifinfo @iftex section @end iftex discusses the GNU-specific regexp operators, and provides more detail concerning how command line options affect the way @code{gawk} interprets the characters in regular expressions. @node GNU Regexp Operators, Case-sensitivity, Regexp Operators, Regexp @section Additional Regexp Operators Only in @code{gawk} @c This section adapted from the regex-0.12 manual @cindex regexp operators, GNU specific GNU software that deals with regular expressions provides a number of additional regexp operators. These operators are described in this section, and are specific to @code{gawk}; they are not available in other @code{awk} implementations. @cindex word, regexp definition of Most of the additional operators are for dealing with word matching. For our purposes, a @dfn{word} is a sequence of one or more letters, digits, or underscores (@samp{_}). @table @code @cindex @code{\w} regexp operator @item \w This operator matches any word-constituent character, i.e.@: any letter, digit, or underscore. Think of it as a short-hand for @c @w{@code{[A-Za-z0-9_]}} or @w{@code{[[:alnum:]_]}}. @cindex @code{\W} regexp operator @item \W This operator matches any character that is not word-constituent. Think of it as a short-hand for @c @w{@code{[^A-Za-z0-9_]}} or @w{@code{[^[:alnum:]_]}}. @cindex @code{\<} regexp operator @item \< This operator matches the empty string at the beginning of a word. For example, @code{/\} regexp operator @item \> This operator matches the empty string at the end of a word. For example, @code{/stow\>/} matches @samp{stow}, but not @samp{stowaway}. @cindex @code{\y} regexp operator @cindex word boundaries, matching @item \y This operator matches the empty string at either the beginning or the end of a word (the word boundar@strong{y}). For example, @samp{\yballs?\y} matches either @samp{ball} or @samp{balls} as a separate word. @cindex @code{\B} regexp operator @item \B This operator matches the empty string within a word. In other words, @samp{\B} matches the empty string that occurs between two word-constituent characters. For example, @code{/\Brat\B/} matches @samp{crate}, but it does not match @samp{dirty rat}. @samp{\B} is essentially the opposite of @samp{\y}. @end table There are two other operators that work on buffers. In Emacs, a @dfn{buffer} is, naturally, an Emacs buffer. For other programs, the regexp library routines that @code{gawk} uses consider the entire string to be matched as the buffer. For @code{awk}, since @samp{^} and @samp{$} always work in terms of the beginning and end of strings, these operators don't add any new capabilities. They are provided for compatibility with other GNU software. @cindex buffer matching operators @table @code @cindex @code{\`} regexp operator @item \` This operator matches the empty string at the beginning of the buffer. @cindex @code{\'} regexp operator @item \' This operator matches the empty string at the end of the buffer. @end table In other GNU software, the word boundary operator is @samp{\b}. However, that conflicts with the @code{awk} language's definition of @samp{\b} as backspace, so @code{gawk} uses a different letter. An alternative method would have been to require two backslashes in the GNU operators, but this was deemed to be too confusing, and the current method of using @samp{\y} for the GNU @samp{\b} appears to be the lesser of two evils. @c NOTE!!! Keep this in sync with the same table in the summary appendix! @cindex regexp, effect of command line options The various command line options (@pxref{Options, ,Command Line Options}) control how @code{gawk} interprets characters in regexps. @table @asis @item No options In the default case, @code{gawk} provide all the facilities of POSIX regexps and the GNU regexp operators described @iftex above. @end iftex @ifinfo in @ref{Regexp Operators, ,Regular Expression Operators}. @end ifinfo However, interval expressions are not supported. @item @code{--posix} Only POSIX regexps are supported, the GNU operators are not special (e.g., @samp{\w} matches a literal @samp{w}). Interval expressions are allowed. @item @code{--traditional} Traditional Unix @code{awk} regexps are matched. The GNU operators are not special, interval expressions are not available, and neither are the POSIX character classes (@code{[[:alnum:]]} and so on). Characters described by octal and hexadecimal escape sequences are treated literally, even if they represent regexp metacharacters. @item @code{--re-interval} Allow interval expressions in regexps, even if @samp{--traditional} has been provided. @end table @node Case-sensitivity, Leftmost Longest, GNU Regexp Operators, Regexp @section Case-sensitivity in Matching @cindex case sensitivity @cindex ignoring case Case is normally significant in regular expressions, both when matching ordinary characters (i.e.@: not metacharacters), and inside character sets. Thus a @samp{w} in a regular expression matches only a lower-case @samp{w} and not an upper-case @samp{W}. The simplest way to do a case-independent match is to use a character list: @samp{[Ww]}. However, this can be cumbersome if you need to use it often; and it can make the regular expressions harder to read. There are two alternatives that you might prefer. One way to do a case-insensitive match at a particular point in the program is to convert the data to a single case, using the @code{tolower} or @code{toupper} built-in string functions (which we haven't discussed yet; @pxref{String Functions, ,Built-in Functions for String Manipulation}). For example: @example tolower($1) ~ /foo/ @{ @dots{} @} @end example @noindent converts the first field to lower-case before matching against it. This will work in any POSIX-compliant implementation of @code{awk}. @cindex differences between @code{gawk} and @code{awk} @cindex @code{~} operator @cindex @code{!~} operator @vindex IGNORECASE Another method, specific to @code{gawk}, is to set the variable @code{IGNORECASE} to a non-zero value (@pxref{Built-in Variables}). When @code{IGNORECASE} is not zero, @emph{all} regexp and string operations ignore case. Changing the value of @code{IGNORECASE} dynamically controls the case sensitivity of your program as it runs. Case is significant by default because @code{IGNORECASE} (like most variables) is initialized to zero. @example @group x = "aB" if (x ~ /ab/) @dots{} # this test will fail @end group @group IGNORECASE = 1 if (x ~ /ab/) @dots{} # now it will succeed @end group @end example In general, you cannot use @code{IGNORECASE} to make certain rules case-insensitive and other rules case-sensitive, because there is no way to set @code{IGNORECASE} just for the pattern of a particular rule. @ignore This isn't quite true. Consider: IGNORECASE=1 && /foObAr/ { .... } IGNORECASE=0 || /foobar/ { .... } But that's pretty bad style and I don't want to get into it at this late date. @end ignore To do this, you must use character lists or @code{tolower}. However, one thing you can do only with @code{IGNORECASE} is turn case-sensitivity on or off dynamically for all the rules at once. @code{IGNORECASE} can be set on the command line, or in a @code{BEGIN} rule (@pxref{Other Arguments, ,Other Command Line Arguments}; also @pxref{Using BEGIN/END, ,Startup and Cleanup Actions}). Setting @code{IGNORECASE} from the command line is a way to make a program case-insensitive without having to edit it. Prior to version 3.0 of @code{gawk}, the value of @code{IGNORECASE} only affected regexp operations. It did not affect string comparison with @samp{==}, @samp{!=}, and so on. Beginning with version 3.0, both regexp and string comparison operations are affected by @code{IGNORECASE}. @cindex ISO 8859-1 @cindex ISO Latin-1 Beginning with version 3.0 of @code{gawk}, the equivalences between upper-case and lower-case characters are based on the ISO-8859-1 (ISO Latin-1) character set. This character set is a superset of the traditional 128 ASCII characters, that also provides a number of characters suitable for use with European languages. @ignore A pure ASCII character set can be used instead if @code{gawk} is compiled with @samp{-DUSE_PURE_ASCII}. @end ignore The value of @code{IGNORECASE} has no effect if @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}). Case is always significant in compatibility mode. @node Leftmost Longest, Computed Regexps, Case-sensitivity, Regexp @section How Much Text Matches? @cindex leftmost longest match @cindex matching, leftmost longest Consider the following example: @example echo aaaabcd | awk '@{ sub(/a+/, ""); print @}' @end example This example uses the @code{sub} function (which we haven't discussed yet, @pxref{String Functions, ,Built-in Functions for String Manipulation}) to make a change to the input record. Here, the regexp @code{/a+/} indicates ``one or more @samp{a} characters,'' and the replacement text is @samp{}. The input contains four @samp{a} characters. What will the output be? In other words, how many is ``one or more''---will @code{awk} match two, three, or all four @samp{a} characters? The answer is, @code{awk} (and POSIX) regular expressions always match the leftmost, @emph{longest} sequence of input characters that can match. Thus, in this example, all four @samp{a} characters are replaced with @samp{}. @example $ echo aaaabcd | awk '@{ sub(/a+/, ""); print @}' @print{} bcd @end example For simple match/no-match tests, this is not so important. But when doing regexp-based field and record splitting, and text matching and substitutions with the @code{match}, @code{sub}, @code{gsub}, and @code{gensub} functions, it is very important. @ifinfo @xref{String Functions, ,Built-in Functions for String Manipulation}, for more information on these functions. @end ifinfo Understanding this principle is also important for regexp-based record and field splitting (@pxref{Records, ,How Input is Split into Records}, and also @pxref{Field Separators, ,Specifying How Fields are Separated}). @node Computed Regexps, , Leftmost Longest, Regexp @section Using Dynamic Regexps @cindex computed regular expressions @cindex regular expressions, computed @cindex dynamic regular expressions @cindex regexp, dynamic @cindex @code{~} operator @cindex @code{!~} operator The right hand side of a @samp{~} or @samp{!~} operator need not be a regexp constant (i.e.@: a string of characters between slashes). It may be any expression. The expression is evaluated, and converted if necessary to a string; the contents of the string are used as the regexp. A regexp that is computed in this way is called a @dfn{dynamic regexp}. For example: @example BEGIN @{ identifier_regexp = "[A-Za-z_][A-Za-z_0-9]+" @} $0 ~ identifier_regexp @{ print @} @end example @noindent sets @code{identifier_regexp} to a regexp that describes @code{awk} variable names, and tests if the input record matches this regexp. @strong{Caution:} When using the @samp{~} and @samp{!~} operators, there is a difference between a regexp constant enclosed in slashes, and a string constant enclosed in double quotes. If you are going to use a string constant, you have to understand that the string is in essence scanned @emph{twice}; the first time when @code{awk} reads your program, and the second time when it goes to match the string on the left-hand side of the operator with the pattern on the right. This is true of any string valued expression (such as @code{identifier_regexp} above), not just string constants. @cindex regexp constants, difference between slashes and quotes What difference does it make if the string is scanned twice? The answer has to do with escape sequences, and particularly with backslashes. To get a backslash into a regular expression inside a string, you have to type two backslashes. For example, @code{/\*/} is a regexp constant for a literal @samp{*}. Only one backslash is needed. To do the same thing with a string, you would have to type @code{"\\*"}. The first backslash escapes the second one, so that the string actually contains the two characters @samp{\} and @samp{*}. @cindex common mistakes @cindex mistakes, common @cindex errors, common Given that you can use both regexp and string constants to describe regular expressions, which should you use? The answer is ``regexp constants,'' for several reasons. @enumerate 1 @item String constants are more complicated to write, and more difficult to read. Using regexp constants makes your programs less error-prone. Not understanding the difference between the two kinds of constants is a common source of errors. @item It is also more efficient to use regexp constants: @code{awk} can note that you have supplied a regexp and store it internally in a form that makes pattern matching more efficient. When using a string constant, @code{awk} must first convert the string into this internal form, and then perform the pattern matching. @item Using regexp constants is better style; it shows clearly that you intend a regexp match. @end enumerate @node Reading Files, Printing, Regexp, Top @chapter Reading Input Files @cindex reading files @cindex input @cindex standard input @vindex FILENAME In the typical @code{awk} program, all input is read either from the standard input (by default the keyboard, but often a pipe from another command) or from files whose names you specify on the @code{awk} command line. If you specify input files, @code{awk} reads them in order, reading all the data from one before going on to the next. The name of the current input file can be found in the built-in variable @code{FILENAME} (@pxref{Built-in Variables}). The input is read in units called @dfn{records}, and processed by the rules of your program one record at a time. By default, each record is one line. Each record is automatically split into chunks called @dfn{fields}. This makes it more convenient for programs to work on the parts of a record. On rare occasions you will need to use the @code{getline} command. The @code{getline} command is valuable, both because it can do explicit input from any number of files, and because the files used with it do not have to be named on the @code{awk} command line (@pxref{Getline, ,Explicit Input with @code{getline}}). @menu * Records:: Controlling how data is split into records. * Fields:: An introduction to fields. * Non-Constant Fields:: Non-constant Field Numbers. * Changing Fields:: Changing the Contents of a Field. * Field Separators:: The field separator and how to change it. * Constant Size:: Reading constant width data. * Multiple Line:: Reading multi-line records. * Getline:: Reading files under explicit program control using the @code{getline} function. @end menu @node Records, Fields, Reading Files, Reading Files @section How Input is Split into Records @cindex record separator, @code{RS} @cindex changing the record separator @cindex record, definition of @vindex RS The @code{awk} utility divides the input for your @code{awk} program into records and fields. Records are separated by a character called the @dfn{record separator}. By default, the record separator is the newline character. This is why records are, by default, single lines. You can use a different character for the record separator by assigning the character to the built-in variable @code{RS}. You can change the value of @code{RS} in the @code{awk} program, like any other variable, with the assignment operator, @samp{=} (@pxref{Assignment Ops, ,Assignment Expressions}). The new record-separator character should be enclosed in quotation marks, which indicate a string constant. Often the right time to do this is at the beginning of execution, before any input has been processed, so that the very first record will be read with the proper separator. To do this, use the special @code{BEGIN} pattern (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}). For example: @example awk 'BEGIN @{ RS = "/" @} ; @{ print $0 @}' BBS-list @end example @noindent changes the value of @code{RS} to @code{"/"}, before reading any input. This is a string whose first character is a slash; as a result, records are separated by slashes. Then the input file is read, and the second rule in the @code{awk} program (the action with no pattern) prints each record. Since each @code{print} statement adds a newline at the end of its output, the effect of this @code{awk} program is to copy the input with each slash changed to a newline. Here are the results of running the program on @file{BBS-list}: @example @group $ awk 'BEGIN @{ RS = "/" @} ; @{ print $0 @}' BBS-list @print{} aardvark 555-5553 1200 @print{} 300 B @print{} alpo-net 555-3412 2400 @print{} 1200 @print{} 300 A @print{} barfly 555-7685 1200 @print{} 300 A @print{} bites 555-1675 2400 @print{} 1200 @print{} 300 A @print{} camelot 555-0542 300 C @print{} core 555-2912 1200 @print{} 300 C @print{} fooey 555-1234 2400 @print{} 1200 @print{} 300 B @print{} foot 555-6699 1200 @print{} 300 B @print{} macfoo 555-6480 1200 @print{} 300 A @print{} sdace 555-3430 2400 @print{} 1200 @print{} 300 A @print{} sabafoo 555-2127 1200 @print{} 300 C @print{} @end group @end example @noindent Note that the entry for the @samp{camelot} BBS is not split. In the original data file (@pxref{Sample Data Files, , Data Files for the Examples}), the line looks like this: @example camelot 555-0542 300 C @end example @noindent It only has one baud rate; there are no slashes in the record. Another way to change the record separator is on the command line, using the variable-assignment feature (@pxref{Other Arguments, ,Other Command Line Arguments}). @example awk '@{ print $0 @}' RS="/" BBS-list @end example @noindent This sets @code{RS} to @samp{/} before processing @file{BBS-list}. Using an unusual character such as @samp{/} for the record separator produces correct behavior in the vast majority of cases. However, the following (extreme) pipeline prints a surprising @samp{1}. There is one field, consisting of a newline. The value of the built-in variable @code{NF} is the number of fields in the current record. @example $ echo | awk 'BEGIN @{ RS = "a" @} ; @{ print NF @}' @print{} 1 @end example @cindex dark corner @noindent Reaching the end of an input file terminates the current input record, even if the last character in the file is not the character in @code{RS} (d.c.). @cindex empty string The empty string, @code{""} (a string of no characters), has a special meaning as the value of @code{RS}: it means that records are separated by one or more blank lines, and nothing else. @xref{Multiple Line, ,Multiple-Line Records}, for more details. If you change the value of @code{RS} in the middle of an @code{awk} run, the new value is used to delimit subsequent records, but the record currently being processed (and records already processed) are not affected. @vindex RT @cindex record terminator, @code{RT} @cindex terminator, record @cindex differences between @code{gawk} and @code{awk} After the end of the record has been determined, @code{gawk} sets the variable @code{RT} to the text in the input that matched @code{RS}. @cindex regular expressions as record separators The value of @code{RS} is in fact not limited to a one-character string. It can be any regular expression (@pxref{Regexp, ,Regular Expressions}). In general, each record ends at the next string that matches the regular expression; the next record starts at the end of the matching string. This general rule is actually at work in the usual case, where @code{RS} contains just a newline: a record ends at the beginning of the next matching string (the next newline in the input) and the following record starts just after the end of this string (at the first character of the following line). The newline, since it matches @code{RS}, is not part of either record. When @code{RS} is a single character, @code{RT} will contain the same single character. However, when @code{RS} is a regular expression, then @code{RT} becomes more useful; it contains the actual input text that matched the regular expression. The following example illustrates both of these features. It sets @code{RS} equal to a regular expression that matches either a newline, or a series of one or more upper-case letters with optional leading and/or trailing white space (@pxref{Regexp, , Regular Expressions}). @example $ echo record 1 AAAA record 2 BBBB record 3 | > gawk 'BEGIN @{ RS = "\n|( *[[:upper:]]+ *)" @} > @{ print "Record =", $0, "and RT =", RT @}' @print{} Record = record 1 and RT = AAAA @print{} Record = record 2 and RT = BBBB @print{} Record = record 3 and RT = @print{} @end example @noindent The final line of output has an extra blank line. This is because the value of @code{RT} is a newline, and then the @code{print} statement supplies its own terminating newline. @xref{Simple Sed, ,A Simple Stream Editor}, for a more useful example of @code{RS} as a regexp and @code{RT}. @cindex differences between @code{gawk} and @code{awk} The use of @code{RS} as a regular expression and the @code{RT} variable are @code{gawk} extensions; they are not available in compatibility mode (@pxref{Options, ,Command Line Options}). In compatibility mode, only the first character of the value of @code{RS} is used to determine the end of the record. @cindex number of records, @code{NR}, @code{FNR} @vindex NR @vindex FNR The @code{awk} utility keeps track of the number of records that have been read so far from the current input file. This value is stored in a built-in variable called @code{FNR}. It is reset to zero when a new file is started. Another built-in variable, @code{NR}, is the total number of input records read so far from all data files. It starts at zero but is never automatically reset to zero. @node Fields, Non-Constant Fields, Records, Reading Files @section Examining Fields @cindex examining fields @cindex fields @cindex accessing fields When @code{awk} reads an input record, the record is automatically separated or @dfn{parsed} by the interpreter into chunks called @dfn{fields}. By default, fields are separated by whitespace, like words in a line. Whitespace in @code{awk} means any string of one or more spaces, tabs or newlines;@footnote{In POSIX @code{awk}, newlines are not considered whitespace for separating fields.} other characters such as formfeed, and so on, that are considered whitespace by other languages are @emph{not} considered whitespace by @code{awk}. The purpose of fields is to make it more convenient for you to refer to these pieces of the record. You don't have to use them---you can operate on the whole record if you wish---but fields are what make simple @code{awk} programs so powerful. @cindex @code{$} (field operator) @cindex field operator @code{$} To refer to a field in an @code{awk} program, you use a dollar-sign, @samp{$}, followed by the number of the field you want. Thus, @code{$1} refers to the first field, @code{$2} to the second, and so on. For example, suppose the following is a line of input: @example This seems like a pretty nice example. @end example @noindent Here the first field, or @code{$1}, is @samp{This}; the second field, or @code{$2}, is @samp{seems}; and so on. Note that the last field, @code{$7}, is @samp{example.}. Because there is no space between the @samp{e} and the @samp{.}, the period is considered part of the seventh field. @vindex NF @cindex number of fields, @code{NF} @code{NF} is a built-in variable whose value is the number of fields in the current record. @code{awk} updates the value of @code{NF} automatically, each time a record is read. No matter how many fields there are, the last field in a record can be represented by @code{$NF}. So, in the example above, @code{$NF} would be the same as @code{$7}, which is @samp{example.}. Why this works is explained below (@pxref{Non-Constant Fields, ,Non-constant Field Numbers}). If you try to reference a field beyond the last one, such as @code{$8} when the record has only seven fields, you get the empty string. @c the empty string acts like 0 in some contexts, but I don't want to @c get into that here.... @code{$0}, which looks like a reference to the ``zeroth'' field, is a special case: it represents the whole input record. @code{$0} is used when you are not interested in fields. Here are some more examples: @example @group $ awk '$1 ~ /foo/ @{ print $0 @}' BBS-list @print{} fooey 555-1234 2400/1200/300 B @print{} foot 555-6699 1200/300 B @print{} macfoo 555-6480 1200/300 A @print{} sabafoo 555-2127 1200/300 C @end group @end example @noindent This example prints each record in the file @file{BBS-list} whose first field contains the string @samp{foo}. The operator @samp{~} is called a @dfn{matching operator} (@pxref{Regexp Usage, , How to Use Regular Expressions}); it tests whether a string (here, the field @code{$1}) matches a given regular expression. By contrast, the following example looks for @samp{foo} in @emph{the entire record} and prints the first field and the last field for each input record containing a match. @example @group $ awk '/foo/ @{ print $1, $NF @}' BBS-list @print{} fooey B @print{} foot B @print{} macfoo A @print{} sabafoo C @end group @end example @node Non-Constant Fields, Changing Fields, Fields, Reading Files @section Non-constant Field Numbers The number of a field does not need to be a constant. Any expression in the @code{awk} language can be used after a @samp{$} to refer to a field. The value of the expression specifies the field number. If the value is a string, rather than a number, it is converted to a number. Consider this example: @example awk '@{ print $NR @}' @end example @noindent Recall that @code{NR} is the number of records read so far: one in the first record, two in the second, etc. So this example prints the first field of the first record, the second field of the second record, and so on. For the twentieth record, field number 20 is printed; most likely, the record has fewer than 20 fields, so this prints a blank line. Here is another example of using expressions as field numbers: @example awk '@{ print $(2*2) @}' BBS-list @end example @code{awk} must evaluate the expression @samp{(2*2)} and use its value as the number of the field to print. The @samp{*} sign represents multiplication, so the expression @samp{2*2} evaluates to four. The parentheses are used so that the multiplication is done before the @samp{$} operation; they are necessary whenever there is a binary operator in the field-number expression. This example, then, prints the hours of operation (the fourth field) for every line of the file @file{BBS-list}. (All of the @code{awk} operators are listed, in order of decreasing precedence, in @ref{Precedence, , Operator Precedence (How Operators Nest)}.) If the field number you compute is zero, you get the entire record. Thus, @code{$(2-2)} has the same value as @code{$0}. Negative field numbers are not allowed; trying to reference one will usually terminate your running @code{awk} program. (The POSIX standard does not define what happens when you reference a negative field number. @code{gawk} will notice this and terminate your program. Other @code{awk} implementations may behave differently.) As mentioned in @ref{Fields, ,Examining Fields}, the number of fields in the current record is stored in the built-in variable @code{NF} (also @pxref{Built-in Variables}). The expression @code{$NF} is not a special feature: it is the direct consequence of evaluating @code{NF} and using its value as a field number. @node Changing Fields, Field Separators, Non-Constant Fields, Reading Files @section Changing the Contents of a Field @cindex field, changing contents of @cindex changing contents of a field @cindex assignment to fields You can change the contents of a field as seen by @code{awk} within an @code{awk} program; this changes what @code{awk} perceives as the current input record. (The actual input is untouched; @code{awk} @emph{never} modifies the input file.) Consider this example and its output: @example @group $ awk '@{ $3 = $2 - 10; print $2, $3 @}' inventory-shipped @print{} 13 3 @print{} 15 5 @print{} 15 5 @dots{} @end group @end example @noindent The @samp{-} sign represents subtraction, so this program reassigns field three, @code{$3}, to be the value of field two minus ten, @samp{$2 - 10}. (@xref{Arithmetic Ops, ,Arithmetic Operators}.) Then field two, and the new value for field three, are printed. In order for this to work, the text in field @code{$2} must make sense as a number; the string of characters must be converted to a number in order for the computer to do arithmetic on it. The number resulting from the subtraction is converted back to a string of characters which then becomes field three. @xref{Conversion, ,Conversion of Strings and Numbers}. When you change the value of a field (as perceived by @code{awk}), the text of the input record is recalculated to contain the new field where the old one was. Therefore, @code{$0} changes to reflect the altered field. Thus, this program prints a copy of the input file, with 10 subtracted from the second field of each line. @example @group $ awk '@{ $2 = $2 - 10; print $0 @}' inventory-shipped @print{} Jan 3 25 15 115 @print{} Feb 5 32 24 226 @print{} Mar 5 24 34 228 @dots{} @end group @end example You can also assign contents to fields that are out of range. For example: @example $ awk '@{ $6 = ($5 + $4 + $3 + $2) > print $6 @}' inventory-shipped @print{} 168 @print{} 297 @print{} 301 @dots{} @end example @noindent We've just created @code{$6}, whose value is the sum of fields @code{$2}, @code{$3}, @code{$4}, and @code{$5}. The @samp{+} sign represents addition. For the file @file{inventory-shipped}, @code{$6} represents the total number of parcels shipped for a particular month. Creating a new field changes @code{awk}'s internal copy of the current input record---the value of @code{$0}. Thus, if you do @samp{print $0} after adding a field, the record printed includes the new field, with the appropriate number of field separators between it and the previously existing fields. This recomputation affects and is affected by @code{NF} (the number of fields; @pxref{Fields, ,Examining Fields}), and by a feature that has not been discussed yet, the @dfn{output field separator}, @code{OFS}, which is used to separate the fields (@pxref{Output Separators}). For example, the value of @code{NF} is set to the number of the highest field you create. Note, however, that merely @emph{referencing} an out-of-range field does @emph{not} change the value of either @code{$0} or @code{NF}. Referencing an out-of-range field only produces an empty string. For example: @example if ($(NF+1) != "") print "can't happen" else print "everything is normal" @end example @noindent should print @samp{everything is normal}, because @code{NF+1} is certain to be out of range. (@xref{If Statement, ,The @code{if}-@code{else} Statement}, for more information about @code{awk}'s @code{if-else} statements. @xref{Typing and Comparison, ,Variable Typing and Comparison Expressions}, for more information about the @samp{!=} operator.) It is important to note that making an assignment to an existing field will change the value of @code{$0}, but will not change the value of @code{NF}, even when you assign the empty string to a field. For example: @example @group $ echo a b c d | awk '@{ OFS = ":"; $2 = "" > print $0; print NF @}' @print{} a::c:d @print{} 4 @end group @end example @noindent The field is still there; it just has an empty value. You can tell because there are two colons in a row. This example shows what happens if you create a new field. @example $ echo a b c d | awk '@{ OFS = ":"; $2 = ""; $6 = "new" > print $0; print NF @}' @print{} a::c:d::new @print{} 6 @end example @noindent The intervening field, @code{$5} is created with an empty value (indicated by the second pair of adjacent colons), and @code{NF} is updated with the value six. Finally, decrementing @code{NF} will lose the values of the fields after the new value of @code{NF}, and @code{$0} will be recomputed. Here is an example: @example $ echo a b c d e f | ../gawk '@{ print "NF =", NF; > NF = 3; print $0 @}' @print{} NF = 6 @print{} a b c @end example @node Field Separators, Constant Size, Changing Fields, Reading Files @section Specifying How Fields are Separated This section is rather long; it describes one of the most fundamental operations in @code{awk}. @menu * Basic Field Splitting:: How fields are split with single characters or simple strings. * Regexp Field Splitting:: Using regexps as the field separator. * Single Character Fields:: Making each character a separate field. * Command Line Field Separator:: Setting @code{FS} from the command line. * Field Splitting Summary:: Some final points and a summary table. @end menu @node Basic Field Splitting, Regexp Field Splitting, Field Separators, Field Separators @subsection The Basics of Field Separating @vindex FS @cindex fields, separating @cindex field separator, @code{FS} The @dfn{field separator}, which is either a single character or a regular expression, controls the way @code{awk} splits an input record into fields. @code{awk} scans the input record for character sequences that match the separator; the fields themselves are the text between the matches. In the examples below, we use the bullet symbol ``@bullet{}'' to represent spaces in the output. If the field separator is @samp{oo}, then the following line: @example moo goo gai pan @end example @noindent would be split into three fields: @samp{m}, @samp{@bullet{}g} and @samp{@bullet{}gai@bullet{}pan}. Note the leading spaces in the values of the second and third fields. @cindex common mistakes @cindex mistakes, common @cindex errors, common The field separator is represented by the built-in variable @code{FS}. Shell programmers take note! @code{awk} does @emph{not} use the name @code{IFS} which is used by the POSIX compatible shells (such as the Bourne shell, @code{sh}, or the GNU Bourne-Again Shell, Bash). You can change the value of @code{FS} in the @code{awk} program with the assignment operator, @samp{=} (@pxref{Assignment Ops, ,Assignment Expressions}). Often the right time to do this is at the beginning of execution, before any input has been processed, so that the very first record will be read with the proper separator. To do this, use the special @code{BEGIN} pattern (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}). For example, here we set the value of @code{FS} to the string @code{","}: @example awk 'BEGIN @{ FS = "," @} ; @{ print $2 @}' @end example @noindent Given the input line, @example John Q. Smith, 29 Oak St., Walamazoo, MI 42139 @end example @noindent this @code{awk} program extracts and prints the string @samp{@bullet{}29@bullet{}Oak@bullet{}St.}. @cindex field separator, choice of @cindex regular expressions as field separators Sometimes your input data will contain separator characters that don't separate fields the way you thought they would. For instance, the person's name in the example we just used might have a title or suffix attached, such as @samp{John Q. Smith, LXIX}. From input containing such a name: @example John Q. Smith, LXIX, 29 Oak St., Walamazoo, MI 42139 @end example @noindent @c careful of an overfull hbox here! the above program would extract @samp{@bullet{}LXIX}, instead of @samp{@bullet{}29@bullet{}Oak@bullet{}St.}. If you were expecting the program to print the address, you would be surprised. The moral is: choose your data layout and separator characters carefully to prevent such problems. @iftex As you know, normally, @end iftex @ifinfo Normally, @end ifinfo fields are separated by whitespace sequences (spaces, tabs and newlines), not by single spaces: two spaces in a row do not delimit an empty field. The default value of the field separator @code{FS} is a string containing a single space, @w{@code{" "}}. If this value were interpreted in the usual way, each space character would separate fields, so two spaces in a row would make an empty field between them. The reason this does not happen is that a single space as the value of @code{FS} is a special case: it is taken to specify the default manner of delimiting fields. If @code{FS} is any other single character, such as @code{","}, then each occurrence of that character separates two fields. Two consecutive occurrences delimit an empty field. If the character occurs at the beginning or the end of the line, that too delimits an empty field. The space character is the only single character which does not follow these rules. @node Regexp Field Splitting, Single Character Fields, Basic Field Splitting, Field Separators @subsection Using Regular Expressions to Separate Fields The previous @iftex subsection @end iftex @ifinfo node @end ifinfo discussed the use of single characters or simple strings as the value of @code{FS}. More generally, the value of @code{FS} may be a string containing any regular expression. In this case, each match in the record for the regular expression separates fields. For example, the assignment: @example FS = ", \t" @end example @noindent makes every area of an input line that consists of a comma followed by a space and a tab, into a field separator. (@samp{\t} is an @dfn{escape sequence} that stands for a tab; @pxref{Escape Sequences}, for the complete list of similar escape sequences.) For a less trivial example of a regular expression, suppose you want single spaces to separate fields the way single commas were used above. You can set @code{FS} to @w{@code{"[@ ]"}} (left bracket, space, right bracket). This regular expression matches a single space and nothing else (@pxref{Regexp, ,Regular Expressions}). There is an important difference between the two cases of @samp{FS = @w{" "}} (a single space) and @samp{FS = @w{"[ \t\n]+"}} (left bracket, space, backslash, ``t'', backslash, ``n'', right bracket, which is a regular expression matching one or more spaces, tabs, or newlines). For both values of @code{FS}, fields are separated by runs of spaces, tabs and/or newlines. However, when the value of @code{FS} is @w{@code{" "}}, @code{awk} will first strip leading and trailing whitespace from the record, and then decide where the fields are. For example, the following pipeline prints @samp{b}: @example $ echo ' a b c d ' | awk '@{ print $2 @}' @print{} b @end example @noindent However, this pipeline prints @samp{a} (note the extra spaces around each letter): @example $ echo ' a b c d ' | awk 'BEGIN @{ FS = "[ \t]+" @} > @{ print $2 @}' @print{} a @end example @noindent @cindex null string @cindex empty string In this case, the first field is @dfn{null}, or empty. The stripping of leading and trailing whitespace also comes into play whenever @code{$0} is recomputed. For instance, study this pipeline: @example $ echo ' a b c d' | awk '@{ print; $2 = $2; print @}' @print{} a b c d @print{} a b c d @end example @noindent The first @code{print} statement prints the record as it was read, with leading whitespace intact. The assignment to @code{$2} rebuilds @code{$0} by concatenating @code{$1} through @code{$NF} together, separated by the value of @code{OFS}. Since the leading whitespace was ignored when finding @code{$1}, it is not part of the new @code{$0}. Finally, the last @code{print} statement prints the new @code{$0}. @node Single Character Fields, Command Line Field Separator, Regexp Field Splitting, Field Separators @subsection Making Each Character a Separate Field @cindex differences between @code{gawk} and @code{awk} @cindex single character fields There are times when you may want to examine each character of a record separately. In @code{gawk}, this is easy to do, you simply assign the null string (@code{""}) to @code{FS}. In this case, each individual character in the record will become a separate field. Here is an example: @example @group $ echo a b | gawk 'BEGIN @{ FS = "" @} > @{ > for (i = 1; i <= NF; i = i + 1) > print "Field", i, "is", $i > @}' @print{} Field 1 is a @print{} Field 2 is @print{} Field 3 is b @end group @end example @cindex dark corner Traditionally, the behavior for @code{FS} equal to @code{""} was not defined. In this case, Unix @code{awk} would simply treat the entire record as only having one field (d.c.). In compatibility mode (@pxref{Options, ,Command Line Options}), if @code{FS} is the null string, then @code{gawk} will also behave this way. @node Command Line Field Separator, Field Splitting Summary, Single Character Fields, Field Separators @subsection Setting @code{FS} from the Command Line @cindex @code{-F} option @cindex field separator, on command line @cindex command line, setting @code{FS} on @code{FS} can be set on the command line. You use the @samp{-F} option to do so. For example: @example awk -F, '@var{program}' @var{input-files} @end example @noindent sets @code{FS} to be the @samp{,} character. Notice that the option uses a capital @samp{F}. Contrast this with @samp{-f}, which specifies a file containing an @code{awk} program. Case is significant in command line options: the @samp{-F} and @samp{-f} options have nothing to do with each other. You can use both options at the same time to set the @code{FS} variable @emph{and} get an @code{awk} program from a file. The value used for the argument to @samp{-F} is processed in exactly the same way as assignments to the built-in variable @code{FS}. This means that if the field separator contains special characters, they must be escaped appropriately. For example, to use a @samp{\} as the field separator, you would have to type: @example # same as FS = "\\" awk -F\\\\ '@dots{}' files @dots{} @end example @noindent Since @samp{\} is used for quoting in the shell, @code{awk} will see @samp{-F\\}. Then @code{awk} processes the @samp{\\} for escape characters (@pxref{Escape Sequences}), finally yielding a single @samp{\} to be used for the field separator. @cindex historical features As a special case, in compatibility mode (@pxref{Options, ,Command Line Options}), if the argument to @samp{-F} is @samp{t}, then @code{FS} is set to the tab character. This is because if you type @samp{-F\t} at the shell, without any quotes, the @samp{\} gets deleted, so @code{awk} figures that you really want your fields to be separated with tabs, and not @samp{t}s. Use @samp{-v FS="t"} on the command line if you really do want to separate your fields with @samp{t}s (@pxref{Options, ,Command Line Options}). For example, let's use an @code{awk} program file called @file{baud.awk} that contains the pattern @code{/300/}, and the action @samp{print $1}. Here is the program: @example /300/ @{ print $1 @} @end example Let's also set @code{FS} to be the @samp{-} character, and run the program on the file @file{BBS-list}. The following command prints a list of the names of the bulletin boards that operate at 300 baud and the first three digits of their phone numbers: @c tweaked to make the tex output look better in @smallbook @example @group $ awk -F- -f baud.awk BBS-list @print{} aardvark 555 @print{} alpo @print{} barfly 555 @dots{} @end group @ignore @print{} bites 555 @print{} camelot 555 @print{} core 555 @print{} fooey 555 @print{} foot 555 @print{} macfoo 555 @print{} sdace 555 @print{} sabafoo 555 @end ignore @end example @noindent Note the second line of output. In the original file (@pxref{Sample Data Files, ,Data Files for the Examples}), the second line looked like this: @example alpo-net 555-3412 2400/1200/300 A @end example The @samp{-} as part of the system's name was used as the field separator, instead of the @samp{-} in the phone number that was originally intended. This demonstrates why you have to be careful in choosing your field and record separators. On many Unix systems, each user has a separate entry in the system password file, one line per user. The information in these lines is separated by colons. The first field is the user's logon name, and the second is the user's encrypted password. A password file entry might look like this: @example arnold:xyzzy:2076:10:Arnold Robbins:/home/arnold:/bin/sh @end example The following program searches the system password file, and prints the entries for users who have no password: @example awk -F: '$2 == ""' /etc/passwd @end example @node Field Splitting Summary, , Command Line Field Separator, Field Separators @subsection Field Splitting Summary @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} According to the POSIX standard, @code{awk} is supposed to behave as if each record is split into fields at the time that it is read. In particular, this means that you can change the value of @code{FS} after a record is read, and the value of the fields (i.e.@: how they were split) should reflect the old value of @code{FS}, not the new one. @cindex dark corner @cindex @code{sed} utility @cindex stream editor However, many implementations of @code{awk} do not work this way. Instead, they defer splitting the fields until a field is actually referenced. The fields will be split using the @emph{current} value of @code{FS}! (d.c.) This behavior can be difficult to diagnose. The following example illustrates the difference between the two methods. (The @code{sed}@footnote{The @code{sed} utility is a ``stream editor.'' Its behavior is also defined by the POSIX standard.} command prints just the first line of @file{/etc/passwd}.) @example sed 1q /etc/passwd | awk '@{ FS = ":" ; print $1 @}' @end example @noindent will usually print @example root @end example @noindent on an incorrect implementation of @code{awk}, while @code{gawk} will print something like @example root:nSijPlPhZZwgE:0:0:Root:/: @end example The following table summarizes how fields are split, based on the value of @code{FS}. (@samp{==} means ``is equal to.'') @c @cartouche @table @code @item FS == " " Fields are separated by runs of whitespace. Leading and trailing whitespace are ignored. This is the default. @item FS == @var{any other single character} Fields are separated by each occurrence of the character. Multiple successive occurrences delimit empty fields, as do leading and trailing occurrences. The character can even be a regexp metacharacter; it does not need to be escaped. @item FS == @var{regexp} Fields are separated by occurrences of characters that match @var{regexp}. Leading and trailing matches of @var{regexp} delimit empty fields. @item FS == "" Each individual character in the record becomes a separate field. @end table @c @end cartouche @node Constant Size, Multiple Line, Field Separators, Reading Files @section Reading Fixed-width Data (This section discusses an advanced, experimental feature. If you are a novice @code{awk} user, you may wish to skip it on the first reading.) @code{gawk} version 2.13 introduced a new facility for dealing with fixed-width fields with no distinctive field separator. Data of this nature arises, for example, in the input for old FORTRAN programs where numbers are run together; or in the output of programs that did not anticipate the use of their output as input for other programs. An example of the latter is a table where all the columns are lined up by the use of a variable number of spaces and @emph{empty fields are just spaces}. Clearly, @code{awk}'s normal field splitting based on @code{FS} will not work well in this case. Although a portable @code{awk} program can use a series of @code{substr} calls on @code{$0} (@pxref{String Functions, ,Built-in Functions for String Manipulation}), this is awkward and inefficient for a large number of fields. The splitting of an input record into fixed-width fields is specified by assigning a string containing space-separated numbers to the built-in variable @code{FIELDWIDTHS}. Each number specifies the width of the field @emph{including} columns between fields. If you want to ignore the columns between fields, you can specify the width as a separate field that is subsequently ignored. The following data is the output of the Unix @code{w} utility. It is useful to illustrate the use of @code{FIELDWIDTHS}. @example @group 10:06pm up 21 days, 14:04, 23 users User tty login@ idle JCPU PCPU what hzuo ttyV0 8:58pm 9 5 vi p24.tex hzang ttyV3 6:37pm 50 -csh eklye ttyV5 9:53pm 7 1 em thes.tex dportein ttyV6 8:17pm 1:47 -csh gierd ttyD3 10:00pm 1 elm dave ttyD4 9:47pm 4 4 w brent ttyp0 26Jun91 4:46 26:46 4:41 bash dave ttyq4 26Jun9115days 46 46 wnewmail @end group @end example The following program takes the above input, converts the idle time to number of seconds and prints out the first two fields and the calculated idle time. (This program uses a number of @code{awk} features that haven't been introduced yet.) @example @group BEGIN @{ FIELDWIDTHS = "9 6 10 6 7 7 35" @} NR > 2 @{ idle = $4 sub(/^ */, "", idle) # strip leading spaces if (idle == "") idle = 0 if (idle ~ /:/) @{ split(idle, t, ":") idle = t[1] * 60 + t[2] @} if (idle ~ /days/) idle *= 24 * 60 * 60 print $1, $2, idle @} @end group @end example Here is the result of running the program on the data: @example hzuo ttyV0 0 hzang ttyV3 50 eklye ttyV5 0 dportein ttyV6 107 gierd ttyD3 1 dave ttyD4 0 brent ttyp0 286 dave ttyq4 1296000 @end example Another (possibly more practical) example of fixed-width input data would be the input from a deck of balloting cards. In some parts of the United States, voters mark their choices by punching holes in computer cards. These cards are then processed to count the votes for any particular candidate or on any particular issue. Since a voter may choose not to vote on some issue, any column on the card may be empty. An @code{awk} program for processing such data could use the @code{FIELDWIDTHS} feature to simplify reading the data. (Of course, getting @code{gawk} to run on a system with card readers is another story!) @ignore Exercise: Write a ballot card reading program @end ignore Assigning a value to @code{FS} causes @code{gawk} to return to using @code{FS} for field splitting. Use @samp{FS = FS} to make this happen, without having to know the current value of @code{FS}. This feature is still experimental, and may evolve over time. Note that in particular, @code{gawk} does not attempt to verify the sanity of the values used in the value of @code{FIELDWIDTHS}. @node Multiple Line, Getline, Constant Size, Reading Files @section Multiple-Line Records @cindex multiple line records @cindex input, multiple line records @cindex reading files, multiple line records @cindex records, multiple line In some data bases, a single line cannot conveniently hold all the information in one entry. In such cases, you can use multi-line records. The first step in doing this is to choose your data format: when records are not defined as single lines, how do you want to define them? What should separate records? One technique is to use an unusual character or string to separate records. For example, you could use the formfeed character (written @samp{\f} in @code{awk}, as in C) to separate them, making each record a page of the file. To do this, just set the variable @code{RS} to @code{"\f"} (a string containing the formfeed character). Any other character could equally well be used, as long as it won't be part of the data in a record. Another technique is to have blank lines separate records. By a special dispensation, an empty string as the value of @code{RS} indicates that records are separated by one or more blank lines. If you set @code{RS} to the empty string, a record always ends at the first blank line encountered. And the next record doesn't start until the first non-blank line that follows---no matter how many blank lines appear in a row, they are considered one record-separator. @cindex leftmost longest match @cindex matching, leftmost longest You can achieve the same effect as @samp{RS = ""} by assigning the string @code{"\n\n+"} to @code{RS}. This regexp matches the newline at the end of the record, and one or more blank lines after the record. In addition, a regular expression always matches the longest possible sequence when there is a choice (@pxref{Leftmost Longest, ,How Much Text Matches?}). So the next record doesn't start until the first non-blank line that follows---no matter how many blank lines appear in a row, they are considered one record-separator. @cindex dark corner There is an important difference between @samp{RS = ""} and @samp{RS = "\n\n+"}. In the first case, leading newlines in the input data file are ignored, and if a file ends without extra blank lines after the last record, the final newline is removed from the record. In the second case, this special processing is not done (d.c.). Now that the input is separated into records, the second step is to separate the fields in the record. One way to do this is to divide each of the lines into fields in the normal manner. This happens by default as the result of a special feature: when @code{RS} is set to the empty string, the newline character @emph{always} acts as a field separator. This is in addition to whatever field separations result from @code{FS}. The original motivation for this special exception was probably to provide useful behavior in the default case (i.e.@: @code{FS} is equal to @w{@code{" "}}). This feature can be a problem if you really don't want the newline character to separate fields, since there is no way to prevent it. However, you can work around this by using the @code{split} function to break up the record manually (@pxref{String Functions, ,Built-in Functions for String Manipulation}). Another way to separate fields is to put each field on a separate line: to do this, just set the variable @code{FS} to the string @code{"\n"}. (This simple regular expression matches a single newline.) A practical example of a data file organized this way might be a mailing list, where each entry is separated by blank lines. If we have a mailing list in a file named @file{addresses}, that looks like this: @example Jane Doe 123 Main Street Anywhere, SE 12345-6789 John Smith 456 Tree-lined Avenue Smallville, MW 98765-4321 @dots{} @end example @noindent A simple program to process this file would look like this: @example @group # addrs.awk --- simple mailing list program # Records are separated by blank lines. # Each line is one field. BEGIN @{ RS = "" ; FS = "\n" @} @{ print "Name is:", $1 print "Address is:", $2 print "City and State are:", $3 print "" @} @end group @end example Running the program produces the following output: @example @group $ awk -f addrs.awk addresses @print{} Name is: Jane Doe @print{} Address is: 123 Main Street @print{} City and State are: Anywhere, SE 12345-6789 @print{} @end group @group @print{} Name is: John Smith @print{} Address is: 456 Tree-lined Avenue @print{} City and State are: Smallville, MW 98765-4321 @print{} @dots{} @end group @end example @xref{Labels Program, ,Printing Mailing Labels}, for a more realistic program that deals with address lists. The following table summarizes how records are split, based on the value of @code{RS}. (@samp{==} means ``is equal to.'') @c @cartouche @table @code @item RS == "\n" Records are separated by the newline character (@samp{\n}). In effect, every line in the data file is a separate record, including blank lines. This is the default. @item RS == @var{any single character} Records are separated by each occurrence of the character. Multiple successive occurrences delimit empty records. @item RS == "" Records are separated by runs of blank lines. The newline character always serves as a field separator, in addition to whatever value @code{FS} may have. Leading and trailing newlines in a file are ignored. @item RS == @var{regexp} Records are separated by occurrences of characters that match @var{regexp}. Leading and trailing matches of @var{regexp} delimit empty records. @end table @c @end cartouche @vindex RT In all cases, @code{gawk} sets @code{RT} to the input text that matched the value specified by @code{RS}. @node Getline, , Multiple Line, Reading Files @section Explicit Input with @code{getline} @findex getline @cindex input, explicit @cindex explicit input @cindex input, @code{getline} command @cindex reading files, @code{getline} command So far we have been getting our input data from @code{awk}'s main input stream---either the standard input (usually your terminal, sometimes the output from another program) or from the files specified on the command line. The @code{awk} language has a special built-in command called @code{getline} that can be used to read input under your explicit control. @menu * Getline Intro:: Introduction to the @code{getline} function. * Plain Getline:: Using @code{getline} with no arguments. * Getline/Variable:: Using @code{getline} into a variable. * Getline/File:: Using @code{getline} from a file. * Getline/Variable/File:: Using @code{getline} into a variable from a file. * Getline/Pipe:: Using @code{getline} from a pipe. * Getline/Variable/Pipe:: Using @code{getline} into a variable from a pipe. * Getline Summary:: Summary Of @code{getline} Variants. @end menu @node Getline Intro, Plain Getline, Getline, Getline @subsection Introduction to @code{getline} This command is used in several different ways, and should @emph{not} be used by beginners. It is covered here because this is the chapter on input. The examples that follow the explanation of the @code{getline} command include material that has not been covered yet. Therefore, come back and study the @code{getline} command @emph{after} you have reviewed the rest of this @value{DOCUMENT} and have a good knowledge of how @code{awk} works. @vindex ERRNO @cindex differences between @code{gawk} and @code{awk} @cindex @code{getline}, return values @code{getline} returns one if it finds a record, and zero if the end of the file is encountered. If there is some error in getting a record, such as a file that cannot be opened, then @code{getline} returns @minus{}1. In this case, @code{gawk} sets the variable @code{ERRNO} to a string describing the error that occurred. In the following examples, @var{command} stands for a string value that represents a shell command. @node Plain Getline, Getline/Variable, Getline Intro, Getline @subsection Using @code{getline} with No Arguments The @code{getline} command can be used without arguments to read input from the current input file. All it does in this case is read the next input record and split it up into fields. This is useful if you've finished processing the current record, but you want to do some special processing @emph{right now} on the next record. Here's an example: @example @group awk '@{ if ((t = index($0, "/*")) != 0) @{ # value will be "" if t is 1 tmp = substr($0, 1, t - 1) u = index(substr($0, t + 2), "*/") while (u == 0) @{ if (getline <= 0) @{ m = "unexpected EOF or error" m = (m ": " ERRNO) print m > "/dev/stderr" exit @} t = -1 u = index($0, "*/") @} @end group @group # substr expression will be "" if */ # occurred at end of line $0 = tmp substr($0, t + u + 3) @} print $0 @}' @end group @end example This @code{awk} program deletes all C-style comments, @samp{/* @dots{} */}, from the input. By replacing the @samp{print $0} with other statements, you could perform more complicated processing on the decommented input, like searching for matches of a regular expression. This program has a subtle problem---it does not work if one comment ends and another begins on the same line. @ignore Exercise, write a program that does handle multiple comments on the line. @end ignore This form of the @code{getline} command sets @code{NF} (the number of fields; @pxref{Fields, ,Examining Fields}), @code{NR} (the number of records read so far; @pxref{Records, ,How Input is Split into Records}), @code{FNR} (the number of records read from this input file), and the value of @code{$0}. @cindex dark corner @strong{Note:} the new value of @code{$0} is used in testing the patterns of any subsequent rules. The original value of @code{$0} that triggered the rule which executed @code{getline} is lost (d.c.). By contrast, the @code{next} statement reads a new record but immediately begins processing it normally, starting with the first rule in the program. @xref{Next Statement, ,The @code{next} Statement}. @node Getline/Variable, Getline/File, Plain Getline, Getline @subsection Using @code{getline} Into a Variable You can use @samp{getline @var{var}} to read the next record from @code{awk}'s input into the variable @var{var}. No other processing is done. For example, suppose the next line is a comment, or a special string, and you want to read it, without triggering any rules. This form of @code{getline} allows you to read that line and store it in a variable so that the main read-a-line-and-check-each-rule loop of @code{awk} never sees it. The following example swaps every two lines of input. For example, given: @example wan tew free phore @end example @noindent it outputs: @example tew wan phore free @end example @noindent Here's the program: @example @group awk '@{ if ((getline tmp) > 0) @{ print tmp print $0 @} else print $0 @}' @end group @end example The @code{getline} command used in this way sets only the variables @code{NR} and @code{FNR} (and of course, @var{var}). The record is not split into fields, so the values of the fields (including @code{$0}) and the value of @code{NF} do not change. @node Getline/File, Getline/Variable/File, Getline/Variable, Getline @subsection Using @code{getline} from a File @cindex input redirection @cindex redirection of input Use @samp{getline < @var{file}} to read the next record from the file @var{file}. Here @var{file} is a string-valued expression that specifies the file name. @samp{< @var{file}} is called a @dfn{redirection} since it directs input to come from a different place. For example, the following program reads its input record from the file @file{secondary.input} when it encounters a first field with a value equal to 10 in the current input file. @example @group awk '@{ if ($1 == 10) @{ getline < "secondary.input" print @} else print @}' @end group @end example Since the main input stream is not used, the values of @code{NR} and @code{FNR} are not changed. But the record read is split into fields in the normal manner, so the values of @code{$0} and other fields are changed. So is the value of @code{NF}. @c Thanks to Paul Eggert for initial wording here According to POSIX, @samp{getline < @var{expression}} is ambiguous if @var{expression} contains unparenthesized operators other than @samp{$}; for example, @samp{getline < dir "/" file} is ambiguous because the concatenation operator is not parenthesized, and you should write it as @samp{getline < (dir "/" file)} if you want your program to be portable to other @code{awk} implementations. @node Getline/Variable/File, Getline/Pipe, Getline/File, Getline @subsection Using @code{getline} Into a Variable from a File Use @samp{getline @var{var} < @var{file}} to read input the file @var{file} and put it in the variable @var{var}. As above, @var{file} is a string-valued expression that specifies the file from which to read. In this version of @code{getline}, none of the built-in variables are changed, and the record is not split into fields. The only variable changed is @var{var}. @ifinfo @c Thanks to Paul Eggert for initial wording here According to POSIX, @samp{getline @var{var} < @var{expression}} is ambiguous if @var{expression} contains unparenthesized operators other than @samp{$}; for example, @samp{getline < dir "/" file} is ambiguous because the concatenation operator is not parenthesized, and you should write it as @samp{getline < (dir "/" file)} if you want your program to be portable to other @code{awk} implementations. @end ifinfo For example, the following program copies all the input files to the output, except for records that say @w{@samp{@@include @var{filename}}}. Such a record is replaced by the contents of the file @var{filename}. @example @group awk '@{ if (NF == 2 && $1 == "@@include") @{ while ((getline line < $2) > 0) print line close($2) @} else print @}' @end group @end example Note here how the name of the extra input file is not built into the program; it is taken directly from the data, from the second field on the @samp{@@include} line. The @code{close} function is called to ensure that if two identical @samp{@@include} lines appear in the input, the entire specified file is included twice. @xref{Close Files And Pipes, ,Closing Input and Output Files and Pipes}. One deficiency of this program is that it does not process nested @samp{@@include} statements (@samp{@@include} statements in included files) the way a true macro preprocessor would. @xref{Igawk Program, ,An Easy Way to Use Library Functions}, for a program that does handle nested @samp{@@include} statements. @node Getline/Pipe, Getline/Variable/Pipe, Getline/Variable/File, Getline @subsection Using @code{getline} from a Pipe @cindex input pipeline @cindex pipeline, input You can pipe the output of a command into @code{getline}, using @samp{@var{command} | getline}. In this case, the string @var{command} is run as a shell command and its output is piped into @code{awk} to be used as input. This form of @code{getline} reads one record at a time from the pipe. For example, the following program copies its input to its output, except for lines that begin with @samp{@@execute}, which are replaced by the output produced by running the rest of the line as a shell command: @example @group awk '@{ if ($1 == "@@execute") @{ tmp = substr($0, 10) while ((tmp | getline) > 0) print close(tmp) @} else print @}' @end group @end example @noindent The @code{close} function is called to ensure that if two identical @samp{@@execute} lines appear in the input, the command is run for each one. @xref{Close Files And Pipes, ,Closing Input and Output Files and Pipes}. @c Exercise!! @c This example is unrealistic, since you could just use system @c NEEDED @page Given the input: @example @group foo bar baz @@execute who bletch @end group @end example @noindent the program might produce: @example @group foo bar baz arnold ttyv0 Jul 13 14:22 miriam ttyp0 Jul 13 14:23 (murphy:0) bill ttyp1 Jul 13 14:23 (murphy:0) bletch @end group @end example @noindent Notice that this program ran the command @code{who} and printed the result. (If you try this program yourself, you will of course get different results, showing you who is logged in on your system.) This variation of @code{getline} splits the record into fields, sets the value of @code{NF} and recomputes the value of @code{$0}. The values of @code{NR} and @code{FNR} are not changed. @c Thanks to Paul Eggert for initial wording here According to POSIX, @samp{@var{expression} | getline} is ambiguous if @var{expression} contains unparenthesized operators other than @samp{$}; for example, @samp{"echo " "date" | getline} is ambiguous because the concatenation operator is not parenthesized, and you should write it as @samp{("echo " "date") | getline} if you want your program to be portable to other @code{awk} implementations. (It happens that @code{gawk} gets it right, but you should not rely on this. Parentheses make it easier to read, anyway.) @node Getline/Variable/Pipe, Getline Summary, Getline/Pipe, Getline @subsection Using @code{getline} Into a Variable from a Pipe When you use @samp{@var{command} | getline @var{var}}, the output of the command @var{command} is sent through a pipe to @code{getline} and into the variable @var{var}. For example, the following program reads the current date and time into the variable @code{current_time}, using the @code{date} utility, and then prints it. @example @group awk 'BEGIN @{ "date" | getline current_time close("date") print "Report printed on " current_time @}' @end group @end example In this version of @code{getline}, none of the built-in variables are changed, and the record is not split into fields. @ifinfo @c Thanks to Paul Eggert for initial wording here According to POSIX, @samp{@var{expression} | getline @var{var}} is ambiguous if @var{expression} contains unparenthesized operators other than @samp{$}; for example, @samp{"echo " "date" | getline @var{var}} is ambiguous because the concatenation operator is not parenthesized, and you should write it as @samp{("echo " "date") | getline @var{var}} if you want your program to be portable to other @code{awk} implementations. (It happens that @code{gawk} gets it right, but you should not rely on this. Parentheses make it easier to read, anyway.) @end ifinfo @node Getline Summary, , Getline/Variable/Pipe, Getline @subsection Summary of @code{getline} Variants With all the forms of @code{getline}, even though @code{$0} and @code{NF}, may be updated, the record will not be tested against all the patterns in the @code{awk} program, in the way that would happen if the record were read normally by the main processing loop of @code{awk}. However the new record is tested against any subsequent rules. @cindex differences between @code{gawk} and @code{awk} @cindex limitations @cindex implementation limits Many @code{awk} implementations limit the number of pipelines an @code{awk} program may have open to just one! In @code{gawk}, there is no such limit. You can open as many pipelines as the underlying operating system will permit. @vindex FILENAME @cindex dark corner @cindex @code{getline}, setting @code{FILENAME} @cindex @code{FILENAME}, being set by @code{getline} An interesting side-effect occurs if you use @code{getline} (without a redirection) inside a @code{BEGIN} rule. Since an unredirected @code{getline} reads from the command line data files, the first @code{getline} command causes @code{awk} to set the value of @code{FILENAME}. Normally, @code{FILENAME} does not have a value inside @code{BEGIN} rules, since you have not yet started to process the command line data files (d.c.). (@xref{BEGIN/END, , The @code{BEGIN} and @code{END} Special Patterns}, also @pxref{Auto-set, , Built-in Variables that Convey Information}.) The following table summarizes the six variants of @code{getline}, listing which built-in variables are set by each one. @c @cartouche @table @code @item getline sets @code{$0}, @code{NF}, @code{FNR}, and @code{NR}. @item getline @var{var} sets @var{var}, @code{FNR}, and @code{NR}. @item getline < @var{file} sets @code{$0}, and @code{NF}. @item getline @var{var} < @var{file} sets @var{var}. @item @var{command} | getline sets @code{$0}, and @code{NF}. @item @var{command} | getline @var{var} sets @var{var}. @end table @c @end cartouche @node Printing, Expressions, Reading Files, Top @chapter Printing Output @cindex printing @cindex output One of the most common actions is to @dfn{print}, or output, some or all of the input. You use the @code{print} statement for simple output. You use the @code{printf} statement for fancier formatting. Both are described in this chapter. @menu * Print:: The @code{print} statement. * Print Examples:: Simple examples of @code{print} statements. * Output Separators:: The output separators and how to change them. * OFMT:: Controlling Numeric Output With @code{print}. * Printf:: The @code{printf} statement. * Redirection:: How to redirect output to multiple files and pipes. * Special Files:: File name interpretation in @code{gawk}. @code{gawk} allows access to inherited file descriptors. * Close Files And Pipes:: Closing Input and Output Files and Pipes. @end menu @node Print, Print Examples, Printing, Printing @section The @code{print} Statement @cindex @code{print} statement The @code{print} statement does output with simple, standardized formatting. You specify only the strings or numbers to be printed, in a list separated by commas. They are output, separated by single spaces, followed by a newline. The statement looks like this: @example print @var{item1}, @var{item2}, @dots{} @end example @noindent The entire list of items may optionally be enclosed in parentheses. The parentheses are necessary if any of the item expressions uses the @samp{>} relational operator; otherwise it could be confused with a redirection (@pxref{Redirection, ,Redirecting Output of @code{print} and @code{printf}}). The items to be printed can be constant strings or numbers, fields of the current record (such as @code{$1}), variables, or any @code{awk} expressions. Numeric values are converted to strings, and then printed. The @code{print} statement is completely general for computing @emph{what} values to print. However, with two exceptions, you cannot specify @emph{how} to print them---how many columns, whether to use exponential notation or not, and so on. (For the exceptions, @pxref{Output Separators}, and @ref{OFMT, ,Controlling Numeric Output with @code{print}}.) For that, you need the @code{printf} statement (@pxref{Printf, ,Using @code{printf} Statements for Fancier Printing}). The simple statement @samp{print} with no items is equivalent to @samp{print $0}: it prints the entire current record. To print a blank line, use @samp{print ""}, where @code{""} is the empty string. To print a fixed piece of text, use a string constant such as @w{@code{"Don't Panic"}} as one item. If you forget to use the double-quote characters, your text will be taken as an @code{awk} expression, and you will probably get an error. Keep in mind that a space is printed between any two items. Each @code{print} statement makes at least one line of output. But it isn't limited to one line. If an item value is a string that contains a newline, the newline is output along with the rest of the string. A single @code{print} can make any number of lines this way. @node Print Examples, Output Separators, Print, Printing @section Examples of @code{print} Statements Here is an example of printing a string that contains embedded newlines (the @samp{\n} is an escape sequence, used to represent the newline character; @pxref{Escape Sequences}): @example @group $ awk 'BEGIN @{ print "line one\nline two\nline three" @}' @print{} line one @print{} line two @print{} line three @end group @end example Here is an example that prints the first two fields of each input record, with a space between them: @example @group $ awk '@{ print $1, $2 @}' inventory-shipped @print{} Jan 13 @print{} Feb 15 @print{} Mar 15 @dots{} @end group @end example @cindex common mistakes @cindex mistakes, common @cindex errors, common A common mistake in using the @code{print} statement is to omit the comma between two items. This often has the effect of making the items run together in the output, with no space. The reason for this is that juxtaposing two string expressions in @code{awk} means to concatenate them. Here is the same program, without the comma: @example @group $ awk '@{ print $1 $2 @}' inventory-shipped @print{} Jan13 @print{} Feb15 @print{} Mar15 @dots{} @end group @end example To someone unfamiliar with the file @file{inventory-shipped}, neither example's output makes much sense. A heading line at the beginning would make it clearer. Let's add some headings to our table of months (@code{$1}) and green crates shipped (@code{$2}). We do this using the @code{BEGIN} pattern (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}) to force the headings to be printed only once: @example awk 'BEGIN @{ print "Month Crates" print "----- ------" @} @{ print $1, $2 @}' inventory-shipped @end example @noindent Did you already guess what happens? When run, the program prints the following: @example @group Month Crates ----- ------ Jan 13 Feb 15 Mar 15 @dots{} @end group @end example @noindent The headings and the table data don't line up! We can fix this by printing some spaces between the two fields: @example awk 'BEGIN @{ print "Month Crates" print "----- ------" @} @{ print $1, " ", $2 @}' inventory-shipped @end example You can imagine that this way of lining up columns can get pretty complicated when you have many columns to fix. Counting spaces for two or three columns can be simple, but more than this and you can get lost quite easily. This is why the @code{printf} statement was created (@pxref{Printf, ,Using @code{printf} Statements for Fancier Printing}); one of its specialties is lining up columns of data. @cindex line continuation As a side point, you can continue either a @code{print} or @code{printf} statement simply by putting a newline after any comma (@pxref{Statements/Lines, ,@code{awk} Statements Versus Lines}). @node Output Separators, OFMT, Print Examples, Printing @section Output Separators @cindex output field separator, @code{OFS} @cindex output record separator, @code{ORS} @vindex OFS @vindex ORS As mentioned previously, a @code{print} statement contains a list of items, separated by commas. In the output, the items are normally separated by single spaces. This need not be the case; a single space is only the default. You can specify any string of characters to use as the @dfn{output field separator} by setting the built-in variable @code{OFS}. The initial value of this variable is the string @w{@code{" "}}, that is, a single space. The output from an entire @code{print} statement is called an @dfn{output record}. Each @code{print} statement outputs one output record and then outputs a string called the @dfn{output record separator}. The built-in variable @code{ORS} specifies this string. The initial value of @code{ORS} is the string @code{"\n"}, i.e.@: a newline character; thus, normally each @code{print} statement makes a separate line. You can change how output fields and records are separated by assigning new values to the variables @code{OFS} and/or @code{ORS}. The usual place to do this is in the @code{BEGIN} rule (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}), so that it happens before any input is processed. You may also do this with assignments on the command line, before the names of your input files, or using the @samp{-v} command line option (@pxref{Options, ,Command Line Options}). @ignore Exercise, Rewrite the @example awk 'BEGIN @{ print "Month Crates" print "----- ------" @} @{ print $1, " ", $2 @}' inventory-shipped @end example program by using a new value of @code{OFS}. @end ignore The following example prints the first and second fields of each input record separated by a semicolon, with a blank line added after each line: @example @group $ awk 'BEGIN @{ OFS = ";"; ORS = "\n\n" @} > @{ print $1, $2 @}' BBS-list @print{} aardvark;555-5553 @print{} @print{} alpo-net;555-3412 @print{} @print{} barfly;555-7685 @dots{} @end group @end example If the value of @code{ORS} does not contain a newline, all your output will be run together on a single line, unless you output newlines some other way. @node OFMT, Printf, Output Separators, Printing @section Controlling Numeric Output with @code{print} @vindex OFMT @cindex numeric output format @cindex format, numeric output @cindex output format specifier, @code{OFMT} When you use the @code{print} statement to print numeric values, @code{awk} internally converts the number to a string of characters, and prints that string. @code{awk} uses the @code{sprintf} function to do this conversion (@pxref{String Functions, ,Built-in Functions for String Manipulation}). For now, it suffices to say that the @code{sprintf} function accepts a @dfn{format specification} that tells it how to format numbers (or strings), and that there are a number of different ways in which numbers can be formatted. The different format specifications are discussed more fully in @ref{Control Letters, , Format-Control Letters}. The built-in variable @code{OFMT} contains the default format specification that @code{print} uses with @code{sprintf} when it wants to convert a number to a string for printing. The default value of @code{OFMT} is @code{"%.6g"}. By supplying different format specifications as the value of @code{OFMT}, you can change how @code{print} will print your numbers. As a brief example: @example @group $ awk 'BEGIN @{ > OFMT = "%.0f" # print numbers as integers (rounds) > print 17.23 @}' @print{} 17 @end group @end example @noindent @cindex dark corner @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} According to the POSIX standard, @code{awk}'s behavior will be undefined if @code{OFMT} contains anything but a floating point conversion specification (d.c.). @node Printf, Redirection, OFMT, Printing @section Using @code{printf} Statements for Fancier Printing @cindex formatted output @cindex output, formatted If you want more precise control over the output format than @code{print} gives you, use @code{printf}. With @code{printf} you can specify the width to use for each item, and you can specify various formatting choices for numbers (such as what radix to use, whether to print an exponent, whether to print a sign, and how many digits to print after the decimal point). You do this by supplying a string, called the @dfn{format string}, which controls how and where to print the other arguments. @menu * Basic Printf:: Syntax of the @code{printf} statement. * Control Letters:: Format-control letters. * Format Modifiers:: Format-specification modifiers. * Printf Examples:: Several examples. @end menu @node Basic Printf, Control Letters, Printf, Printf @subsection Introduction to the @code{printf} Statement @cindex @code{printf} statement, syntax of The @code{printf} statement looks like this: @example printf @var{format}, @var{item1}, @var{item2}, @dots{} @end example @noindent The entire list of arguments may optionally be enclosed in parentheses. The parentheses are necessary if any of the item expressions use the @samp{>} relational operator; otherwise it could be confused with a redirection (@pxref{Redirection, ,Redirecting Output of @code{print} and @code{printf}}). @cindex format string The difference between @code{printf} and @code{print} is the @var{format} argument. This is an expression whose value is taken as a string; it specifies how to output each of the other arguments. It is called the @dfn{format string}. The format string is very similar to that in the ANSI C library function @code{printf}. Most of @var{format} is text to be output verbatim. Scattered among this text are @dfn{format specifiers}, one per item. Each format specifier says to output the next item in the argument list at that place in the format. The @code{printf} statement does not automatically append a newline to its output. It outputs only what the format string specifies. So if you want a newline, you must include one in the format string. The output separator variables @code{OFS} and @code{ORS} have no effect on @code{printf} statements. For example: @example @group BEGIN @{ ORS = "\nOUCH!\n"; OFS = "!" msg = "Don't Panic!"; printf "%s\n", msg @} @end group @end example This program still prints the familiar @samp{Don't Panic!} message. @node Control Letters, Format Modifiers, Basic Printf, Printf @subsection Format-Control Letters @cindex @code{printf}, format-control characters @cindex format specifier A format specifier starts with the character @samp{%} and ends with a @dfn{format-control letter}; it tells the @code{printf} statement how to output one item. (If you actually want to output a @samp{%}, write @samp{%%}.) The format-control letter specifies what kind of value to print. The rest of the format specifier is made up of optional @dfn{modifiers} which are parameters to use, such as the field width. Here is a list of the format-control letters: @table @code @item c This prints a number as an ASCII character. Thus, @samp{printf "%c", 65} outputs the letter @samp{A}. The output for a string value is the first character of the string. @item d @itemx i These are equivalent. They both print a decimal integer. The @samp{%i} specification is for compatibility with ANSI C. @item e @itemx E This prints a number in scientific (exponential) notation. For example, @example printf "%4.3e\n", 1950 @end example @noindent prints @samp{1.950e+03}, with a total of four significant figures of which three follow the decimal point. The @samp{4.3} are modifiers, discussed below. @samp{%E} uses @samp{E} instead of @samp{e} in the output. @item f This prints a number in floating point notation. For example, @example printf "%4.3f", 1950 @end example @noindent prints @samp{1950.000}, with a total of four significant figures of which three follow the decimal point. The @samp{4.3} are modifiers, discussed below. @item g @itemx G This prints a number in either scientific notation or floating point notation, whichever uses fewer characters. If the result is printed in scientific notation, @samp{%G} uses @samp{E} instead of @samp{e}. @item o This prints an unsigned octal integer. (In octal, or base-eight notation, the digits run from @samp{0} to @samp{7}; the decimal number eight is represented as @samp{10} in octal.) @item s This prints a string. @item x @itemx X This prints an unsigned hexadecimal integer. (In hexadecimal, or base-16 notation, the digits are @samp{0} through @samp{9} and @samp{a} through @samp{f}. The hexadecimal digit @samp{f} represents the decimal number 15.) @samp{%X} uses the letters @samp{A} through @samp{F} instead of @samp{a} through @samp{f}. @item % This isn't really a format-control letter, but it does have a meaning when used after a @samp{%}: the sequence @samp{%%} outputs one @samp{%}. It does not consume an argument, and it ignores any modifiers. @end table @cindex dark corner When using the integer format-control letters for values that are outside the range of a C @code{long} integer, @code{gawk} will switch to the @samp{%g} format specifier. Other versions of @code{awk} may print invalid values, or do something else entirely (d.c.). @node Format Modifiers, Printf Examples, Control Letters, Printf @subsection Modifiers for @code{printf} Formats @cindex @code{printf}, modifiers @cindex modifiers (in format specifiers) A format specification can also include @dfn{modifiers} that can control how much of the item's value is printed and how much space it gets. The modifiers come between the @samp{%} and the format-control letter. In the examples below, we use the bullet symbol ``@bullet{}'' to represent spaces in the output. Here are the possible modifiers, in the order in which they may appear: @table @code @item - The minus sign, used before the width modifier (see below), says to left-justify the argument within its specified width. Normally the argument is printed right-justified in the specified width. Thus, @example printf "%-4s", "foo" @end example @noindent prints @samp{foo@bullet{}}. @item @var{space} For numeric conversions, prefix positive values with a space, and negative values with a minus sign. @item + The plus sign, used before the width modifier (see below), says to always supply a sign for numeric conversions, even if the data to be formatted is positive. The @samp{+} overrides the space modifier. @item # Use an ``alternate form'' for certain control letters. For @samp{%o}, supply a leading zero. For @samp{%x}, and @samp{%X}, supply a leading @samp{0x} or @samp{0X} for a non-zero result. For @samp{%e}, @samp{%E}, and @samp{%f}, the result will always contain a decimal point. For @samp{%g}, and @samp{%G}, trailing zeros are not removed from the result. @cindex dark corner @item 0 A leading @samp{0} (zero) acts as a flag, that indicates output should be padded with zeros instead of spaces. This applies even to non-numeric output formats (d.c.). This flag only has an effect when the field width is wider than the value to be printed. @item @var{width} This is a number specifying the desired minimum width of a field. Inserting any number between the @samp{%} sign and the format control character forces the field to be expanded to this width. The default way to do this is to pad with spaces on the left. For example, @example printf "%4s", "foo" @end example @noindent prints @samp{@bullet{}foo}. The value of @var{width} is a minimum width, not a maximum. If the item value requires more than @var{width} characters, it can be as wide as necessary. Thus, @example printf "%4s", "foobar" @end example @noindent prints @samp{foobar}. Preceding the @var{width} with a minus sign causes the output to be padded with spaces on the right, instead of on the left. @item .@var{prec} This is a number that specifies the precision to use when printing. For the @samp{e}, @samp{E}, and @samp{f} formats, this specifies the number of digits you want printed to the right of the decimal point. For the @samp{g}, and @samp{G} formats, it specifies the maximum number of significant digits. For the @samp{d}, @samp{o}, @samp{i}, @samp{u}, @samp{x}, and @samp{X} formats, it specifies the minimum number of digits to print. For a string, it specifies the maximum number of characters from the string that should be printed. Thus, @example printf "%.4s", "foobar" @end example @noindent prints @samp{foob}. @end table The C library @code{printf}'s dynamic @var{width} and @var{prec} capability (for example, @code{"%*.*s"}) is supported. Instead of supplying explicit @var{width} and/or @var{prec} values in the format string, you pass them in the argument list. For example: @example w = 5 p = 3 s = "abcdefg" printf "%*.*s\n", w, p, s @end example @noindent is exactly equivalent to @example s = "abcdefg" printf "%5.3s\n", s @end example @noindent Both programs output @samp{@w{@bullet{}@bullet{}abc}}. Earlier versions of @code{awk} did not support this capability. If you must use such a version, you may simulate this feature by using concatenation to build up the format string, like so: @example w = 5 p = 3 s = "abcdefg" printf "%" w "." p "s\n", s @end example @noindent This is not particularly easy to read, but it does work. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} C programmers may be used to supplying additional @samp{l} and @samp{h} flags in @code{printf} format strings. These are not valid in @code{awk}. Most @code{awk} implementations silently ignore these flags. If @samp{--lint} is provided on the command line (@pxref{Options, ,Command Line Options}), @code{gawk} will warn about their use. If @samp{--posix} is supplied, their use is a fatal error. @node Printf Examples, , Format Modifiers, Printf @subsection Examples Using @code{printf} Here is how to use @code{printf} to make an aligned table: @example awk '@{ printf "%-10s %s\n", $1, $2 @}' BBS-list @end example @noindent prints the names of bulletin boards (@code{$1}) of the file @file{BBS-list} as a string of 10 characters, left justified. It also prints the phone numbers (@code{$2}) afterward on the line. This produces an aligned two-column table of names and phone numbers: @example @group $ awk '@{ printf "%-10s %s\n", $1, $2 @}' BBS-list @print{} aardvark 555-5553 @print{} alpo-net 555-3412 @print{} barfly 555-7685 @print{} bites 555-1675 @print{} camelot 555-0542 @print{} core 555-2912 @print{} fooey 555-1234 @print{} foot 555-6699 @print{} macfoo 555-6480 @print{} sdace 555-3430 @print{} sabafoo 555-2127 @end group @end example Did you notice that we did not specify that the phone numbers be printed as numbers? They had to be printed as strings because the numbers are separated by a dash. If we had tried to print the phone numbers as numbers, all we would have gotten would have been the first three digits, @samp{555}. This would have been pretty confusing. We did not specify a width for the phone numbers because they are the last things on their lines. We don't need to put spaces after them. We could make our table look even nicer by adding headings to the tops of the columns. To do this, we use the @code{BEGIN} pattern (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}) to force the header to be printed only once, at the beginning of the @code{awk} program: @example @group awk 'BEGIN @{ print "Name Number" print "---- ------" @} @{ printf "%-10s %s\n", $1, $2 @}' BBS-list @end group @end example Did you notice that we mixed @code{print} and @code{printf} statements in the above example? We could have used just @code{printf} statements to get the same results: @example @group awk 'BEGIN @{ printf "%-10s %s\n", "Name", "Number" printf "%-10s %s\n", "----", "------" @} @{ printf "%-10s %s\n", $1, $2 @}' BBS-list @end group @end example @noindent By printing each column heading with the same format specification used for the elements of the column, we have made sure that the headings are aligned just like the columns. The fact that the same format specification is used three times can be emphasized by storing it in a variable, like this: @example @group awk 'BEGIN @{ format = "%-10s %s\n" printf format, "Name", "Number" printf format, "----", "------" @} @{ printf format, $1, $2 @}' BBS-list @end group @end example @c !!! exercise See if you can use the @code{printf} statement to line up the headings and table data for our @file{inventory-shipped} example covered earlier in the section on the @code{print} statement (@pxref{Print, ,The @code{print} Statement}). @node Redirection, Special Files, Printf, Printing @section Redirecting Output of @code{print} and @code{printf} @cindex output redirection @cindex redirection of output So far we have been dealing only with output that prints to the standard output, usually your terminal. Both @code{print} and @code{printf} can also send their output to other places. This is called @dfn{redirection}. A redirection appears after the @code{print} or @code{printf} statement. Redirections in @code{awk} are written just like redirections in shell commands, except that they are written inside the @code{awk} program. There are three forms of output redirection: output to a file, output appended to a file, and output through a pipe to another command. They are all shown for the @code{print} statement, but they work identically for @code{printf} also. @table @code @item print @var{items} > @var{output-file} This type of redirection prints the items into the output file @var{output-file}. The file name @var{output-file} can be any expression. Its value is changed to a string and then used as a file name (@pxref{Expressions}). When this type of redirection is used, the @var{output-file} is erased before the first output is written to it. Subsequent writes to the same @var{output-file} do not erase @var{output-file}, but append to it. If @var{output-file} does not exist, then it is created. For example, here is how an @code{awk} program can write a list of BBS names to a file @file{name-list} and a list of phone numbers to a file @file{phone-list}. Each output file contains one name or number per line. @example @group $ awk '@{ print $2 > "phone-list" > print $1 > "name-list" @}' BBS-list @end group @group $ cat phone-list @print{} 555-5553 @print{} 555-3412 @dots{} @end group @group $ cat name-list @print{} aardvark @print{} alpo-net @dots{} @end group @end example @item print @var{items} >> @var{output-file} This type of redirection prints the items into the pre-existing output file @var{output-file}. The difference between this and the single-@samp{>} redirection is that the old contents (if any) of @var{output-file} are not erased. Instead, the @code{awk} output is appended to the file. If @var{output-file} does not exist, then it is created. @cindex pipes for output @cindex output, piping @item print @var{items} | @var{command} It is also possible to send output to another program through a pipe instead of into a file. This type of redirection opens a pipe to @var{command} and writes the values of @var{items} through this pipe, to another process created to execute @var{command}. The redirection argument @var{command} is actually an @code{awk} expression. Its value is converted to a string, whose contents give the shell command to be run. For example, this produces two files, one unsorted list of BBS names and one list sorted in reverse alphabetical order: @example awk '@{ print $1 > "names.unsorted" command = "sort -r > names.sorted" print $1 | command @}' BBS-list @end example Here the unsorted list is written with an ordinary redirection while the sorted list is written by piping through the @code{sort} utility. This example uses redirection to mail a message to a mailing list @samp{bug-system}. This might be useful when trouble is encountered in an @code{awk} script run periodically for system maintenance. @example report = "mail bug-system" print "Awk script failed:", $0 | report m = ("at record number " FNR " of " FILENAME) print m | report close(report) @end example The message is built using string concatenation and saved in the variable @code{m}. It is then sent down the pipeline to the @code{mail} program. We call the @code{close} function here because it's a good idea to close the pipe as soon as all the intended output has been sent to it. @xref{Close Files And Pipes, ,Closing Input and Output Files and Pipes}, for more information on this. This example also illustrates the use of a variable to represent a @var{file} or @var{command}: it is not necessary to always use a string constant. Using a variable is generally a good idea, since @code{awk} requires you to spell the string value identically every time. @end table Redirecting output using @samp{>}, @samp{>>}, or @samp{|} asks the system to open a file or pipe only if the particular @var{file} or @var{command} you've specified has not already been written to by your program, or if it has been closed since it was last written to. @cindex differences between @code{gawk} and @code{awk} @cindex limitations @cindex implementation limits @iftex As mentioned earlier (@pxref{Getline Summary, , Summary of @code{getline} Variants}), many @end iftex @ifinfo Many @end ifinfo @code{awk} implementations limit the number of pipelines an @code{awk} program may have open to just one! In @code{gawk}, there is no such limit. You can open as many pipelines as the underlying operating system will permit. @node Special Files, Close Files And Pipes , Redirection, Printing @section Special File Names in @code{gawk} @cindex standard input @cindex standard output @cindex standard error output @cindex file descriptors Running programs conventionally have three input and output streams already available to them for reading and writing. These are known as the @dfn{standard input}, @dfn{standard output}, and @dfn{standard error output}. These streams are, by default, connected to your terminal, but they are often redirected with the shell, via the @samp{<}, @samp{<<}, @samp{>}, @samp{>>}, @samp{>&} and @samp{|} operators. Standard error is typically used for writing error messages; the reason we have two separate streams, standard output and standard error, is so that they can be redirected separately. @cindex differences between @code{gawk} and @code{awk} In other implementations of @code{awk}, the only way to write an error message to standard error in an @code{awk} program is as follows: @example print "Serious error detected!" | "cat 1>&2" @end example @noindent This works by opening a pipeline to a shell command which can access the standard error stream which it inherits from the @code{awk} process. This is far from elegant, and is also inefficient, since it requires a separate process. So people writing @code{awk} programs often neglect to do this. Instead, they send the error messages to the terminal, like this: @example @group print "Serious error detected!" > "/dev/tty" @end group @end example @noindent This usually has the same effect, but not always: although the standard error stream is usually the terminal, it can be redirected, and when that happens, writing to the terminal is not correct. In fact, if @code{awk} is run from a background job, it may not have a terminal at all. Then opening @file{/dev/tty} will fail. @code{gawk} provides special file names for accessing the three standard streams. When you redirect input or output in @code{gawk}, if the file name matches one of these special names, then @code{gawk} directly uses the stream it stands for. @cindex @file{/dev/stdin} @cindex @file{/dev/stdout} @cindex @file{/dev/stderr} @cindex @file{/dev/fd} @c @cartouche @table @file @item /dev/stdin The standard input (file descriptor 0). @item /dev/stdout The standard output (file descriptor 1). @item /dev/stderr The standard error output (file descriptor 2). @item /dev/fd/@var{N} The file associated with file descriptor @var{N}. Such a file must have been opened by the program initiating the @code{awk} execution (typically the shell). Unless you take special pains in the shell from which you invoke @code{gawk}, only descriptors 0, 1 and 2 are available. @end table @c @end cartouche The file names @file{/dev/stdin}, @file{/dev/stdout}, and @file{/dev/stderr} are aliases for @file{/dev/fd/0}, @file{/dev/fd/1}, and @file{/dev/fd/2}, respectively, but they are more self-explanatory. The proper way to write an error message in a @code{gawk} program is to use @file{/dev/stderr}, like this: @example print "Serious error detected!" > "/dev/stderr" @end example @code{gawk} also provides special file names that give access to information about the running @code{gawk} process. Each of these ``files'' provides a single record of information. To read them more than once, you must first close them with the @code{close} function (@pxref{Close Files And Pipes, ,Closing Input and Output Files and Pipes}). The filenames are: @cindex process information @cindex @file{/dev/pid} @cindex @file{/dev/pgrpid} @cindex @file{/dev/ppid} @cindex @file{/dev/user} @c @cartouche @table @file @item /dev/pid Reading this file returns the process ID of the current process, in decimal, terminated with a newline. @item /dev/ppid Reading this file returns the parent process ID of the current process, in decimal, terminated with a newline. @item /dev/pgrpid Reading this file returns the process group ID of the current process, in decimal, terminated with a newline. @item /dev/user Reading this file returns a single record terminated with a newline. The fields are separated with spaces. The fields represent the following information: @table @code @item $1 The return value of the @code{getuid} system call (the real user ID number). @item $2 The return value of the @code{geteuid} system call (the effective user ID number). @item $3 The return value of the @code{getgid} system call (the real group ID number). @item $4 The return value of the @code{getegid} system call (the effective group ID number). @end table If there are any additional fields, they are the group IDs returned by @code{getgroups} system call. (Multiple groups may not be supported on all systems.) @end table @c @end cartouche These special file names may be used on the command line as data files, as well as for I/O redirections within an @code{awk} program. They may not be used as source files with the @samp{-f} option. Recognition of these special file names is disabled if @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}). @strong{Caution}: Unless your system actually has a @file{/dev/fd} directory (or any of the other above listed special files), the interpretation of these file names is done by @code{gawk} itself. For example, using @samp{/dev/fd/4} for output will actually write on file descriptor 4, and not on a new file descriptor that was @code{dup}'ed from file descriptor 4. Most of the time this does not matter; however, it is important to @emph{not} close any of the files related to file descriptors 0, 1, and 2. If you do close one of these files, unpredictable behavior will result. The special files that provide process-related information may disappear in a future version of @code{gawk}. @xref{Future Extensions, ,Probable Future Extensions}. @node Close Files And Pipes, , Special Files, Printing @section Closing Input and Output Files and Pipes @cindex closing input files and pipes @cindex closing output files and pipes @findex close If the same file name or the same shell command is used with @code{getline} (@pxref{Getline, ,Explicit Input with @code{getline}}) more than once during the execution of an @code{awk} program, the file is opened (or the command is executed) only the first time. At that time, the first record of input is read from that file or command. The next time the same file or command is used in @code{getline}, another record is read from it, and so on. Similarly, when a file or pipe is opened for output, the file name or command associated with it is remembered by @code{awk} and subsequent writes to the same file or command are appended to the previous writes. The file or pipe stays open until @code{awk} exits. This implies that if you want to start reading the same file again from the beginning, or if you want to rerun a shell command (rather than reading more output from the command), you must take special steps. What you must do is use the @code{close} function, as follows: @example close(@var{filename}) @end example @noindent or @example close(@var{command}) @end example The argument @var{filename} or @var{command} can be any expression. Its value must @emph{exactly} match the string that was used to open the file or start the command (spaces and other ``irrelevant'' characters included). For example, if you open a pipe with this: @example "sort -r names" | getline foo @end example @noindent then you must close it with this: @example close("sort -r names") @end example Once this function call is executed, the next @code{getline} from that file or command, or the next @code{print} or @code{printf} to that file or command, will reopen the file or rerun the command. Because the expression that you use to close a file or pipeline must exactly match the expression used to open the file or run the command, it is good practice to use a variable to store the file name or command. The previous example would become @example sortcom = "sort -r names" sortcom | getline foo @dots{} close(sortcom) @end example @noindent This helps avoid hard-to-find typographical errors in your @code{awk} programs. Here are some reasons why you might need to close an output file: @itemize @bullet @item To write a file and read it back later on in the same @code{awk} program. Close the file when you are finished writing it; then you can start reading it with @code{getline}. @item To write numerous files, successively, in the same @code{awk} program. If you don't close the files, eventually you may exceed a system limit on the number of open files in one process. So close each one when you are finished writing it. @item To make a command finish. When you redirect output through a pipe, the command reading the pipe normally continues to try to read input as long as the pipe is open. Often this means the command cannot really do its work until the pipe is closed. For example, if you redirect output to the @code{mail} program, the message is not actually sent until the pipe is closed. @item To run the same program a second time, with the same arguments. This is not the same thing as giving more input to the first run! For example, suppose you pipe output to the @code{mail} program. If you output several lines redirected to this pipe without closing it, they make a single message of several lines. By contrast, if you close the pipe after each line of output, then each line makes a separate message. @end itemize @vindex ERRNO @cindex differences between @code{gawk} and @code{awk} @code{close} returns a value of zero if the close succeeded. Otherwise, the value will be non-zero. In this case, @code{gawk} sets the variable @code{ERRNO} to a string describing the error that occurred. @cindex differences between @code{gawk} and @code{awk} @cindex portability issues If you use more files than the system allows you to have open, @code{gawk} will attempt to multiplex the available open files among your data files. @code{gawk}'s ability to do this depends upon the facilities of your operating system: it may not always work. It is therefore both good practice and good portability advice to always use @code{close} on your files when you are done with them. @node Expressions, Patterns and Actions, Printing, Top @chapter Expressions @cindex expression Expressions are the basic building blocks of @code{awk} patterns and actions. An expression evaluates to a value, which you can print, test, store in a variable or pass to a function. Additionally, an expression can assign a new value to a variable or a field, with an assignment operator. An expression can serve as a pattern or action statement on its own. Most other kinds of statements contain one or more expressions which specify data on which to operate. As in other languages, expressions in @code{awk} include variables, array references, constants, and function calls, as well as combinations of these with various operators. @menu * Constants:: String, numeric, and regexp constants. * Using Constant Regexps:: When and how to use a regexp constant. * Variables:: Variables give names to values for later use. * Conversion:: The conversion of strings to numbers and vice versa. * Arithmetic Ops:: Arithmetic operations (@samp{+}, @samp{-}, etc.) * Concatenation:: Concatenating strings. * Assignment Ops:: Changing the value of a variable or a field. * Increment Ops:: Incrementing the numeric value of a variable. * Truth Values:: What is ``true'' and what is ``false''. * Typing and Comparison:: How variables acquire types, and how this affects comparison of numbers and strings with @samp{<}, etc. * Boolean Ops:: Combining comparison expressions using boolean operators @samp{||} (``or''), @samp{&&} (``and'') and @samp{!} (``not''). * Conditional Exp:: Conditional expressions select between two subexpressions under control of a third subexpression. * Function Calls:: A function call is an expression. * Precedence:: How various operators nest. @end menu @node Constants, Using Constant Regexps, Expressions, Expressions @section Constant Expressions @cindex constants, types of @cindex string constants The simplest type of expression is the @dfn{constant}, which always has the same value. There are three types of constants: numeric constants, string constants, and regular expression constants. @menu * Scalar Constants:: Numeric and string constants. * Regexp Constants:: Regular Expression constants. @end menu @node Scalar Constants, Regexp Constants, Constants, Constants @subsection Numeric and String Constants @cindex numeric constant @cindex numeric value A @dfn{numeric constant} stands for a number. This number can be an integer, a decimal fraction, or a number in scientific (exponential) notation.@footnote{The internal representation uses double-precision floating point numbers. If you don't know what that means, then don't worry about it.} Here are some examples of numeric constants, which all have the same value: @example 105 1.05e+2 1050e-1 @end example A string constant consists of a sequence of characters enclosed in double-quote marks. For example: @example "parrot" @end example @noindent @cindex differences between @code{gawk} and @code{awk} represents the string whose contents are @samp{parrot}. Strings in @code{gawk} can be of any length and they can contain any of the possible eight-bit ASCII characters including ASCII NUL (character code zero). Other @code{awk} implementations may have difficulty with some character codes. @node Regexp Constants, , Scalar Constants, Constants @subsection Regular Expression Constants @cindex @code{~} operator @cindex @code{!~} operator A regexp constant is a regular expression description enclosed in slashes, such as @code{@w{/^beginning and end$/}}. Most regexps used in @code{awk} programs are constant, but the @samp{~} and @samp{!~} matching operators can also match computed or ``dynamic'' regexps (which are just ordinary strings or variables that contain a regexp). @node Using Constant Regexps, Variables, Constants, Expressions @section Using Regular Expression Constants When used on the right hand side of the @samp{~} or @samp{!~} operators, a regexp constant merely stands for the regexp that is to be matched. @cindex dark corner Regexp constants (such as @code{/foo/}) may be used like simple expressions. When a regexp constant appears by itself, it has the same meaning as if it appeared in a pattern, i.e.@: @samp{($0 ~ /foo/)} (d.c.) (@pxref{Expression Patterns, ,Expressions as Patterns}). This means that the two code segments, @example if ($0 ~ /barfly/ || $0 ~ /camelot/) print "found" @end example @noindent and @example if (/barfly/ || /camelot/) print "found" @end example @noindent are exactly equivalent. One rather bizarre consequence of this rule is that the following boolean expression is valid, but does not do what the user probably intended: @example # note that /foo/ is on the left of the ~ if (/foo/ ~ $1) print "found foo" @end example @noindent This code is ``obviously'' testing @code{$1} for a match against the regexp @code{/foo/}. But in fact, the expression @samp{/foo/ ~ $1} actually means @samp{($0 ~ /foo/) ~ $1}. In other words, first match the input record against the regexp @code{/foo/}. The result will be either zero or one, depending upon the success or failure of the match. Then match that result against the first field in the record. Since it is unlikely that you would ever really wish to make this kind of test, @code{gawk} will issue a warning when it sees this construct in a program. Another consequence of this rule is that the assignment statement @example matches = /foo/ @end example @noindent will assign either zero or one to the variable @code{matches}, depending upon the contents of the current input record. This feature of the language was never well documented until the POSIX specification. @cindex differences between @code{gawk} and @code{awk} @cindex dark corner Constant regular expressions are also used as the first argument for the @code{gensub}, @code{sub} and @code{gsub} functions, and as the second argument of the @code{match} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). Modern implementations of @code{awk}, including @code{gawk}, allow the third argument of @code{split} to be a regexp constant, while some older implementations do not (d.c.). This can lead to confusion when attempting to use regexp constants as arguments to user defined functions (@pxref{User-defined, , User-defined Functions}). For example: @example @group function mysub(pat, repl, str, global) @{ if (global) gsub(pat, repl, str) else sub(pat, repl, str) return str @} @end group @group @{ @dots{} text = "hi! hi yourself!" mysub(/hi/, "howdy", text, 1) @dots{} @} @end group @end example In this example, the programmer wishes to pass a regexp constant to the user-defined function @code{mysub}, which will in turn pass it on to either @code{sub} or @code{gsub}. However, what really happens is that the @code{pat} parameter will be either one or zero, depending upon whether or not @code{$0} matches @code{/hi/}. As it is unlikely that you would ever really wish to pass a truth value in this way, @code{gawk} will issue a warning when it sees a regexp constant used as a parameter to a user-defined function. @node Variables, Conversion, Using Constant Regexps, Expressions @section Variables Variables are ways of storing values at one point in your program for use later in another part of your program. You can manipulate them entirely within your program text, and you can also assign values to them on the @code{awk} command line. @menu * Using Variables:: Using variables in your programs. * Assignment Options:: Setting variables on the command line and a summary of command line syntax. This is an advanced method of input. @end menu @node Using Variables, Assignment Options, Variables, Variables @subsection Using Variables in a Program @cindex variables, user-defined @cindex user-defined variables Variables let you give names to values and refer to them later. You have already seen variables in many of the examples. The name of a variable must be a sequence of letters, digits and underscores, but it may not begin with a digit. Case is significant in variable names; @code{a} and @code{A} are distinct variables. A variable name is a valid expression by itself; it represents the variable's current value. Variables are given new values with @dfn{assignment operators}, @dfn{increment operators} and @dfn{decrement operators}. @xref{Assignment Ops, ,Assignment Expressions}. A few variables have special built-in meanings, such as @code{FS}, the field separator, and @code{NF}, the number of fields in the current input record. @xref{Built-in Variables}, for a list of them. These built-in variables can be used and assigned just like all other variables, but their values are also used or changed automatically by @code{awk}. All built-in variables names are entirely upper-case. Variables in @code{awk} can be assigned either numeric or string values. By default, variables are initialized to the empty string, which is zero if converted to a number. There is no need to ``initialize'' each variable explicitly in @code{awk}, the way you would in C and in most other traditional languages. @node Assignment Options, , Using Variables, Variables @subsection Assigning Variables on the Command Line You can set any @code{awk} variable by including a @dfn{variable assignment} among the arguments on the command line when you invoke @code{awk} (@pxref{Other Arguments, ,Other Command Line Arguments}). Such an assignment has this form: @example @var{variable}=@var{text} @end example @noindent With it, you can set a variable either at the beginning of the @code{awk} run or in between input files. If you precede the assignment with the @samp{-v} option, like this: @example -v @var{variable}=@var{text} @end example @noindent then the variable is set at the very beginning, before even the @code{BEGIN} rules are run. The @samp{-v} option and its assignment must precede all the file name arguments, as well as the program text. (@xref{Options, ,Command Line Options}, for more information about the @samp{-v} option.) Otherwise, the variable assignment is performed at a time determined by its position among the input file arguments: after the processing of the preceding input file argument. For example: @example awk '@{ print $n @}' n=4 inventory-shipped n=2 BBS-list @end example @noindent prints the value of field number @code{n} for all input records. Before the first file is read, the command line sets the variable @code{n} equal to four. This causes the fourth field to be printed in lines from the file @file{inventory-shipped}. After the first file has finished, but before the second file is started, @code{n} is set to two, so that the second field is printed in lines from @file{BBS-list}. @example @group $ awk '@{ print $n @}' n=4 inventory-shipped n=2 BBS-list @print{} 15 @print{} 24 @dots{} @print{} 555-5553 @print{} 555-3412 @dots{} @end group @end example Command line arguments are made available for explicit examination by the @code{awk} program in an array named @code{ARGV} (@pxref{ARGC and ARGV, ,Using @code{ARGC} and @code{ARGV}}). @cindex dark corner @code{awk} processes the values of command line assignments for escape sequences (d.c.) (@pxref{Escape Sequences}). @node Conversion, Arithmetic Ops, Variables, Expressions @section Conversion of Strings and Numbers @cindex conversion of strings and numbers Strings are converted to numbers, and numbers to strings, if the context of the @code{awk} program demands it. For example, if the value of either @code{foo} or @code{bar} in the expression @samp{foo + bar} happens to be a string, it is converted to a number before the addition is performed. If numeric values appear in string concatenation, they are converted to strings. Consider this: @example two = 2; three = 3 print (two three) + 4 @end example @noindent This prints the (numeric) value 27. The numeric values of the variables @code{two} and @code{three} are converted to strings and concatenated together, and the resulting string is converted back to the number 23, to which four is then added. @cindex null string @cindex empty string @cindex type conversion If, for some reason, you need to force a number to be converted to a string, concatenate the empty string, @code{""}, with that number. To force a string to be converted to a number, add zero to that string. A string is converted to a number by interpreting any numeric prefix of the string as numerals: @code{"2.5"} converts to 2.5, @code{"1e3"} converts to 1000, and @code{"25fix"} has a numeric value of 25. Strings that can't be interpreted as valid numbers are converted to zero. @vindex CONVFMT The exact manner in which numbers are converted into strings is controlled by the @code{awk} built-in variable @code{CONVFMT} (@pxref{Built-in Variables}). Numbers are converted using the @code{sprintf} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}) with @code{CONVFMT} as the format specifier. @code{CONVFMT}'s default value is @code{"%.6g"}, which prints a value with at least six significant digits. For some applications you will want to change it to specify more precision. Double precision on most modern machines gives you 16 or 17 decimal digits of precision. Strange results can happen if you set @code{CONVFMT} to a string that doesn't tell @code{sprintf} how to format floating point numbers in a useful way. For example, if you forget the @samp{%} in the format, all numbers will be converted to the same constant string. @cindex dark corner As a special case, if a number is an integer, then the result of converting it to a string is @emph{always} an integer, no matter what the value of @code{CONVFMT} may be. Given the following code fragment: @example CONVFMT = "%2.2f" a = 12 b = a "" @end example @noindent @code{b} has the value @code{"12"}, not @code{"12.00"} (d.c.). @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @vindex OFMT Prior to the POSIX standard, @code{awk} specified that the value of @code{OFMT} was used for converting numbers to strings. @code{OFMT} specifies the output format to use when printing numbers with @code{print}. @code{CONVFMT} was introduced in order to separate the semantics of conversion from the semantics of printing. Both @code{CONVFMT} and @code{OFMT} have the same default value: @code{"%.6g"}. In the vast majority of cases, old @code{awk} programs will not change their behavior. However, this use of @code{OFMT} is something to keep in mind if you must port your program to other implementations of @code{awk}; we recommend that instead of changing your programs, you just port @code{gawk} itself! @xref{Print, ,The @code{print} Statement}, for more information on the @code{print} statement. @node Arithmetic Ops, Concatenation, Conversion, Expressions @section Arithmetic Operators @cindex arithmetic operators @cindex operators, arithmetic @cindex addition @cindex subtraction @cindex multiplication @cindex division @cindex remainder @cindex quotient @cindex exponentiation The @code{awk} language uses the common arithmetic operators when evaluating expressions. All of these arithmetic operators follow normal precedence rules, and work as you would expect them to. Here is a file @file{grades} containing a list of student names and three test scores per student (it's a small class): @example Pat 100 97 58 Sandy 84 72 93 Chris 72 92 89 @end example @noindent This programs takes the file @file{grades}, and prints the average of the scores. @example $ awk '@{ sum = $2 + $3 + $4 ; avg = sum / 3 > print $1, avg @}' grades @print{} Pat 85 @print{} Sandy 83 @print{} Chris 84.3333 @end example This table lists the arithmetic operators in @code{awk}, in order from highest precedence to lowest: @c @cartouche @table @code @item - @var{x} Negation. @item + @var{x} Unary plus. The expression is converted to a number. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @item @var{x} ^ @var{y} @itemx @var{x} ** @var{y} Exponentiation: @var{x} raised to the @var{y} power. @samp{2 ^ 3} has the value eight. The character sequence @samp{**} is equivalent to @samp{^}. (The POSIX standard only specifies the use of @samp{^} for exponentiation.) @item @var{x} * @var{y} Multiplication. @item @var{x} / @var{y} Division. Since all numbers in @code{awk} are real numbers, the result is not rounded to an integer: @samp{3 / 4} has the value 0.75. @item @var{x} % @var{y} @cindex differences between @code{gawk} and @code{awk} Remainder. The quotient is rounded toward zero to an integer, multiplied by @var{y} and this result is subtracted from @var{x}. This operation is sometimes known as ``trunc-mod.'' The following relation always holds: @example b * int(a / b) + (a % b) == a @end example One possibly undesirable effect of this definition of remainder is that @code{@var{x} % @var{y}} is negative if @var{x} is negative. Thus, @example -17 % 8 = -1 @end example In other @code{awk} implementations, the signedness of the remainder may be machine dependent. @c !!! what does posix say? @item @var{x} + @var{y} Addition. @item @var{x} - @var{y} Subtraction. @end table @c @end cartouche For maximum portability, do not use the @samp{**} operator. Unary plus and minus have the same precedence, the multiplication operators all have the same precedence, and addition and subtraction have the same precedence. @node Concatenation, Assignment Ops, Arithmetic Ops, Expressions @section String Concatenation @cindex Kernighan, Brian @display @i{It seemed like a good idea at the time.} Brian Kernighan @end display @sp 1 @cindex string operators @cindex operators, string @cindex concatenation There is only one string operation: concatenation. It does not have a specific operator to represent it. Instead, concatenation is performed by writing expressions next to one another, with no operator. For example: @example @group $ awk '@{ print "Field number one: " $1 @}' BBS-list @print{} Field number one: aardvark @print{} Field number one: alpo-net @dots{} @end group @end example Without the space in the string constant after the @samp{:}, the line would run together. For example: @example @group $ awk '@{ print "Field number one:" $1 @}' BBS-list @print{} Field number one:aardvark @print{} Field number one:alpo-net @dots{} @end group @end example Since string concatenation does not have an explicit operator, it is often necessary to insure that it happens where you want it to by using parentheses to enclose the items to be concatenated. For example, the following code fragment does not concatenate @code{file} and @code{name} as you might expect: @example @group file = "file" name = "name" print "something meaningful" > file name @end group @end example @noindent It is necessary to use the following: @example print "something meaningful" > (file name) @end example We recommend that you use parentheses around concatenation in all but the most common contexts (such as on the right-hand side of @samp{=}). @node Assignment Ops, Increment Ops, Concatenation, Expressions @section Assignment Expressions @cindex assignment operators @cindex operators, assignment @cindex expression, assignment An @dfn{assignment} is an expression that stores a new value into a variable. For example, let's assign the value one to the variable @code{z}: @example z = 1 @end example After this expression is executed, the variable @code{z} has the value one. Whatever old value @code{z} had before the assignment is forgotten. Assignments can store string values also. For example, this would store the value @code{"this food is good"} in the variable @code{message}: @example thing = "food" predicate = "good" message = "this " thing " is " predicate @end example @noindent (This also illustrates string concatenation.) The @samp{=} sign is called an @dfn{assignment operator}. It is the simplest assignment operator because the value of the right-hand operand is stored unchanged. @cindex side effect Most operators (addition, concatenation, and so on) have no effect except to compute a value. If you ignore the value, you might as well not use the operator. An assignment operator is different; it does produce a value, but even if you ignore the value, the assignment still makes itself felt through the alteration of the variable. We call this a @dfn{side effect}. @cindex lvalue @cindex rvalue The left-hand operand of an assignment need not be a variable (@pxref{Variables}); it can also be a field (@pxref{Changing Fields, ,Changing the Contents of a Field}) or an array element (@pxref{Arrays, ,Arrays in @code{awk}}). These are all called @dfn{lvalues}, which means they can appear on the left-hand side of an assignment operator. The right-hand operand may be any expression; it produces the new value which the assignment stores in the specified variable, field or array element. (Such values are called @dfn{rvalues}). @cindex types of variables It is important to note that variables do @emph{not} have permanent types. The type of a variable is simply the type of whatever value it happens to hold at the moment. In the following program fragment, the variable @code{foo} has a numeric value at first, and a string value later on: @example @group foo = 1 print foo foo = "bar" print foo @end group @end example @noindent When the second assignment gives @code{foo} a string value, the fact that it previously had a numeric value is forgotten. String values that do not begin with a digit have a numeric value of zero. After executing this code, the value of @code{foo} is five: @example foo = "a string" foo = foo + 5 @end example @noindent (Note that using a variable as a number and then later as a string can be confusing and is poor programming style. The above examples illustrate how @code{awk} works, @emph{not} how you should write your own programs!) An assignment is an expression, so it has a value: the same value that is assigned. Thus, @samp{z = 1} as an expression has the value one. One consequence of this is that you can write multiple assignments together: @example x = y = z = 0 @end example @noindent stores the value zero in all three variables. It does this because the value of @samp{z = 0}, which is zero, is stored into @code{y}, and then the value of @samp{y = z = 0}, which is zero, is stored into @code{x}. You can use an assignment anywhere an expression is called for. For example, it is valid to write @samp{x != (y = 1)} to set @code{y} to one and then test whether @code{x} equals one. But this style tends to make programs hard to read; except in a one-shot program, you should not use such nesting of assignments. Aside from @samp{=}, there are several other assignment operators that do arithmetic with the old value of the variable. For example, the operator @samp{+=} computes a new value by adding the right-hand value to the old value of the variable. Thus, the following assignment adds five to the value of @code{foo}: @example foo += 5 @end example @noindent This is equivalent to the following: @example foo = foo + 5 @end example @noindent Use whichever one makes the meaning of your program clearer. There are situations where using @samp{+=} (or any assignment operator) is @emph{not} the same as simply repeating the left-hand operand in the right-hand expression. For example: @cindex Rankin, Pat @example @group # Thanks to Pat Rankin for this example BEGIN @{ foo[rand()] += 5 for (x in foo) print x, foo[x] bar[rand()] = bar[rand()] + 5 for (x in bar) print x, bar[x] @} @end group @end example @noindent The indices of @code{bar} are guaranteed to be different, because @code{rand} will return different values each time it is called. (Arrays and the @code{rand} function haven't been covered yet. @xref{Arrays, ,Arrays in @code{awk}}, and see @ref{Numeric Functions, ,Numeric Built-in Functions}, for more information). This example illustrates an important fact about the assignment operators: the left-hand expression is only evaluated @emph{once}. It is also up to the implementation as to which expression is evaluated first, the left-hand one or the right-hand one. Consider this example: @example i = 1 a[i += 2] = i + 1 @end example @noindent The value of @code{a[3]} could be either two or four. Here is a table of the arithmetic assignment operators. In each case, the right-hand operand is an expression whose value is converted to a number. @c @cartouche @table @code @item @var{lvalue} += @var{increment} Adds @var{increment} to the value of @var{lvalue} to make the new value of @var{lvalue}. @item @var{lvalue} -= @var{decrement} Subtracts @var{decrement} from the value of @var{lvalue}. @item @var{lvalue} *= @var{coefficient} Multiplies the value of @var{lvalue} by @var{coefficient}. @item @var{lvalue} /= @var{divisor} Divides the value of @var{lvalue} by @var{divisor}. @item @var{lvalue} %= @var{modulus} Sets @var{lvalue} to its remainder by @var{modulus}. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @item @var{lvalue} ^= @var{power} @itemx @var{lvalue} **= @var{power} Raises @var{lvalue} to the power @var{power}. (Only the @samp{^=} operator is specified by POSIX.) @end table @c @end cartouche For maximum portability, do not use the @samp{**=} operator. @node Increment Ops, Truth Values, Assignment Ops, Expressions @section Increment and Decrement Operators @cindex increment operators @cindex operators, increment @dfn{Increment} and @dfn{decrement operators} increase or decrease the value of a variable by one. You could do the same thing with an assignment operator, so the increment operators add no power to the @code{awk} language; but they are convenient abbreviations for very common operations. The operator to add one is written @samp{++}. It can be used to increment a variable either before or after taking its value. To pre-increment a variable @var{v}, write @samp{++@var{v}}. This adds one to the value of @var{v} and that new value is also the value of this expression. The assignment expression @samp{@var{v} += 1} is completely equivalent. Writing the @samp{++} after the variable specifies post-increment. This increments the variable value just the same; the difference is that the value of the increment expression itself is the variable's @emph{old} value. Thus, if @code{foo} has the value four, then the expression @samp{foo++} has the value four, but it changes the value of @code{foo} to five. The post-increment @samp{foo++} is nearly equivalent to writing @samp{(foo += 1) - 1}. It is not perfectly equivalent because all numbers in @code{awk} are floating point: in floating point, @samp{foo + 1 - 1} does not necessarily equal @code{foo}. But the difference is minute as long as you stick to numbers that are fairly small (less than 10e12). Any lvalue can be incremented. Fields and array elements are incremented just like variables. (Use @samp{$(i++)} when you wish to do a field reference and a variable increment at the same time. The parentheses are necessary because of the precedence of the field reference operator, @samp{$}.) @cindex decrement operators @cindex operators, decrement The decrement operator @samp{--} works just like @samp{++} except that it subtracts one instead of adding. Like @samp{++}, it can be used before the lvalue to pre-decrement or after it to post-decrement. Here is a summary of increment and decrement expressions. @c @cartouche @table @code @item ++@var{lvalue} This expression increments @var{lvalue} and the new value becomes the value of the expression. @item @var{lvalue}++ This expression increments @var{lvalue}, but the value of the expression is the @emph{old} value of @var{lvalue}. @item --@var{lvalue} Like @samp{++@var{lvalue}}, but instead of adding, it subtracts. It decrements @var{lvalue} and delivers the value that results. @item @var{lvalue}-- Like @samp{@var{lvalue}++}, but instead of adding, it subtracts. It decrements @var{lvalue}. The value of the expression is the @emph{old} value of @var{lvalue}. @end table @c @end cartouche @node Truth Values, Typing and Comparison, Increment Ops, Expressions @section True and False in @code{awk} @cindex truth values @cindex logical true @cindex logical false Many programming languages have a special representation for the concepts of ``true'' and ``false.'' Such languages usually use the special constants @code{true} and @code{false}, or perhaps their upper-case equivalents. @cindex null string @cindex empty string @code{awk} is different. It borrows a very simple concept of true and false from C. In @code{awk}, any non-zero numeric value, @emph{or} any non-empty string value is true. Any other value (zero or the null string, @code{""}) is false. The following program will print @samp{A strange truth value} three times: @example @group BEGIN @{ if (3.1415927) print "A strange truth value" if ("Four Score And Seven Years Ago") print "A strange truth value" if (j = 57) print "A strange truth value" @} @end group @end example @cindex dark corner There is a surprising consequence of the ``non-zero or non-null'' rule: The string constant @code{"0"} is actually true, since it is non-null (d.c.). @node Typing and Comparison, Boolean Ops, Truth Values, Expressions @section Variable Typing and Comparison Expressions @cindex comparison expressions @cindex expression, comparison @cindex expression, matching @cindex relational operators @cindex operators, relational @cindex regexp match/non-match operators @cindex variable typing @cindex types of variables @c 2e: consider splitting this section into subsections @display @i{The Guide is definitive. Reality is frequently inaccurate.} The Hitchhiker's Guide to the Galaxy @end display @sp 1 Unlike other programming languages, @code{awk} variables do not have a fixed type. Instead, they can be either a number or a string, depending upon the value that is assigned to them. @cindex numeric string The 1992 POSIX standard introduced the concept of a @dfn{numeric string}, which is simply a string that looks like a number, for example, @code{@w{" +2"}}. This concept is used for determining the type of a variable. The type of the variable is important, since the types of two variables determine how they are compared. In @code{gawk}, variable typing follows these rules. @enumerate 1 @item A numeric literal or the result of a numeric operation has the @var{numeric} attribute. @item A string literal or the result of a string operation has the @var{string} attribute. @item Fields, @code{getline} input, @code{FILENAME}, @code{ARGV} elements, @code{ENVIRON} elements and the elements of an array created by @code{split} that are numeric strings have the @var{strnum} attribute. Otherwise, they have the @var{string} attribute. Uninitialized variables also have the @var{strnum} attribute. @item Attributes propagate across assignments, but are not changed by any use. @c (Although a use may cause the entity to acquire an additional @c value such that it has both a numeric and string value -- this leaves the @c attribute unchanged.) @c This is important but not relevant @end enumerate The last rule is particularly important. In the following program, @code{a} has numeric type, even though it is later used in a string operation. @example BEGIN @{ a = 12.345 b = a " is a cute number" print b @} @end example When two operands are compared, either string comparison or numeric comparison may be used, depending on the attributes of the operands, according to the following, symmetric, matrix: @c thanks to Karl Berry, kb@cs.umb.edu, for major help with TeX tables @tex \centerline{ \vbox{\bigskip % space above the table (about 1 linespace) % Because we have vertical rules, we can't let TeX insert interline space % in its usual way. \offinterlineskip % % Define the table template. & separates columns, and \cr ends the % template (and each row). # is replaced by the text of that entry on % each row. The template for the first column breaks down like this: % \strut -- a way to make each line have the height and depth % of a normal line of type, since we turned off interline spacing. % \hfil -- infinite glue; has the effect of right-justifying in this case. % # -- replaced by the text (for instance, `STRNUM', in the last row). % \quad -- about the width of an `M'. Just separates the columns. % % The second column (\vrule#) is what generates the vertical rule that % spans table rows. % % The doubled && before the next entry means `repeat the following % template as many times as necessary on each line' -- in our case, twice. % % The template itself, \quad#\hfil, left-justifies with a little space before. % \halign{\strut\hfil#\quad&\vrule#&&\quad#\hfil\cr &&STRING &NUMERIC &STRNUM\cr % The \omit tells TeX to skip inserting the template for this column on % this particular row. In this case, we only want a little extra space % to separate the heading row from the rule below it. the depth 2pt -- % `\vrule depth 2pt' is that little space. \omit &depth 2pt\cr % This is the horizontal rule below the heading. Since it has nothing to % do with the columns of the table, we use \noalign to get it in there. \noalign{\hrule} % Like above, this time a little more space. \omit &depth 4pt\cr % The remaining rows have nothing special about them. STRING &&string &string &string\cr NUMERIC &&string &numeric &numeric\cr STRNUM &&string &numeric &numeric\cr }}} @end tex @ifinfo @display +---------------------------------------------- | STRING NUMERIC STRNUM --------+---------------------------------------------- | STRING | string string string | NUMERIC | string numeric numeric | STRNUM | string numeric numeric --------+---------------------------------------------- @end display @end ifinfo The basic idea is that user input that looks numeric, and @emph{only} user input, should be treated as numeric, even though it is actually made of characters, and is therefore also a string. @dfn{Comparison expressions} compare strings or numbers for relationships such as equality. They are written using @dfn{relational operators}, which are a superset of those in C. Here is a table of them: @cindex relational operators @cindex operators, relational @cindex @code{<} operator @cindex @code{<=} operator @cindex @code{>} operator @cindex @code{>=} operator @cindex @code{==} operator @cindex @code{!=} operator @cindex @code{~} operator @cindex @code{!~} operator @cindex @code{in} operator @c @cartouche @table @code @item @var{x} < @var{y} True if @var{x} is less than @var{y}. @item @var{x} <= @var{y} True if @var{x} is less than or equal to @var{y}. @item @var{x} > @var{y} True if @var{x} is greater than @var{y}. @item @var{x} >= @var{y} True if @var{x} is greater than or equal to @var{y}. @item @var{x} == @var{y} True if @var{x} is equal to @var{y}. @item @var{x} != @var{y} True if @var{x} is not equal to @var{y}. @item @var{x} ~ @var{y} True if the string @var{x} matches the regexp denoted by @var{y}. @item @var{x} !~ @var{y} True if the string @var{x} does not match the regexp denoted by @var{y}. @item @var{subscript} in @var{array} True if the array @var{array} has an element with the subscript @var{subscript}. @end table @c @end cartouche Comparison expressions have the value one if true and zero if false. When comparing operands of mixed types, numeric operands are converted to strings using the value of @code{CONVFMT} (@pxref{Conversion, ,Conversion of Strings and Numbers}). Strings are compared by comparing the first character of each, then the second character of each, and so on. Thus @code{"10"} is less than @code{"9"}. If there are two strings where one is a prefix of the other, the shorter string is less than the longer one. Thus @code{"abc"} is less than @code{"abcd"}. @cindex common mistakes @cindex mistakes, common @cindex errors, common It is very easy to accidentally mistype the @samp{==} operator, and leave off one of the @samp{=}s. The result is still valid @code{awk} code, but the program will not do what you mean: @example if (a = b) # oops! should be a == b @dots{} else @dots{} @end example @noindent Unless @code{b} happens to be zero or the null string, the @code{if} part of the test will always succeed. Because the operators are so similar, this kind of error is very difficult to spot when scanning the source code. Here are some sample expressions, how @code{gawk} compares them, and what the result of the comparison is. @table @code @item 1.5 <= 2.0 numeric comparison (true) @item "abc" >= "xyz" string comparison (false) @item 1.5 != " +2" string comparison (true) @item "1e2" < "3" string comparison (true) @item a = 2; b = "2" @itemx a == b string comparison (true) @item a = 2; b = " +2" @itemx a == b string comparison (false) @end table In this example, @example @group $ echo 1e2 3 | awk '@{ print ($1 < $2) ? "true" : "false" @}' @print{} false @end group @end example @noindent the result is @samp{false} since both @code{$1} and @code{$2} are numeric strings and thus both have the @var{strnum} attribute, dictating a numeric comparison. The purpose of the comparison rules and the use of numeric strings is to attempt to produce the behavior that is ``least surprising,'' while still ``doing the right thing.'' @cindex comparisons, string vs. regexp @cindex string comparison vs. regexp comparison @cindex regexp comparison vs. string comparison String comparisons and regular expression comparisons are very different. For example, @example x == "foo" @end example @noindent has the value of one, or is true, if the variable @code{x} is precisely @samp{foo}. By contrast, @example x ~ /foo/ @end example @noindent has the value one if @code{x} contains @samp{foo}, such as @code{"Oh, what a fool am I!"}. The right hand operand of the @samp{~} and @samp{!~} operators may be either a regexp constant (@code{/@dots{}/}), or an ordinary expression, in which case the value of the expression as a string is used as a dynamic regexp (@pxref{Regexp Usage, ,How to Use Regular Expressions}; also @pxref{Computed Regexps, ,Using Dynamic Regexps}). @cindex regexp as expression In recent implementations of @code{awk}, a constant regular expression in slashes by itself is also an expression. The regexp @code{/@var{regexp}/} is an abbreviation for this comparison expression: @example $0 ~ /@var{regexp}/ @end example One special place where @code{/foo/} is @emph{not} an abbreviation for @samp{$0 ~ /foo/} is when it is the right-hand operand of @samp{~} or @samp{!~}! @xref{Using Constant Regexps, ,Using Regular Expression Constants}, where this is discussed in more detail. @c This paragraph has been here since day 1, and has always bothered @c me, especially since the expression doesn't really make a lot of @c sense. So, just take it out. @ignore In some contexts it may be necessary to write parentheses around the regexp to avoid confusing the @code{gawk} parser. For example, @samp{(/x/ - /y/) > threshold} is not allowed, but @samp{((/x/) - (/y/)) > threshold} parses properly. @end ignore @node Boolean Ops, Conditional Exp, Typing and Comparison, Expressions @section Boolean Expressions @cindex expression, boolean @cindex boolean expressions @cindex operators, boolean @cindex boolean operators @cindex logical operations @cindex operations, logical @cindex short-circuit operators @cindex operators, short-circuit @cindex and operator @cindex or operator @cindex not operator @cindex @code{&&} operator @cindex @code{||} operator @cindex @code{!} operator A @dfn{boolean expression} is a combination of comparison expressions or matching expressions, using the boolean operators ``or'' (@samp{||}), ``and'' (@samp{&&}), and ``not'' (@samp{!}), along with parentheses to control nesting. The truth value of the boolean expression is computed by combining the truth values of the component expressions. Boolean expressions are also referred to as @dfn{logical expressions}. The terms are equivalent. Boolean expressions can be used wherever comparison and matching expressions can be used. They can be used in @code{if}, @code{while}, @code{do} and @code{for} statements (@pxref{Statements, ,Control Statements in Actions}). They have numeric values (one if true, zero if false), which come into play if the result of the boolean expression is stored in a variable, or used in arithmetic. In addition, every boolean expression is also a valid pattern, so you can use one as a pattern to control the execution of rules. Here are descriptions of the three boolean operators, with examples. @c @cartouche @table @code @item @var{boolean1} && @var{boolean2} True if both @var{boolean1} and @var{boolean2} are true. For example, the following statement prints the current input record if it contains both @samp{2400} and @samp{foo}. @example if ($0 ~ /2400/ && $0 ~ /foo/) print @end example The subexpression @var{boolean2} is evaluated only if @var{boolean1} is true. This can make a difference when @var{boolean2} contains expressions that have side effects: in the case of @samp{$0 ~ /foo/ && ($2 == bar++)}, the variable @code{bar} is not incremented if there is no @samp{foo} in the record. @item @var{boolean1} || @var{boolean2} True if at least one of @var{boolean1} or @var{boolean2} is true. For example, the following statement prints all records in the input that contain @emph{either} @samp{2400} or @samp{foo}, or both. @example if ($0 ~ /2400/ || $0 ~ /foo/) print @end example The subexpression @var{boolean2} is evaluated only if @var{boolean1} is false. This can make a difference when @var{boolean2} contains expressions that have side effects. @item ! @var{boolean} True if @var{boolean} is false. For example, the following program prints all records in the input file @file{BBS-list} that do @emph{not} contain the string @samp{foo}. @c A better example would be `if (! (subscript in array)) ...' but we @c haven't done anything with arrays or `in' yet. Sigh. @example awk '@{ if (! ($0 ~ /foo/)) print @}' BBS-list @end example @end table @c @end cartouche The @samp{&&} and @samp{||} operators are called @dfn{short-circuit} operators because of the way they work. Evaluation of the full expression is ``short-circuited'' if the result can be determined part way through its evaluation. @cindex line continuation You can continue a statement that uses @samp{&&} or @samp{||} simply by putting a newline after them. But you cannot put a newline in front of either of these operators without using backslash continuation (@pxref{Statements/Lines, ,@code{awk} Statements Versus Lines}). The actual value of an expression using the @samp{!} operator will be either one or zero, depending upon the truth value of the expression it is applied to. The @samp{!} operator is often useful for changing the sense of a flag variable from false to true and back again. For example, the following program is one way to print lines in between special bracketing lines: @example $1 == "START" @{ interested = ! interested @} interested == 1 @{ print @} $1 == "END" @{ interested = ! interested @} @end example @noindent The variable @code{interested}, like all @code{awk} variables, starts out initialized to zero, which is also false. When a line is seen whose first field is @samp{START}, the value of @code{interested} is toggled to true, using @samp{!}. The next rule prints lines as long as @code{interested} is true. When a line is seen whose first field is @samp{END}, @code{interested} is toggled back to false. @ignore We should discuss using `next' in the two rules that toggle the variable, to avoid printing the bracketing lines, but that's more distraction than really needed. @end ignore @node Conditional Exp, Function Calls, Boolean Ops, Expressions @section Conditional Expressions @cindex conditional expression @cindex expression, conditional A @dfn{conditional expression} is a special kind of expression with three operands. It allows you to use one expression's value to select one of two other expressions. The conditional expression is the same as in the C language: @example @var{selector} ? @var{if-true-exp} : @var{if-false-exp} @end example @noindent There are three subexpressions. The first, @var{selector}, is always computed first. If it is ``true'' (not zero and not null) then @var{if-true-exp} is computed next and its value becomes the value of the whole expression. Otherwise, @var{if-false-exp} is computed next and its value becomes the value of the whole expression. For example, this expression produces the absolute value of @code{x}: @example x > 0 ? x : -x @end example Each time the conditional expression is computed, exactly one of @var{if-true-exp} and @var{if-false-exp} is computed; the other is ignored. This is important when the expressions contain side effects. For example, this conditional expression examines element @code{i} of either array @code{a} or array @code{b}, and increments @code{i}. @example x == y ? a[i++] : b[i++] @end example @noindent This is guaranteed to increment @code{i} exactly once, because each time only one of the two increment expressions is executed, and the other is not. @xref{Arrays, ,Arrays in @code{awk}}, for more information about arrays. @cindex differences between @code{gawk} and @code{awk} @cindex line continuation As a minor @code{gawk} extension, you can continue a statement that uses @samp{?:} simply by putting a newline after either character. However, you cannot put a newline in front of either character without using backslash continuation (@pxref{Statements/Lines, ,@code{awk} Statements Versus Lines}). If @samp{--posix} is specified (@pxref{Options, , Command Line Options}), then this extension is disabled. @node Function Calls, Precedence, Conditional Exp, Expressions @section Function Calls @cindex function call @cindex calling a function A @dfn{function} is a name for a particular calculation. Because it has a name, you can ask for it by name at any point in the program. For example, the function @code{sqrt} computes the square root of a number. A fixed set of functions are @dfn{built-in}, which means they are available in every @code{awk} program. The @code{sqrt} function is one of these. @xref{Built-in, ,Built-in Functions}, for a list of built-in functions and their descriptions. In addition, you can define your own functions for use in your program. @xref{User-defined, ,User-defined Functions}, for how to do this. @cindex arguments in function call The way to use a function is with a @dfn{function call} expression, which consists of the function name followed immediately by a list of @dfn{arguments} in parentheses. The arguments are expressions which provide the raw materials for the function's calculations. When there is more than one argument, they are separated by commas. If there are no arguments, write just @samp{()} after the function name. Here are some examples: @example sqrt(x^2 + y^2) @i{one argument} atan2(y, x) @i{two arguments} rand() @i{no arguments} @end example @strong{Do not put any space between the function name and the open-parenthesis!} A user-defined function name looks just like the name of a variable, and space would make the expression look like concatenation of a variable with an expression inside parentheses. Space before the parenthesis is harmless with built-in functions, but it is best not to get into the habit of using space to avoid mistakes with user-defined functions. Each function expects a particular number of arguments. For example, the @code{sqrt} function must be called with a single argument, the number to take the square root of: @example sqrt(@var{argument}) @end example Some of the built-in functions allow you to omit the final argument. If you do so, they use a reasonable default. @xref{Built-in, ,Built-in Functions}, for full details. If arguments are omitted in calls to user-defined functions, then those arguments are treated as local variables, initialized to the empty string (@pxref{User-defined, ,User-defined Functions}). Like every other expression, the function call has a value, which is computed by the function based on the arguments you give it. In this example, the value of @samp{sqrt(@var{argument})} is the square root of @var{argument}. A function can also have side effects, such as assigning values to certain variables or doing I/O. Here is a command to read numbers, one number per line, and print the square root of each one: @example @group $ awk '@{ print "The square root of", $1, "is", sqrt($1) @}' 1 @print{} The square root of 1 is 1 3 @print{} The square root of 3 is 1.73205 5 @print{} The square root of 5 is 2.23607 @kbd{Control-d} @end group @end example @node Precedence, , Function Calls, Expressions @section Operator Precedence (How Operators Nest) @cindex precedence @cindex operator precedence @dfn{Operator precedence} determines how operators are grouped, when different operators appear close by in one expression. For example, @samp{*} has higher precedence than @samp{+}; thus, @samp{a + b * c} means to multiply @code{b} and @code{c}, and then add @code{a} to the product (i.e.@: @samp{a + (b * c)}). You can overrule the precedence of the operators by using parentheses. You can think of the precedence rules as saying where the parentheses are assumed to be if you do not write parentheses yourself. In fact, it is wise to always use parentheses whenever you have an unusual combination of operators, because other people who read the program may not remember what the precedence is in this case. You might forget, too; then you could make a mistake. Explicit parentheses will help prevent any such mistake. When operators of equal precedence are used together, the leftmost operator groups first, except for the assignment, conditional and exponentiation operators, which group in the opposite order. Thus, @samp{a - b + c} groups as @samp{(a - b) + c}, and @samp{a = b = c} groups as @samp{a = (b = c)}. The precedence of prefix unary operators does not matter as long as only unary operators are involved, because there is only one way to interpret them---innermost first. Thus, @samp{$++i} means @samp{$(++i)} and @samp{++$x} means @samp{++($x)}. However, when another operator follows the operand, then the precedence of the unary operators can matter. Thus, @samp{$x^2} means @samp{($x)^2}, but @samp{-x^2} means @samp{-(x^2)}, because @samp{-} has lower precedence than @samp{^} while @samp{$} has higher precedence. Here is a table of @code{awk}'s operators, in order from highest precedence to lowest: @c use @code in the items, looks better in TeX w/o all the quotes @table @code @item (@dots{}) Grouping. @item $ Field. @item ++ -- Increment, decrement. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @item ^ ** Exponentiation. These operators group right-to-left. (The @samp{**} operator is not specified by POSIX.) @item + - ! Unary plus, minus, logical ``not''. @item * / % Multiplication, division, modulus. @item + - Addition, subtraction. @item @r{Concatenation} No special token is used to indicate concatenation. The operands are simply written side by side. @item < <= == != @itemx > >= >> | Relational, and redirection. The relational operators and the redirections have the same precedence level. Characters such as @samp{>} serve both as relationals and as redirections; the context distinguishes between the two meanings. Note that the I/O redirection operators in @code{print} and @code{printf} statements belong to the statement level, not to expressions. The redirection does not produce an expression which could be the operand of another operator. As a result, it does not make sense to use a redirection operator near another operator of lower precedence, without parentheses. Such combinations, for example @samp{print foo > a ? b : c}, result in syntax errors. The correct way to write this statement is @samp{print foo > (a ? b : c)}. @item ~ !~ Matching, non-matching. @item in Array membership. @item && Logical ``and''. @item || Logical ``or''. @item ?: Conditional. This operator groups right-to-left. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @item = += -= *= @itemx /= %= ^= **= Assignment. These operators group right-to-left. (The @samp{**=} operator is not specified by POSIX.) @end table @node Patterns and Actions, Statements, Expressions, Top @chapter Patterns and Actions @cindex pattern, definition of As you have already seen, each @code{awk} statement consists of a pattern with an associated action. This chapter describes how you build patterns and actions. @menu * Pattern Overview:: What goes into a pattern. * Action Overview:: What goes into an action. @end menu @node Pattern Overview, Action Overview, Patterns and Actions, Patterns and Actions @section Pattern Elements Patterns in @code{awk} control the execution of rules: a rule is executed when its pattern matches the current input record. This section explains all about how to write patterns. @menu * Kinds of Patterns:: A list of all kinds of patterns. * Regexp Patterns:: Using regexps as patterns. * Expression Patterns:: Any expression can be used as a pattern. * Ranges:: Pairs of patterns specify record ranges. * BEGIN/END:: Specifying initialization and cleanup rules. * Empty:: The empty pattern, which matches every record. @end menu @node Kinds of Patterns, Regexp Patterns, Pattern Overview, Pattern Overview @subsection Kinds of Patterns @cindex patterns, types of Here is a summary of the types of patterns supported in @code{awk}. @table @code @item /@var{regular expression}/ A regular expression as a pattern. It matches when the text of the input record fits the regular expression. (@xref{Regexp, ,Regular Expressions}.) @item @var{expression} A single expression. It matches when its value is non-zero (if a number) or non-null (if a string). (@xref{Expression Patterns, ,Expressions as Patterns}.) @item @var{pat1}, @var{pat2} A pair of patterns separated by a comma, specifying a range of records. The range includes both the initial record that matches @var{pat1}, and the final record that matches @var{pat2}. (@xref{Ranges, ,Specifying Record Ranges with Patterns}.) @item BEGIN @itemx END Special patterns for you to supply start-up or clean-up actions for your @code{awk} program. (@xref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}.) @item @var{empty} The empty pattern matches every input record. (@xref{Empty, ,The Empty Pattern}.) @end table @node Regexp Patterns, Expression Patterns, Kinds of Patterns, Pattern Overview @subsection Regular Expressions as Patterns We have been using regular expressions as patterns since our early examples. This kind of pattern is simply a regexp constant in the pattern part of a rule. Its meaning is @samp{$0 ~ /@var{pattern}/}. The pattern matches when the input record matches the regexp. For example: @example /foo|bar|baz/ @{ buzzwords++ @} END @{ print buzzwords, "buzzwords seen" @} @end example @node Expression Patterns, Ranges, Regexp Patterns, Pattern Overview @subsection Expressions as Patterns Any @code{awk} expression is valid as an @code{awk} pattern. Then the pattern matches if the expression's value is non-zero (if a number) or non-null (if a string). The expression is reevaluated each time the rule is tested against a new input record. If the expression uses fields such as @code{$1}, the value depends directly on the new input record's text; otherwise, it depends only on what has happened so far in the execution of the @code{awk} program, but that may still be useful. A very common kind of expression used as a pattern is the comparison expression, using the comparison operators described in @ref{Typing and Comparison, ,Variable Typing and Comparison Expressions}. Regexp matching and non-matching are also very common expressions. The left operand of the @samp{~} and @samp{!~} operators is a string. The right operand is either a constant regular expression enclosed in slashes (@code{/@var{regexp}/}), or any expression, whose string value is used as a dynamic regular expression (@pxref{Computed Regexps, , Using Dynamic Regexps}). The following example prints the second field of each input record whose first field is precisely @samp{foo}. @example $ awk '$1 == "foo" @{ print $2 @}' BBS-list @end example @noindent (There is no output, since there is no BBS site named ``foo''.) Contrast this with the following regular expression match, which would accept any record with a first field that contains @samp{foo}: @example @group $ awk '$1 ~ /foo/ @{ print $2 @}' BBS-list @print{} 555-1234 @print{} 555-6699 @print{} 555-6480 @print{} 555-2127 @end group @end example Boolean expressions are also commonly used as patterns. Whether the pattern matches an input record depends on whether its subexpressions match. For example, the following command prints all records in @file{BBS-list} that contain both @samp{2400} and @samp{foo}. @example $ awk '/2400/ && /foo/' BBS-list @print{} fooey 555-1234 2400/1200/300 B @end example The following command prints all records in @file{BBS-list} that contain @emph{either} @samp{2400} or @samp{foo}, or both. @example @group $ awk '/2400/ || /foo/' BBS-list @print{} alpo-net 555-3412 2400/1200/300 A @print{} bites 555-1675 2400/1200/300 A @print{} fooey 555-1234 2400/1200/300 B @print{} foot 555-6699 1200/300 B @print{} macfoo 555-6480 1200/300 A @print{} sdace 555-3430 2400/1200/300 A @print{} sabafoo 555-2127 1200/300 C @end group @end example The following command prints all records in @file{BBS-list} that do @emph{not} contain the string @samp{foo}. @example @group $ awk '! /foo/' BBS-list @print{} aardvark 555-5553 1200/300 B @print{} alpo-net 555-3412 2400/1200/300 A @print{} barfly 555-7685 1200/300 A @print{} bites 555-1675 2400/1200/300 A @print{} camelot 555-0542 300 C @print{} core 555-2912 1200/300 C @print{} sdace 555-3430 2400/1200/300 A @end group @end example The subexpressions of a boolean operator in a pattern can be constant regular expressions, comparisons, or any other @code{awk} expressions. Range patterns are not expressions, so they cannot appear inside boolean patterns. Likewise, the special patterns @code{BEGIN} and @code{END}, which never match any input record, are not expressions and cannot appear inside boolean patterns. A regexp constant as a pattern is also a special case of an expression pattern. @code{/foo/} as an expression has the value one if @samp{foo} appears in the current input record; thus, as a pattern, @code{/foo/} matches any record containing @samp{foo}. @node Ranges, BEGIN/END, Expression Patterns, Pattern Overview @subsection Specifying Record Ranges with Patterns @cindex range pattern @cindex pattern, range @cindex matching ranges of lines A @dfn{range pattern} is made of two patterns separated by a comma, of the form @samp{@var{begpat}, @var{endpat}}. It matches ranges of consecutive input records. The first pattern, @var{begpat}, controls where the range begins, and the second one, @var{endpat}, controls where it ends. For example, @example awk '$1 == "on", $1 == "off"' @end example @noindent prints every record between @samp{on}/@samp{off} pairs, inclusive. A range pattern starts out by matching @var{begpat} against every input record; when a record matches @var{begpat}, the range pattern becomes @dfn{turned on}. The range pattern matches this record. As long as it stays turned on, it automatically matches every input record read. It also matches @var{endpat} against every input record; when that succeeds, the range pattern is turned off again for the following record. Then it goes back to checking @var{begpat} against each record. The record that turns on the range pattern and the one that turns it off both match the range pattern. If you don't want to operate on these records, you can write @code{if} statements in the rule's action to distinguish them from the records you are interested in. It is possible for a pattern to be turned both on and off by the same record, if the record satisfies both conditions. Then the action is executed for just that record. For example, suppose you have text between two identical markers (say the @samp{%} symbol) that you wish to ignore. You might try to combine a range pattern that describes the delimited text with the @code{next} statement (not discussed yet, @pxref{Next Statement, , The @code{next} Statement}), which causes @code{awk} to skip any further processing of the current record and start over again with the next input record. Such a program would look like this: @example /^%$/,/^%$/ @{ next @} @{ print @} @end example @noindent @cindex skipping lines between markers This program fails because the range pattern is both turned on and turned off by the first line with just a @samp{%} on it. To accomplish this task, you must write the program this way, using a flag: @example /^%$/ @{ skip = ! skip; next @} skip == 1 @{ next @} # skip lines with `skip' set @end example Note that in a range pattern, the @samp{,} has the lowest precedence (is evaluated last) of all the operators. Thus, for example, the following program attempts to combine a range pattern with another, simpler test. @example echo Yes | awk '/1/,/2/ || /Yes/' @end example The author of this program intended it to mean @samp{(/1/,/2/) || /Yes/}. However, @code{awk} interprets this as @samp{/1/, (/2/ || /Yes/)}. This cannot be changed or worked around; range patterns do not combine with other patterns. @node BEGIN/END, Empty, Ranges, Pattern Overview @subsection The @code{BEGIN} and @code{END} Special Patterns @cindex @code{BEGIN} special pattern @cindex pattern, @code{BEGIN} @cindex @code{END} special pattern @cindex pattern, @code{END} @code{BEGIN} and @code{END} are special patterns. They are not used to match input records. Rather, they supply start-up or clean-up actions for your @code{awk} script. @menu * Using BEGIN/END:: How and why to use BEGIN/END rules. * I/O And BEGIN/END:: I/O issues in BEGIN/END rules. @end menu @node Using BEGIN/END, I/O And BEGIN/END, BEGIN/END, BEGIN/END @subsubsection Startup and Cleanup Actions A @code{BEGIN} rule is executed, once, before the first input record has been read. An @code{END} rule is executed, once, after all the input has been read. For example: @example @group $ awk ' > BEGIN @{ print "Analysis of \"foo\"" @} > /foo/ @{ ++n @} > END @{ print "\"foo\" appears " n " times." @}' BBS-list @print{} Analysis of "foo" @print{} "foo" appears 4 times. @end group @end example This program finds the number of records in the input file @file{BBS-list} that contain the string @samp{foo}. The @code{BEGIN} rule prints a title for the report. There is no need to use the @code{BEGIN} rule to initialize the counter @code{n} to zero, as @code{awk} does this automatically (@pxref{Variables}). The second rule increments the variable @code{n} every time a record containing the pattern @samp{foo} is read. The @code{END} rule prints the value of @code{n} at the end of the run. The special patterns @code{BEGIN} and @code{END} cannot be used in ranges or with boolean operators (indeed, they cannot be used with any operators). An @code{awk} program may have multiple @code{BEGIN} and/or @code{END} rules. They are executed in the order they appear, all the @code{BEGIN} rules at start-up and all the @code{END} rules at termination. @code{BEGIN} and @code{END} rules may be intermixed with other rules. This feature was added in the 1987 version of @code{awk}, and is included in the POSIX standard. The original (1978) version of @code{awk} required you to put the @code{BEGIN} rule at the beginning of the program, and the @code{END} rule at the end, and only allowed one of each. This is no longer required, but it is a good idea in terms of program organization and readability. Multiple @code{BEGIN} and @code{END} rules are useful for writing library functions, since each library file can have its own @code{BEGIN} and/or @code{END} rule to do its own initialization and/or cleanup. Note that the order in which library functions are named on the command line controls the order in which their @code{BEGIN} and @code{END} rules are executed. Therefore you have to be careful to write such rules in library files so that the order in which they are executed doesn't matter. @xref{Options, ,Command Line Options}, for more information on using library functions. @xref{Library Functions, ,A Library of @code{awk} Functions}, for a number of useful library functions. @cindex dark corner If an @code{awk} program only has a @code{BEGIN} rule, and no other rules, then the program exits after the @code{BEGIN} rule has been run. (The original version of @code{awk} used to keep reading and ignoring input until end of file was seen.) However, if an @code{END} rule exists, then the input will be read, even if there are no other rules in the program. This is necessary in case the @code{END} rule checks the @code{FNR} and @code{NR} variables (d.c.). @code{BEGIN} and @code{END} rules must have actions; there is no default action for these rules since there is no current record when they run. @node I/O And BEGIN/END, , Using BEGIN/END, BEGIN/END @subsubsection Input/Output from @code{BEGIN} and @code{END} Rules @cindex I/O from @code{BEGIN} and @code{END} There are several (sometimes subtle) issues involved when doing I/O from a @code{BEGIN} or @code{END} rule. The first has to do with the value of @code{$0} in a @code{BEGIN} rule. Since @code{BEGIN} rules are executed before any input is read, there simply is no input record, and therefore no fields, when executing @code{BEGIN} rules. References to @code{$0} and the fields yield a null string or zero, depending upon the context. One way to give @code{$0} a real value is to execute a @code{getline} command without a variable (@pxref{Getline, ,Explicit Input with @code{getline}}). Another way is to simply assign a value to it. @cindex differences between @code{gawk} and @code{awk} The second point is similar to the first, but from the other direction. Inside an @code{END} rule, what is the value of @code{$0} and @code{NF}? Traditionally, due largely to implementation issues, @code{$0} and @code{NF} were @emph{undefined} inside an @code{END} rule. The POSIX standard specified that @code{NF} was available in an @code{END} rule, containing the number of fields from the last input record. Due most probably to an oversight, the standard does not say that @code{$0} is also preserved, although logically one would think that it should be. In fact, @code{gawk} does preserve the value of @code{$0} for use in @code{END} rules. Be aware, however, that Unix @code{awk}, and possibly other implementations, do not. The third point follows from the first two. What is the meaning of @samp{print} inside a @code{BEGIN} or @code{END} rule? The meaning is the same as always, @samp{print $0}. If @code{$0} is the null string, then this prints an empty line. Many long time @code{awk} programmers use @samp{print} in @code{BEGIN} and @code{END} rules, to mean @samp{@w{print ""}}, relying on @code{$0} being null. While you might generally get away with this in @code{BEGIN} rules, in @code{gawk} at least, it is a very bad idea in @code{END} rules. It is also poor style, since if you want an empty line in the output, you should say so explicitly in your program. @node Empty, , BEGIN/END, Pattern Overview @subsection The Empty Pattern @cindex empty pattern @cindex pattern, empty An empty (i.e.@: non-existent) pattern is considered to match @emph{every} input record. For example, the program: @example awk '@{ print $1 @}' BBS-list @end example @noindent prints the first field of every record. @node Action Overview, , Pattern Overview, Patterns and Actions @section Overview of Actions @cindex action, definition of @cindex curly braces @cindex action, curly braces @cindex action, separating statements An @code{awk} program or script consists of a series of rules and function definitions, interspersed. (Functions are described later. @xref{User-defined, ,User-defined Functions}.) A rule contains a pattern and an action, either of which (but not both) may be omitted. The purpose of the @dfn{action} is to tell @code{awk} what to do once a match for the pattern is found. Thus, in outline, an @code{awk} program generally looks like this: @example @r{[}@var{pattern}@r{]} @r{[}@{ @var{action} @}@r{]} @r{[}@var{pattern}@r{]} @r{[}@{ @var{action} @}@r{]} @dots{} function @var{name}(@var{args}) @{ @dots{} @} @dots{} @end example An action consists of one or more @code{awk} @dfn{statements}, enclosed in curly braces (@samp{@{} and @samp{@}}). Each statement specifies one thing to be done. The statements are separated by newlines or semicolons. The curly braces around an action must be used even if the action contains only one statement, or even if it contains no statements at all. However, if you omit the action entirely, omit the curly braces as well. An omitted action is equivalent to @samp{@{ print $0 @}}. @example /foo/ @{ @} # match foo, do nothing - empty action /foo/ # match foo, print the record - omitted action @end example Here are the kinds of statements supported in @code{awk}: @itemize @bullet @item Expressions, which can call functions or assign values to variables (@pxref{Expressions}). Executing this kind of statement simply computes the value of the expression. This is useful when the expression has side effects (@pxref{Assignment Ops, ,Assignment Expressions}). @item Control statements, which specify the control flow of @code{awk} programs. The @code{awk} language gives you C-like constructs (@code{if}, @code{for}, @code{while}, and @code{do}) as well as a few special ones (@pxref{Statements, ,Control Statements in Actions}). @item Compound statements, which consist of one or more statements enclosed in curly braces. A compound statement is used in order to put several statements together in the body of an @code{if}, @code{while}, @code{do} or @code{for} statement. @item Input statements, using the @code{getline} command (@pxref{Getline, ,Explicit Input with @code{getline}}), the @code{next} statement (@pxref{Next Statement, ,The @code{next} Statement}), and the @code{nextfile} statement (@pxref{Nextfile Statement, ,The @code{nextfile} Statement}). @item Output statements, @code{print} and @code{printf}. @xref{Printing, ,Printing Output}. @item Deletion statements, for deleting array elements. @xref{Delete, ,The @code{delete} Statement}. @end itemize @iftex The next chapter covers control statements in detail. @end iftex @node Statements, Built-in Variables, Patterns and Actions, Top @chapter Control Statements in Actions @cindex control statement @dfn{Control statements} such as @code{if}, @code{while}, and so on control the flow of execution in @code{awk} programs. Most of the control statements in @code{awk} are patterned on similar statements in C. All the control statements start with special keywords such as @code{if} and @code{while}, to distinguish them from simple expressions. @cindex compound statement @cindex statement, compound Many control statements contain other statements; for example, the @code{if} statement contains another statement which may or may not be executed. The contained statement is called the @dfn{body}. If you want to include more than one statement in the body, group them into a single @dfn{compound statement} with curly braces, separating them with newlines or semicolons. @menu * If Statement:: Conditionally execute some @code{awk} statements. * While Statement:: Loop until some condition is satisfied. * Do Statement:: Do specified action while looping until some condition is satisfied. * For Statement:: Another looping statement, that provides initialization and increment clauses. * Break Statement:: Immediately exit the innermost enclosing loop. * Continue Statement:: Skip to the end of the innermost enclosing loop. * Next Statement:: Stop processing the current input record. * Nextfile Statement:: Stop processing the current file. * Exit Statement:: Stop execution of @code{awk}. @end menu @node If Statement, While Statement, Statements, Statements @section The @code{if}-@code{else} Statement @cindex @code{if}-@code{else} statement The @code{if}-@code{else} statement is @code{awk}'s decision-making statement. It looks like this: @example if (@var{condition}) @var{then-body} @r{[}else @var{else-body}@r{]} @end example @noindent The @var{condition} is an expression that controls what the rest of the statement will do. If @var{condition} is true, @var{then-body} is executed; otherwise, @var{else-body} is executed. The @code{else} part of the statement is optional. The condition is considered false if its value is zero or the null string, and true otherwise. Here is an example: @example if (x % 2 == 0) print "x is even" else print "x is odd" @end example In this example, if the expression @samp{x % 2 == 0} is true (that is, the value of @code{x} is evenly divisible by two), then the first @code{print} statement is executed, otherwise the second @code{print} statement is executed. If the @code{else} appears on the same line as @var{then-body}, and @var{then-body} is not a compound statement (i.e.@: not surrounded by curly braces), then a semicolon must separate @var{then-body} from @code{else}. To illustrate this, let's rewrite the previous example: @example if (x % 2 == 0) print "x is even"; else print "x is odd" @end example @noindent If you forget the @samp{;}, @code{awk} won't be able to interpret the statement, and you will get a syntax error. We would not actually write this example this way, because a human reader might fail to see the @code{else} if it were not the first thing on its line. @node While Statement, Do Statement, If Statement, Statements @section The @code{while} Statement @cindex @code{while} statement @cindex loop @cindex body of a loop In programming, a @dfn{loop} means a part of a program that can be executed two or more times in succession. The @code{while} statement is the simplest looping statement in @code{awk}. It repeatedly executes a statement as long as a condition is true. It looks like this: @example while (@var{condition}) @var{body} @end example @noindent Here @var{body} is a statement that we call the @dfn{body} of the loop, and @var{condition} is an expression that controls how long the loop keeps running. The first thing the @code{while} statement does is test @var{condition}. If @var{condition} is true, it executes the statement @var{body}. @ifinfo (The @var{condition} is true when the value is not zero and not a null string.) @end ifinfo After @var{body} has been executed, @var{condition} is tested again, and if it is still true, @var{body} is executed again. This process repeats until @var{condition} is no longer true. If @var{condition} is initially false, the body of the loop is never executed, and @code{awk} continues with the statement following the loop. This example prints the first three fields of each record, one per line. @example awk '@{ i = 1 while (i <= 3) @{ print $i i++ @} @}' inventory-shipped @end example @noindent Here the body of the loop is a compound statement enclosed in braces, containing two statements. The loop works like this: first, the value of @code{i} is set to one. Then, the @code{while} tests whether @code{i} is less than or equal to three. This is true when @code{i} equals one, so the @code{i}-th field is printed. Then the @samp{i++} increments the value of @code{i} and the loop repeats. The loop terminates when @code{i} reaches four. As you can see, a newline is not required between the condition and the body; but using one makes the program clearer unless the body is a compound statement or is very simple. The newline after the open-brace that begins the compound statement is not required either, but the program would be harder to read without it. @node Do Statement, For Statement, While Statement, Statements @section The @code{do}-@code{while} Statement The @code{do} loop is a variation of the @code{while} looping statement. The @code{do} loop executes the @var{body} once, and then repeats @var{body} as long as @var{condition} is true. It looks like this: @example @group do @var{body} while (@var{condition}) @end group @end example Even if @var{condition} is false at the start, @var{body} is executed at least once (and only once, unless executing @var{body} makes @var{condition} true). Contrast this with the corresponding @code{while} statement: @example while (@var{condition}) @var{body} @end example @noindent This statement does not execute @var{body} even once if @var{condition} is false to begin with. Here is an example of a @code{do} statement: @example awk '@{ i = 1 do @{ print $0 i++ @} while (i <= 10) @}' @end example @noindent This program prints each input record ten times. It isn't a very realistic example, since in this case an ordinary @code{while} would do just as well. But this reflects actual experience; there is only occasionally a real use for a @code{do} statement. @node For Statement, Break Statement, Do Statement, Statements @section The @code{for} Statement @cindex @code{for} statement The @code{for} statement makes it more convenient to count iterations of a loop. The general form of the @code{for} statement looks like this: @example for (@var{initialization}; @var{condition}; @var{increment}) @var{body} @end example @noindent The @var{initialization}, @var{condition} and @var{increment} parts are arbitrary @code{awk} expressions, and @var{body} stands for any @code{awk} statement. The @code{for} statement starts by executing @var{initialization}. Then, as long as @var{condition} is true, it repeatedly executes @var{body} and then @var{increment}. Typically @var{initialization} sets a variable to either zero or one, @var{increment} adds one to it, and @var{condition} compares it against the desired number of iterations. Here is an example of a @code{for} statement: @example @group awk '@{ for (i = 1; i <= 3; i++) print $i @}' inventory-shipped @end group @end example @noindent This prints the first three fields of each input record, one field per line. You cannot set more than one variable in the @var{initialization} part unless you use a multiple assignment statement such as @samp{x = y = 0}, which is possible only if all the initial values are equal. (But you can initialize additional variables by writing their assignments as separate statements preceding the @code{for} loop.) The same is true of the @var{increment} part; to increment additional variables, you must write separate statements at the end of the loop. The C compound expression, using C's comma operator, would be useful in this context, but it is not supported in @code{awk}. Most often, @var{increment} is an increment expression, as in the example above. But this is not required; it can be any expression whatever. For example, this statement prints all the powers of two between one and 100: @example for (i = 1; i <= 100; i *= 2) print i @end example Any of the three expressions in the parentheses following the @code{for} may be omitted if there is nothing to be done there. Thus, @w{@samp{for (; x > 0;)}} is equivalent to @w{@samp{while (x > 0)}}. If the @var{condition} is omitted, it is treated as @var{true}, effectively yielding an @dfn{infinite loop} (i.e.@: a loop that will never terminate). In most cases, a @code{for} loop is an abbreviation for a @code{while} loop, as shown here: @example @var{initialization} while (@var{condition}) @{ @var{body} @var{increment} @} @end example @noindent The only exception is when the @code{continue} statement (@pxref{Continue Statement, ,The @code{continue} Statement}) is used inside the loop; changing a @code{for} statement to a @code{while} statement in this way can change the effect of the @code{continue} statement inside the loop. There is an alternate version of the @code{for} loop, for iterating over all the indices of an array: @example for (i in array) @var{do something with} array[i] @end example @noindent @xref{Scanning an Array, ,Scanning All Elements of an Array}, for more information on this version of the @code{for} loop. The @code{awk} language has a @code{for} statement in addition to a @code{while} statement because often a @code{for} loop is both less work to type and more natural to think of. Counting the number of iterations is very common in loops. It can be easier to think of this counting as part of looping rather than as something to do inside the loop. The next section has more complicated examples of @code{for} loops. @node Break Statement, Continue Statement, For Statement, Statements @section The @code{break} Statement @cindex @code{break} statement @cindex loops, exiting The @code{break} statement jumps out of the innermost @code{for}, @code{while}, or @code{do} loop that encloses it. The following example finds the smallest divisor of any integer, and also identifies prime numbers: @example awk '# find smallest divisor of num @{ num = $1 for (div = 2; div*div <= num; div++) if (num % div == 0) break if (num % div == 0) printf "Smallest divisor of %d is %d\n", num, div else printf "%d is prime\n", num @}' @end example When the remainder is zero in the first @code{if} statement, @code{awk} immediately @dfn{breaks out} of the containing @code{for} loop. This means that @code{awk} proceeds immediately to the statement following the loop and continues processing. (This is very different from the @code{exit} statement which stops the entire @code{awk} program. @xref{Exit Statement, ,The @code{exit} Statement}.) Here is another program equivalent to the previous one. It illustrates how the @var{condition} of a @code{for} or @code{while} could just as well be replaced with a @code{break} inside an @code{if}: @example @group awk '# find smallest divisor of num @{ num = $1 for (div = 2; ; div++) @{ if (num % div == 0) @{ printf "Smallest divisor of %d is %d\n", num, div break @} if (div*div > num) @{ printf "%d is prime\n", num break @} @} @}' @end group @end example @cindex @code{break}, outside of loops @cindex historical features @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @cindex dark corner As described above, the @code{break} statement has no meaning when used outside the body of a loop. However, although it was never documented, historical implementations of @code{awk} have treated the @code{break} statement outside of a loop as if it were a @code{next} statement (@pxref{Next Statement, ,The @code{next} Statement}). Recent versions of Unix @code{awk} no longer allow this usage. @code{gawk} will support this use of @code{break} only if @samp{--traditional} has been specified on the command line (@pxref{Options, ,Command Line Options}). Otherwise, it will be treated as an error, since the POSIX standard specifies that @code{break} should only be used inside the body of a loop (d.c.). @node Continue Statement, Next Statement, Break Statement, Statements @section The @code{continue} Statement @cindex @code{continue} statement The @code{continue} statement, like @code{break}, is used only inside @code{for}, @code{while}, and @code{do} loops. It skips over the rest of the loop body, causing the next cycle around the loop to begin immediately. Contrast this with @code{break}, which jumps out of the loop altogether. @c The point of this program was to illustrate the use of continue with @c a while loop. But Karl Berry points out that that is done adequately @c below, and that this example is very un-awk-like. So for now, we'll @c omit it. @ignore In Texinfo source files, text that the author wishes to ignore can be enclosed between lines that start with @samp{@@ignore} and end with @samp{@@end ignore}. Here is a program that strips out lines between @samp{@@ignore} and @samp{@@end ignore} pairs. @example BEGIN @{ while (getline > 0) @{ if (/^@@ignore/) ignoring = 1 else if (/^@@end[ \t]+ignore/) @{ ignoring = 0 continue @} if (ignoring) continue print @} @} @end example When an @samp{@@ignore} is seen, the @code{ignoring} flag is set to one (true). When @samp{@@end ignore} is seen, the flag is reset to zero (false). As long as the flag is true, the input record is not printed, because the @code{continue} restarts the @code{while} loop, skipping over the @code{print} statement. @c Exercise!!! @c How could this program be written to make better use of the awk language? @end ignore The @code{continue} statement in a @code{for} loop directs @code{awk} to skip the rest of the body of the loop, and resume execution with the increment-expression of the @code{for} statement. The following program illustrates this fact: @example awk 'BEGIN @{ for (x = 0; x <= 20; x++) @{ if (x == 5) continue printf "%d ", x @} print "" @}' @end example @noindent This program prints all the numbers from zero to 20, except for five, for which the @code{printf} is skipped. Since the increment @samp{x++} is not skipped, @code{x} does not remain stuck at five. Contrast the @code{for} loop above with this @code{while} loop: @example awk 'BEGIN @{ x = 0 while (x <= 20) @{ if (x == 5) continue printf "%d ", x x++ @} print "" @}' @end example @noindent This program loops forever once @code{x} gets to five. @cindex @code{continue}, outside of loops @cindex historical features @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @cindex dark corner As described above, the @code{continue} statement has no meaning when used outside the body of a loop. However, although it was never documented, historical implementations of @code{awk} have treated the @code{continue} statement outside of a loop as if it were a @code{next} statement (@pxref{Next Statement, ,The @code{next} Statement}). Recent versions of Unix @code{awk} no longer allow this usage. @code{gawk} will support this use of @code{continue} only if @samp{--traditional} has been specified on the command line (@pxref{Options, ,Command Line Options}). Otherwise, it will be treated as an error, since the POSIX standard specifies that @code{continue} should only be used inside the body of a loop (d.c.). @node Next Statement, Nextfile Statement, Continue Statement, Statements @section The @code{next} Statement @cindex @code{next} statement The @code{next} statement forces @code{awk} to immediately stop processing the current record and go on to the next record. This means that no further rules are executed for the current record. The rest of the current rule's action is not executed either. Contrast this with the effect of the @code{getline} function (@pxref{Getline, ,Explicit Input with @code{getline}}). That too causes @code{awk} to read the next record immediately, but it does not alter the flow of control in any way. So the rest of the current action executes with a new input record. At the highest level, @code{awk} program execution is a loop that reads an input record and then tests each rule's pattern against it. If you think of this loop as a @code{for} statement whose body contains the rules, then the @code{next} statement is analogous to a @code{continue} statement: it skips to the end of the body of this implicit loop, and executes the increment (which reads another record). For example, if your @code{awk} program works only on records with four fields, and you don't want it to fail when given bad input, you might use this rule near the beginning of the program: @example @group NF != 4 @{ err = sprintf("%s:%d: skipped: NF != 4\n", FILENAME, FNR) print err > "/dev/stderr" next @} @end group @end example @noindent so that the following rules will not see the bad record. The error message is redirected to the standard error output stream, as error messages should be. @xref{Special Files, ,Special File Names in @code{gawk}}. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} According to the POSIX standard, the behavior is undefined if the @code{next} statement is used in a @code{BEGIN} or @code{END} rule. @code{gawk} will treat it as a syntax error. Although POSIX permits it, some other @code{awk} implementations don't allow the @code{next} statement inside function bodies (@pxref{User-defined, ,User-defined Functions}). Just as any other @code{next} statement, a @code{next} inside a function body reads the next record and starts processing it with the first rule in the program. If the @code{next} statement causes the end of the input to be reached, then the code in any @code{END} rules will be executed. @xref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}. @cindex @code{next}, inside a user-defined function @strong{Caution:} Some @code{awk} implementations generate a run-time error if you use the @code{next} statement inside a user-defined function (@pxref{User-defined, , User-defined Functions}). @code{gawk} does not have this problem. @node Nextfile Statement, Exit Statement, Next Statement, Statements @section The @code{nextfile} Statement @cindex @code{nextfile} statement @cindex differences between @code{gawk} and @code{awk} @code{gawk} provides the @code{nextfile} statement, which is similar to the @code{next} statement. However, instead of abandoning processing of the current record, the @code{nextfile} statement instructs @code{gawk} to stop processing the current data file. Upon execution of the @code{nextfile} statement, @code{FILENAME} is updated to the name of the next data file listed on the command line, @code{FNR} is reset to one, @code{ARGIND} is incremented, and processing starts over with the first rule in the progam. @xref{Built-in Variables}. If the @code{nextfile} statement causes the end of the input to be reached, then the code in any @code{END} rules will be executed. @xref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}. The @code{nextfile} statement is a @code{gawk} extension; it is not (currently) available in any other @code{awk} implementation. @xref{Nextfile Function, ,Implementing @code{nextfile} as a Function}, for a user-defined function you can use to simulate the @code{nextfile} statement. The @code{nextfile} statement would be useful if you have many data files to process, and you expect that you would not want to process every record in every file. Normally, in order to move on to the next data file, you would have to continue scanning the unwanted records. The @code{nextfile} statement accomplishes this much more efficiently. @cindex @code{next file} statement @strong{Caution:} Versions of @code{gawk} prior to 3.0 used two words (@samp{next file}) for the @code{nextfile} statement. This was changed in 3.0 to one word, since the treatment of @samp{file} was inconsistent. When it appeared after @code{next}, it was a keyword. Otherwise, it was a regular identifier. The old usage is still accepted. However, @code{gawk} will generate a warning message, and support for @code{next file} will eventually be discontinued in a future version of @code{gawk}. @node Exit Statement, , Nextfile Statement, Statements @section The @code{exit} Statement @cindex @code{exit} statement The @code{exit} statement causes @code{awk} to immediately stop executing the current rule and to stop processing input; any remaining input is ignored. It looks like this: @example exit @r{[}@var{return code}@r{]} @end example If an @code{exit} statement is executed from a @code{BEGIN} rule the program stops processing everything immediately. No input records are read. However, if an @code{END} rule is present, it is executed (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}). If @code{exit} is used as part of an @code{END} rule, it causes the program to stop immediately. An @code{exit} statement that is not part of a @code{BEGIN} or @code{END} rule stops the execution of any further automatic rules for the current record, skips reading any remaining input records, and executes the @code{END} rule if there is one. If you do not want the @code{END} rule to do its job in this case, you can set a variable to non-zero before the @code{exit} statement, and check that variable in the @code{END} rule. @xref{Assert Function, ,Assertions}, for an example that does this. @cindex dark corner If an argument is supplied to @code{exit}, its value is used as the exit status code for the @code{awk} process. If no argument is supplied, @code{exit} returns status zero (success). In the case where an argument is supplied to a first @code{exit} statement, and then @code{exit} is called a second time with no argument, the previously supplied exit value is used (d.c.). For example, let's say you've discovered an error condition you really don't know how to handle. Conventionally, programs report this by exiting with a non-zero status. Your @code{awk} program can do this using an @code{exit} statement with a non-zero argument. Here is an example: @example @group BEGIN @{ if (("date" | getline date_now) <= 0) @{ print "Can't get system date" > "/dev/stderr" exit 1 @} print "current date is", date_now close("date") @} @end group @end example @node Built-in Variables, Arrays, Statements, Top @chapter Built-in Variables @cindex built-in variables Most @code{awk} variables are available for you to use for your own purposes; they never change except when your program assigns values to them, and never affect anything except when your program examines them. However, a few variables in @code{awk} have special built-in meanings. Some of them @code{awk} examines automatically, so that they enable you to tell @code{awk} how to do certain things. Others are set automatically by @code{awk}, so that they carry information from the internal workings of @code{awk} to your program. This chapter documents all the built-in variables of @code{gawk}. Most of them are also documented in the chapters describing their areas of activity. @menu * User-modified:: Built-in variables that you change to control @code{awk}. * Auto-set:: Built-in variables where @code{awk} gives you information. * ARGC and ARGV:: Ways to use @code{ARGC} and @code{ARGV}. @end menu @node User-modified, Auto-set, Built-in Variables, Built-in Variables @section Built-in Variables that Control @code{awk} @cindex built-in variables, user modifiable This is an alphabetical list of the variables which you can change to control how @code{awk} does certain things. Those variables that are specific to @code{gawk} are marked with an asterisk, @samp{*}. @table @code @vindex CONVFMT @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} @item CONVFMT This string controls conversion of numbers to strings (@pxref{Conversion, ,Conversion of Strings and Numbers}). It works by being passed, in effect, as the first argument to the @code{sprintf} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). Its default value is @code{"%.6g"}. @code{CONVFMT} was introduced by the POSIX standard. @vindex FIELDWIDTHS @item FIELDWIDTHS * This is a space separated list of columns that tells @code{gawk} how to split input with fixed, columnar boundaries. It is an experimental feature. Assigning to @code{FIELDWIDTHS} overrides the use of @code{FS} for field splitting. @xref{Constant Size, ,Reading Fixed-width Data}, for more information. If @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}), then @code{FIELDWIDTHS} has no special meaning, and field splitting operations are done based exclusively on the value of @code{FS}. @vindex FS @item FS @code{FS} is the input field separator (@pxref{Field Separators, ,Specifying How Fields are Separated}). The value is a single-character string or a multi-character regular expression that matches the separations between fields in an input record. If the value is the null string (@code{""}), then each character in the record becomes a separate field. The default value is @w{@code{" "}}, a string consisting of a single space. As a special exception, this value means that any sequence of spaces, tabs, and/or newlines is a single separator.@footnote{In POSIX @code{awk}, newline does not count as whitespace.} It also causes spaces, tabs, and newlines at the beginning and end of a record to be ignored. You can set the value of @code{FS} on the command line using the @samp{-F} option: @example awk -F, '@var{program}' @var{input-files} @end example If @code{gawk} is using @code{FIELDWIDTHS} for field-splitting, assigning a value to @code{FS} will cause @code{gawk} to return to the normal, @code{FS}-based, field splitting. An easy way to do this is to simply say @samp{FS = FS}, perhaps with an explanatory comment. @vindex IGNORECASE @item IGNORECASE * If @code{IGNORECASE} is non-zero or non-null, then all string comparisons, and all regular expression matching are case-independent. Thus, regexp matching with @samp{~} and @samp{!~}, and the @code{gensub}, @code{gsub}, @code{index}, @code{match}, @code{split} and @code{sub} functions, record termination with @code{RS}, and field splitting with @code{FS} all ignore case when doing their particular regexp operations. The value of @code{IGNORECASE} does @emph{not} affect array subscripting. @xref{Case-sensitivity, ,Case-sensitivity in Matching}. If @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}), then @code{IGNORECASE} has no special meaning, and string and regexp operations are always case-sensitive. @vindex OFMT @item OFMT This string controls conversion of numbers to strings (@pxref{Conversion, ,Conversion of Strings and Numbers}) for printing with the @code{print} statement. It works by being passed, in effect, as the first argument to the @code{sprintf} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). Its default value is @code{"%.6g"}. Earlier versions of @code{awk} also used @code{OFMT} to specify the format for converting numbers to strings in general expressions; this is now done by @code{CONVFMT}. @vindex OFS @item OFS This is the output field separator (@pxref{Output Separators}). It is output between the fields output by a @code{print} statement. Its default value is @w{@code{" "}}, a string consisting of a single space. @vindex ORS @item ORS This is the output record separator. It is output at the end of every @code{print} statement. Its default value is @code{"\n"}. (@xref{Output Separators}.) @vindex RS @item RS This is @code{awk}'s input record separator. Its default value is a string containing a single newline character, which means that an input record consists of a single line of text. It can also be the null string, in which case records are separated by runs of blank lines, or a regexp, in which case records are separated by matches of the regexp in the input text. (@xref{Records, ,How Input is Split into Records}.) @vindex SUBSEP @item SUBSEP @code{SUBSEP} is the subscript separator. It has the default value of @code{"\034"}, and is used to separate the parts of the indices of a multi-dimensional array. Thus, the expression @code{@w{foo["A", "B"]}} really accesses @code{foo["A\034B"]} (@pxref{Multi-dimensional, ,Multi-dimensional Arrays}). @end table @node Auto-set, ARGC and ARGV, User-modified, Built-in Variables @section Built-in Variables that Convey Information @cindex built-in variables, convey information This is an alphabetical list of the variables that are set automatically by @code{awk} on certain occasions in order to provide information to your program. Those variables that are specific to @code{gawk} are marked with an asterisk, @samp{*}. @table @code @vindex ARGC @vindex ARGV @item ARGC @itemx ARGV The command-line arguments available to @code{awk} programs are stored in an array called @code{ARGV}. @code{ARGC} is the number of command-line arguments present. @xref{Other Arguments, ,Other Command Line Arguments}. Unlike most @code{awk} arrays, @code{ARGV} is indexed from zero to @code{ARGC} @minus{} 1. For example: @example @group $ awk 'BEGIN @{ > for (i = 0; i < ARGC; i++) > print ARGV[i] > @}' inventory-shipped BBS-list @print{} awk @print{} inventory-shipped @print{} BBS-list @end group @end example @noindent In this example, @code{ARGV[0]} contains @code{"awk"}, @code{ARGV[1]} contains @code{"inventory-shipped"}, and @code{ARGV[2]} contains @code{"BBS-list"}. The value of @code{ARGC} is three, one more than the index of the last element in @code{ARGV}, since the elements are numbered from zero. The names @code{ARGC} and @code{ARGV}, as well as the convention of indexing the array from zero to @code{ARGC} @minus{} 1, are derived from the C language's method of accessing command line arguments. @xref{ARGC and ARGV, , Using @code{ARGC} and @code{ARGV}}, for information about how @code{awk} uses these variables. @vindex ARGIND @item ARGIND * The index in @code{ARGV} of the current file being processed. Every time @code{gawk} opens a new data file for processing, it sets @code{ARGIND} to the index in @code{ARGV} of the file name. When @code{gawk} is processing the input files, it is always true that @samp{FILENAME == ARGV[ARGIND]}. This variable is useful in file processing; it allows you to tell how far along you are in the list of data files, and to distinguish between successive instances of the same filename on the command line. While you can change the value of @code{ARGIND} within your @code{awk} program, @code{gawk} will automatically set it to a new value when the next file is opened. This variable is a @code{gawk} extension. In other @code{awk} implementations, or if @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}), it is not special. @vindex ENVIRON @item ENVIRON An associative array that contains the values of the environment. The array indices are the environment variable names; the values are the values of the particular environment variables. For example, @code{ENVIRON["HOME"]} might be @file{/home/arnold}. Changing this array does not affect the environment passed on to any programs that @code{awk} may spawn via redirection or the @code{system} function. (In a future version of @code{gawk}, it may do so.) Some operating systems may not have environment variables. On such systems, the @code{ENVIRON} array is empty (except for @w{@code{ENVIRON["AWKPATH"]}}). @vindex ERRNO @item ERRNO * If a system error occurs either doing a redirection for @code{getline}, during a read for @code{getline}, or during a @code{close} operation, then @code{ERRNO} will contain a string describing the error. This variable is a @code{gawk} extension. In other @code{awk} implementations, or if @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}), it is not special. @cindex dark corner @vindex FILENAME @item FILENAME This is the name of the file that @code{awk} is currently reading. When no data files are listed on the command line, @code{awk} reads from the standard input, and @code{FILENAME} is set to @code{"-"}. @code{FILENAME} is changed each time a new file is read (@pxref{Reading Files, ,Reading Input Files}). Inside a @code{BEGIN} rule, the value of @code{FILENAME} is @code{""}, since there are no input files being processed yet.@footnote{Some early implementations of Unix @code{awk} initialized @code{FILENAME} to @code{"-"}, even if there were data files to be processed. This behavior was incorrect, and should not be relied upon in your programs.} (d.c.) @vindex FNR @item FNR @code{FNR} is the current record number in the current file. @code{FNR} is incremented each time a new record is read (@pxref{Getline, ,Explicit Input with @code{getline}}). It is reinitialized to zero each time a new input file is started. @vindex NF @item NF @code{NF} is the number of fields in the current input record. @code{NF} is set each time a new record is read, when a new field is created, or when @code{$0} changes (@pxref{Fields, ,Examining Fields}). @vindex NR @item NR This is the number of input records @code{awk} has processed since the beginning of the program's execution (@pxref{Records, ,How Input is Split into Records}). @code{NR} is set each time a new record is read. @vindex RLENGTH @item RLENGTH @code{RLENGTH} is the length of the substring matched by the @code{match} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). @code{RLENGTH} is set by invoking the @code{match} function. Its value is the length of the matched string, or @minus{}1 if no match was found. @vindex RSTART @item RSTART @code{RSTART} is the start-index in characters of the substring matched by the @code{match} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). @code{RSTART} is set by invoking the @code{match} function. Its value is the position of the string where the matched substring starts, or zero if no match was found. @vindex RT @item RT * @code{RT} is set each time a record is read. It contains the input text that matched the text denoted by @code{RS}, the record separator. This variable is a @code{gawk} extension. In other @code{awk} implementations, or if @code{gawk} is in compatibility mode (@pxref{Options, ,Command Line Options}), it is not special. @end table @cindex dark corner A side note about @code{NR} and @code{FNR}. @code{awk} simply increments both of these variables each time it reads a record, instead of setting them to the absolute value of the number of records read. This means that your program can change these variables, and their new values will be incremented for each record (d.c.). For example: @example @group $ echo '1 > 2 > 3 > 4' | awk 'NR == 2 @{ NR = 17 @} > @{ print NR @}' @print{} 1 @print{} 17 @print{} 18 @print{} 19 @end group @end example @noindent Before @code{FNR} was added to the @code{awk} language (@pxref{V7/SVR3.1, ,Major Changes between V7 and SVR3.1}), many @code{awk} programs used this feature to track the number of records in a file by resetting @code{NR} to zero when @code{FILENAME} changed. @node ARGC and ARGV, , Auto-set, Built-in Variables @section Using @code{ARGC} and @code{ARGV} In @ref{Auto-set, , Built-in Variables that Convey Information}, you saw this program describing the information contained in @code{ARGC} and @code{ARGV}: @example @group $ awk 'BEGIN @{ > for (i = 0; i < ARGC; i++) > print ARGV[i] > @}' inventory-shipped BBS-list @print{} awk @print{} inventory-shipped @print{} BBS-list @end group @end example @noindent In this example, @code{ARGV[0]} contains @code{"awk"}, @code{ARGV[1]} contains @code{"inventory-shipped"}, and @code{ARGV[2]} contains @code{"BBS-list"}. Notice that the @code{awk} program is not entered in @code{ARGV}. The other special command line options, with their arguments, are also not entered. This includes variable assignments done with the @samp{-v} option (@pxref{Options, ,Command Line Options}). Normal variable assignments on the command line @emph{are} treated as arguments, and do show up in the @code{ARGV} array. @example $ cat showargs.awk @print{} BEGIN @{ @print{} printf "A=%d, B=%d\n", A, B @print{} for (i = 0; i < ARGC; i++) @print{} printf "\tARGV[%d] = %s\n", i, ARGV[i] @print{} @} @print{} END @{ printf "A=%d, B=%d\n", A, B @} $ awk -v A=1 -f showargs.awk B=2 /dev/null @print{} A=1, B=0 @print{} ARGV[0] = awk @print{} ARGV[1] = B=2 @print{} ARGV[2] = /dev/null @print{} A=1, B=2 @end example Your program can alter @code{ARGC} and the elements of @code{ARGV}. Each time @code{awk} reaches the end of an input file, it uses the next element of @code{ARGV} as the name of the next input file. By storing a different string there, your program can change which files are read. You can use @code{"-"} to represent the standard input. By storing additional elements and incrementing @code{ARGC} you can cause additional files to be read. If you decrease the value of @code{ARGC}, that eliminates input files from the end of the list. By recording the old value of @code{ARGC} elsewhere, your program can treat the eliminated arguments as something other than file names. To eliminate a file from the middle of the list, store the null string (@code{""}) into @code{ARGV} in place of the file's name. As a special feature, @code{awk} ignores file names that have been replaced with the null string. You may also use the @code{delete} statement to remove elements from @code{ARGV} (@pxref{Delete, ,The @code{delete} Statement}). All of these actions are typically done from the @code{BEGIN} rule, before actual processing of the input begins. @xref{Split Program, ,Splitting a Large File Into Pieces}, and see @ref{Tee Program, ,Duplicating Output Into Multiple Files}, for an example of each way of removing elements from @code{ARGV}. The following fragment processes @code{ARGV} in order to examine, and then remove, command line options. @example @group BEGIN @{ for (i = 1; i < ARGC; i++) @{ if (ARGV[i] == "-v") verbose = 1 else if (ARGV[i] == "-d") debug = 1 @end group @group else if (ARGV[i] ~ /^-?/) @{ e = sprintf("%s: unrecognized option -- %c", ARGV[0], substr(ARGV[i], 1, ,1)) print e > "/dev/stderr" @} else break delete ARGV[i] @} @} @end group @end example To actually get the options into the @code{awk} program, you have to end the @code{awk} options with @samp{--}, and then supply your options, like so: @example awk -f myprog -- -v -d file1 file2 @dots{} @end example @cindex differences between @code{gawk} and @code{awk} This is not necessary in @code{gawk}: Unless @samp{--posix} has been specified, @code{gawk} silently puts any unrecognized options into @code{ARGV} for the @code{awk} program to deal with. As soon as it sees an unknown option, @code{gawk} stops looking for other options it might otherwise recognize. The above example with @code{gawk} would be: @example gawk -f myprog -d -v file1 file2 @dots{} @end example @noindent Since @samp{-d} is not a valid @code{gawk} option, the following @samp{-v} is passed on to the @code{awk} program. @node Arrays, Built-in, Built-in Variables, Top @chapter Arrays in @code{awk} An @dfn{array} is a table of values, called @dfn{elements}. The elements of an array are distinguished by their indices. @dfn{Indices} may be either numbers or strings. @code{awk} maintains a single set of names that may be used for naming variables, arrays and functions (@pxref{User-defined, ,User-defined Functions}). Thus, you cannot have a variable and an array with the same name in the same @code{awk} program. @menu * Array Intro:: Introduction to Arrays * Reference to Elements:: How to examine one element of an array. * Assigning Elements:: How to change an element of an array. * Array Example:: Basic Example of an Array * Scanning an Array:: A variation of the @code{for} statement. It loops through the indices of an array's existing elements. * Delete:: The @code{delete} statement removes an element from an array. * Numeric Array Subscripts:: How to use numbers as subscripts in @code{awk}. * Uninitialized Subscripts:: Using Uninitialized variables as subscripts. * Multi-dimensional:: Emulating multi-dimensional arrays in @code{awk}. * Multi-scanning:: Scanning multi-dimensional arrays. @end menu @node Array Intro, Reference to Elements, Arrays, Arrays @section Introduction to Arrays @cindex arrays The @code{awk} language provides one-dimensional @dfn{arrays} for storing groups of related strings or numbers. Every @code{awk} array must have a name. Array names have the same syntax as variable names; any valid variable name would also be a valid array name. But you cannot use one name in both ways (as an array and as a variable) in one @code{awk} program. Arrays in @code{awk} superficially resemble arrays in other programming languages; but there are fundamental differences. In @code{awk}, you don't need to specify the size of an array before you start to use it. Additionally, any number or string in @code{awk} may be used as an array index, not just consecutive integers. In most other languages, you have to @dfn{declare} an array and specify how many elements or components it contains. In such languages, the declaration causes a contiguous block of memory to be allocated for that many elements. An index in the array usually must be a positive integer; for example, the index zero specifies the first element in the array, which is actually stored at the beginning of the block of memory. Index one specifies the second element, which is stored in memory right after the first element, and so on. It is impossible to add more elements to the array, because it has room for only as many elements as you declared. (Some languages allow arbitrary starting and ending indices, e.g., @samp{15 .. 27}, but the size of the array is still fixed when the array is declared.) A contiguous array of four elements might look like this, conceptually, if the element values are eight, @code{"foo"}, @code{""} and 30: @iftex @c from Karl Berry, much thanks for the help. @tex \bigskip % space above the table (about 1 linespace) \offinterlineskip \newdimen\width \width = 1.5cm \newdimen\hwidth \hwidth = 4\width \advance\hwidth by 2pt % 5 * 0.4pt \centerline{\vbox{ \halign{\strut\hfil\ignorespaces#&&\vrule#&\hbox to\width{\hfil#\unskip\hfil}\cr \noalign{\hrule width\hwidth} &&{\tt 8} &&{\tt "foo"} &&{\tt ""} &&{\tt 30} &&\quad value\cr \noalign{\hrule width\hwidth} \noalign{\smallskip} &\omit&0&\omit &1 &\omit&2 &\omit&3 &\omit&\quad index\cr } }} @end tex @end iftex @ifinfo @example +---------+---------+--------+---------+ | 8 | "foo" | "" | 30 | @r{value} +---------+---------+--------+---------+ 0 1 2 3 @r{index} @end example @end ifinfo @noindent Only the values are stored; the indices are implicit from the order of the values. Eight is the value at index zero, because eight appears in the position with zero elements before it. @cindex arrays, definition of @cindex associative arrays @cindex arrays, associative Arrays in @code{awk} are different: they are @dfn{associative}. This means that each array is a collection of pairs: an index, and its corresponding array element value: @example @r{Element} 4 @r{Value} 30 @r{Element} 2 @r{Value} "foo" @r{Element} 1 @r{Value} 8 @r{Element} 3 @r{Value} "" @end example @noindent We have shown the pairs in jumbled order because their order is irrelevant. One advantage of associative arrays is that new pairs can be added at any time. For example, suppose we add to the above array a tenth element whose value is @w{@code{"number ten"}}. The result is this: @example @r{Element} 10 @r{Value} "number ten" @r{Element} 4 @r{Value} 30 @r{Element} 2 @r{Value} "foo" @r{Element} 1 @r{Value} 8 @r{Element} 3 @r{Value} "" @end example @noindent @cindex sparse arrays @cindex arrays, sparse Now the array is @dfn{sparse}, which just means some indices are missing: it has elements 1--4 and 10, but doesn't have elements 5, 6, 7, 8, or 9. @c ok, I should spell out the above, but ... Another consequence of associative arrays is that the indices don't have to be positive integers. Any number, or even a string, can be an index. For example, here is an array which translates words from English into French: @example @r{Element} "dog" @r{Value} "chien" @r{Element} "cat" @r{Value} "chat" @r{Element} "one" @r{Value} "un" @r{Element} 1 @r{Value} "un" @end example @noindent Here we decided to translate the number one in both spelled-out and numeric form---thus illustrating that a single array can have both numbers and strings as indices. (In fact, array subscripts are always strings; this is discussed in more detail in @ref{Numeric Array Subscripts, ,Using Numbers to Subscript Arrays}.) @cindex Array subscripts and @code{IGNORECASE} @cindex @code{IGNORECASE} and array subscripts @vindex IGNORECASE The value of @code{IGNORECASE} has no effect upon array subscripting. You must use the exact same string value to retrieve an array element as you used to store it. When @code{awk} creates an array for you, e.g., with the @code{split} built-in function, that array's indices are consecutive integers starting at one. (@xref{String Functions, ,Built-in Functions for String Manipulation}.) @node Reference to Elements, Assigning Elements, Array Intro, Arrays @section Referring to an Array Element @cindex array reference @cindex element of array @cindex reference to array The principal way of using an array is to refer to one of its elements. An array reference is an expression which looks like this: @example @var{array}[@var{index}] @end example @noindent Here, @var{array} is the name of an array. The expression @var{index} is the index of the element of the array that you want. The value of the array reference is the current value of that array element. For example, @code{foo[4.3]} is an expression for the element of array @code{foo} at index @samp{4.3}. If you refer to an array element that has no recorded value, the value of the reference is @code{""}, the null string. This includes elements to which you have not assigned any value, and elements that have been deleted (@pxref{Delete, ,The @code{delete} Statement}). Such a reference automatically creates that array element, with the null string as its value. (In some cases, this is unfortunate, because it might waste memory inside @code{awk}.) @cindex arrays, presence of elements @cindex arrays, the @code{in} operator You can find out if an element exists in an array at a certain index with the expression: @example @var{index} in @var{array} @end example @noindent This expression tests whether or not the particular index exists, without the side effect of creating that element if it is not present. The expression has the value one (true) if @code{@var{array}[@var{index}]} exists, and zero (false) if it does not exist. For example, to test whether the array @code{frequencies} contains the index @samp{2}, you could write this statement: @example if (2 in frequencies) print "Subscript 2 is present." @end example Note that this is @emph{not} a test of whether or not the array @code{frequencies} contains an element whose @emph{value} is two. (There is no way to do that except to scan all the elements.) Also, this @emph{does not} create @code{frequencies[2]}, while the following (incorrect) alternative would do so: @example if (frequencies[2] != "") print "Subscript 2 is present." @end example @node Assigning Elements, Array Example, Reference to Elements, Arrays @section Assigning Array Elements @cindex array assignment @cindex element assignment Array elements are lvalues: they can be assigned values just like @code{awk} variables: @example @var{array}[@var{subscript}] = @var{value} @end example @noindent Here @var{array} is the name of your array. The expression @var{subscript} is the index of the element of the array that you want to assign a value. The expression @var{value} is the value you are assigning to that element of the array. @node Array Example, Scanning an Array, Assigning Elements, Arrays @section Basic Array Example The following program takes a list of lines, each beginning with a line number, and prints them out in order of line number. The line numbers are not in order, however, when they are first read: they are scrambled. This program sorts the lines by making an array using the line numbers as subscripts. It then prints out the lines in sorted order of their numbers. It is a very simple program, and gets confused if it encounters repeated numbers, gaps, or lines that don't begin with a number. @example @c file eg/misc/arraymax.awk @{ if ($1 > max) max = $1 arr[$1] = $0 @} END @{ for (x = 1; x <= max; x++) print arr[x] @} @c endfile @end example The first rule keeps track of the largest line number seen so far; it also stores each line into the array @code{arr}, at an index that is the line's number. The second rule runs after all the input has been read, to print out all the lines. When this program is run with the following input: @example @group @c file eg/misc/arraymax.data 5 I am the Five man 2 Who are you? The new number two! 4 . . . And four on the floor 1 Who is number one? 3 I three you. @c endfile @end group @end example @noindent its output is this: @example 1 Who is number one? 2 Who are you? The new number two! 3 I three you. 4 . . . And four on the floor 5 I am the Five man @end example If a line number is repeated, the last line with a given number overrides the others. Gaps in the line numbers can be handled with an easy improvement to the program's @code{END} rule: @example END @{ for (x = 1; x <= max; x++) if (x in arr) print arr[x] @} @end example @node Scanning an Array, Delete, Array Example, Arrays @section Scanning All Elements of an Array @cindex @code{for (x in @dots{})} @cindex arrays, special @code{for} statement @cindex scanning an array In programs that use arrays, you often need a loop that executes once for each element of an array. In other languages, where arrays are contiguous and indices are limited to positive integers, this is easy: you can find all the valid indices by counting from the lowest index up to the highest. This technique won't do the job in @code{awk}, since any number or string can be an array index. So @code{awk} has a special kind of @code{for} statement for scanning an array: @example for (@var{var} in @var{array}) @var{body} @end example @noindent This loop executes @var{body} once for each index in @var{array} that your program has previously used, with the variable @var{var} set to that index. Here is a program that uses this form of the @code{for} statement. The first rule scans the input records and notes which words appear (at least once) in the input, by storing a one into the array @code{used} with the word as index. The second rule scans the elements of @code{used} to find all the distinct words that appear in the input. It prints each word that is more than 10 characters long, and also prints the number of such words. @xref{String Functions, ,Built-in Functions for String Manipulation}, for more information on the built-in function @code{length}. @example # Record a 1 for each word that is used at least once. @{ for (i = 1; i <= NF; i++) used[$i] = 1 @} # Find number of distinct words more than 10 characters long. END @{ for (x in used) if (length(x) > 10) @{ ++num_long_words print x @} print num_long_words, "words longer than 10 characters" @} @end example @noindent @xref{Word Sorting, ,Generating Word Usage Counts}, for a more detailed example of this type. The order in which elements of the array are accessed by this statement is determined by the internal arrangement of the array elements within @code{awk} and cannot be controlled or changed. This can lead to problems if new elements are added to @var{array} by statements in the loop body; you cannot predict whether or not the @code{for} loop will reach them. Similarly, changing @var{var} inside the loop may produce strange results. It is best to avoid such things. @node Delete, Numeric Array Subscripts, Scanning an Array, Arrays @section The @code{delete} Statement @cindex @code{delete} statement @cindex deleting elements of arrays @cindex removing elements of arrays @cindex arrays, deleting an element You can remove an individual element of an array using the @code{delete} statement: @example delete @var{array}[@var{index}] @end example Once you have deleted an array element, you can no longer obtain any value the element once had. It is as if you had never referred to it and had never given it any value. Here is an example of deleting elements in an array: @example for (i in frequencies) delete frequencies[i] @end example @noindent This example removes all the elements from the array @code{frequencies}. If you delete an element, a subsequent @code{for} statement to scan the array will not report that element, and the @code{in} operator to check for the presence of that element will return zero (i.e.@: false): @example delete foo[4] if (4 in foo) print "This will never be printed" @end example It is important to note that deleting an element is @emph{not} the same as assigning it a null value (the empty string, @code{""}). @example foo[4] = "" if (4 in foo) print "This is printed, even though foo[4] is empty" @end example It is not an error to delete an element that does not exist. @cindex arrays, deleting entire contents @cindex deleting entire arrays @cindex differences between @code{gawk} and @code{awk} You can delete all the elements of an array with a single statement, by leaving off the subscript in the @code{delete} statement. @example delete @var{array} @end example This ability is a @code{gawk} extension; it is not available in compatibility mode (@pxref{Options, ,Command Line Options}). Using this version of the @code{delete} statement is about three times more efficient than the equivalent loop that deletes each element one at a time. @cindex portability issues The following statement provides a portable, but non-obvious way to clear out an array. @cindex Brennan, Michael @example @group # thanks to Michael Brennan for pointing this out split("", array) @end group @end example The @code{split} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}) clears out the target array first. This call asks it to split apart the null string. Since there is no data to split out, the function simply clears the array and then returns. @strong{Caution:} Deleting an array does not change its type; you cannot delete an array and then use the array's name as a scalar. For example, this will not work: @example a[1] = 3; delete a; a = 3 @end example @node Numeric Array Subscripts, Uninitialized Subscripts, Delete, Arrays @section Using Numbers to Subscript Arrays An important aspect of arrays to remember is that @emph{array subscripts are always strings}. If you use a numeric value as a subscript, it will be converted to a string value before it is used for subscripting (@pxref{Conversion, ,Conversion of Strings and Numbers}). @cindex conversions, during subscripting @cindex numbers, used as subscripts @vindex CONVFMT This means that the value of the built-in variable @code{CONVFMT} can potentially affect how your program accesses elements of an array. For example: @example xyz = 12.153 data[xyz] = 1 CONVFMT = "%2.2f" @group if (xyz in data) printf "%s is in data\n", xyz else printf "%s is not in data\n", xyz @end group @end example @noindent This prints @samp{12.15 is not in data}. The first statement gives @code{xyz} a numeric value. Assigning to @code{data[xyz]} subscripts @code{data} with the string value @code{"12.153"} (using the default conversion value of @code{CONVFMT}, @code{"%.6g"}), and assigns one to @code{data["12.153"]}. The program then changes the value of @code{CONVFMT}. The test @samp{(xyz in data)} generates a new string value from @code{xyz}, this time @code{"12.15"}, since the value of @code{CONVFMT} only allows two significant digits. This test fails, since @code{"12.15"} is a different string from @code{"12.153"}. According to the rules for conversions (@pxref{Conversion, ,Conversion of Strings and Numbers}), integer values are always converted to strings as integers, no matter what the value of @code{CONVFMT} may happen to be. So the usual case of: @example for (i = 1; i <= maxsub; i++) @i{do something with} array[i] @end example @noindent will work, no matter what the value of @code{CONVFMT}. Like many things in @code{awk}, the majority of the time things work as you would expect them to work. But it is useful to have a precise knowledge of the actual rules, since sometimes they can have a subtle effect on your programs. @node Uninitialized Subscripts, Multi-dimensional, Numeric Array Subscripts, Arrays @section Using Uninitialized Variables as Subscripts @cindex uninitialized variables, as array subscripts @cindex array subscripts, uninitialized variables Suppose you want to print your input data in reverse order. A reasonable attempt at a program to do so (with some test data) might look like this: @example @group $ echo 'line 1 > line 2 > line 3' | awk '@{ l[lines] = $0; ++lines @} > END @{ > for (i = lines-1; i >= 0; --i) > print l[i] > @}' @print{} line 3 @print{} line 2 @end group @end example Unfortunately, the very first line of input data did not come out in the output! At first glance, this program should have worked. The variable @code{lines} is uninitialized, and uninitialized variables have the numeric value zero. So, the value of @code{l[0]} should have been printed. The issue here is that subscripts for @code{awk} arrays are @strong{always} strings. And uninitialized variables, when used as strings, have the value @code{""}, not zero. Thus, @samp{line 1} ended up stored in @code{l[""]}. The following version of the program works correctly: @example @{ l[lines++] = $0 @} END @{ for (i = lines - 1; i >= 0; --i) print l[i] @} @end example Here, the @samp{++} forces @code{lines} to be numeric, thus making the ``old value'' numeric zero, which is then converted to @code{"0"} as the array subscript. @cindex null string, as array subscript @cindex dark corner As we have just seen, even though it is somewhat unusual, the null string (@code{""}) is a valid array subscript (d.c.). If @samp{--lint} is provided on the command line (@pxref{Options, ,Command Line Options}), @code{gawk} will warn about the use of the null string as a subscript. @node Multi-dimensional, Multi-scanning, Uninitialized Subscripts, Arrays @section Multi-dimensional Arrays @cindex subscripts in arrays @cindex arrays, multi-dimensional subscripts @cindex multi-dimensional subscripts A multi-dimensional array is an array in which an element is identified by a sequence of indices, instead of a single index. For example, a two-dimensional array requires two indices. The usual way (in most languages, including @code{awk}) to refer to an element of a two-dimensional array named @code{grid} is with @code{grid[@var{x},@var{y}]}. @vindex SUBSEP Multi-dimensional arrays are supported in @code{awk} through concatenation of indices into one string. What happens is that @code{awk} converts the indices into strings (@pxref{Conversion, ,Conversion of Strings and Numbers}) and concatenates them together, with a separator between them. This creates a single string that describes the values of the separate indices. The combined string is used as a single index into an ordinary, one-dimensional array. The separator used is the value of the built-in variable @code{SUBSEP}. For example, suppose we evaluate the expression @samp{foo[5,12] = "value"} when the value of @code{SUBSEP} is @code{"@@"}. The numbers five and 12 are converted to strings and concatenated with an @samp{@@} between them, yielding @code{"5@@12"}; thus, the array element @code{foo["5@@12"]} is set to @code{"value"}. Once the element's value is stored, @code{awk} has no record of whether it was stored with a single index or a sequence of indices. The two expressions @samp{foo[5,12]} and @w{@samp{foo[5 SUBSEP 12]}} are always equivalent. The default value of @code{SUBSEP} is the string @code{"\034"}, which contains a non-printing character that is unlikely to appear in an @code{awk} program or in most input data. The usefulness of choosing an unlikely character comes from the fact that index values that contain a string matching @code{SUBSEP} lead to combined strings that are ambiguous. Suppose that @code{SUBSEP} were @code{"@@"}; then @w{@samp{foo["a@@b", "c"]}} and @w{@samp{foo["a", "b@@c"]}} would be indistinguishable because both would actually be stored as @samp{foo["a@@b@@c"]}. You can test whether a particular index-sequence exists in a ``multi-dimensional'' array with the same operator @samp{in} used for single dimensional arrays. Instead of a single index as the left-hand operand, write the whole sequence of indices, separated by commas, in parentheses: @example (@var{subscript1}, @var{subscript2}, @dots{}) in @var{array} @end example The following example treats its input as a two-dimensional array of fields; it rotates this array 90 degrees clockwise and prints the result. It assumes that all lines have the same number of elements. @example @group awk '@{ if (max_nf < NF) max_nf = NF max_nr = NR for (x = 1; x <= NF; x++) vector[x, NR] = $x @} @end group @group END @{ for (x = 1; x <= max_nf; x++) @{ for (y = max_nr; y >= 1; --y) printf("%s ", vector[x, y]) printf("\n") @} @}' @end group @end example @noindent When given the input: @example @group 1 2 3 4 5 6 2 3 4 5 6 1 3 4 5 6 1 2 4 5 6 1 2 3 @end group @end example @noindent it produces: @example @group 4 3 2 1 5 4 3 2 6 5 4 3 1 6 5 4 2 1 6 5 3 2 1 6 @end group @end example @node Multi-scanning, , Multi-dimensional, Arrays @section Scanning Multi-dimensional Arrays There is no special @code{for} statement for scanning a ``multi-dimensional'' array; there cannot be one, because in truth there are no multi-dimensional arrays or elements; there is only a multi-dimensional @emph{way of accessing} an array. However, if your program has an array that is always accessed as multi-dimensional, you can get the effect of scanning it by combining the scanning @code{for} statement (@pxref{Scanning an Array, ,Scanning All Elements of an Array}) with the @code{split} built-in function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). It works like this: @example for (combined in array) @{ split(combined, separate, SUBSEP) @dots{} @} @end example @noindent This sets @code{combined} to each concatenated, combined index in the array, and splits it into the individual indices by breaking it apart where the value of @code{SUBSEP} appears. The split-out indices become the elements of the array @code{separate}. Thus, suppose you have previously stored a value in @code{array[1, "foo"]}; then an element with index @code{"1\034foo"} exists in @code{array}. (Recall that the default value of @code{SUBSEP} is the character with code 034.) Sooner or later the @code{for} statement will find that index and do an iteration with @code{combined} set to @code{"1\034foo"}. Then the @code{split} function is called as follows: @example split("1\034foo", separate, "\034") @end example @noindent The result of this is to set @code{separate[1]} to @code{"1"} and @code{separate[2]} to @code{"foo"}. Presto, the original sequence of separate indices has been recovered. @node Built-in, User-defined, Arrays, Top @chapter Built-in Functions @c 2e: USE TEXINFO-2 FUNCTION DEFINITION STUFF!!!!!!!!!!!!! @cindex built-in functions @dfn{Built-in} functions are functions that are always available for your @code{awk} program to call. This chapter defines all the built-in functions in @code{awk}; some of them are mentioned in other sections, but they are summarized here for your convenience. (You can also define new functions yourself. @xref{User-defined, ,User-defined Functions}.) @menu * Calling Built-in:: How to call built-in functions. * Numeric Functions:: Functions that work with numbers, including @code{int}, @code{sin} and @code{rand}. * String Functions:: Functions for string manipulation, such as @code{split}, @code{match}, and @code{sprintf}. * I/O Functions:: Functions for files and shell commands. * Time Functions:: Functions for dealing with time stamps. @end menu @node Calling Built-in, Numeric Functions, Built-in, Built-in @section Calling Built-in Functions To call a built-in function, write the name of the function followed by arguments in parentheses. For example, @samp{atan2(y + z, 1)} is a call to the function @code{atan2}, with two arguments. Whitespace is ignored between the built-in function name and the open-parenthesis, but we recommend that you avoid using whitespace there. User-defined functions do not permit whitespace in this way, and you will find it easier to avoid mistakes by following a simple convention which always works: no whitespace after a function name. @cindex differences between @code{gawk} and @code{awk} Each built-in function accepts a certain number of arguments. In some cases, arguments can be omitted. The defaults for omitted arguments vary from function to function and are described under the individual functions. In some @code{awk} implementations, extra arguments given to built-in functions are ignored. However, in @code{gawk}, it is a fatal error to give extra arguments to a built-in function. When a function is called, expressions that create the function's actual parameters are evaluated completely before the function call is performed. For example, in the code fragment: @example i = 4 j = sqrt(i++) @end example @noindent the variable @code{i} is set to five before @code{sqrt} is called with a value of four for its actual parameter. @cindex evaluation, order of @cindex order of evaluation The order of evaluation of the expressions used for the function's parameters is undefined. Thus, you should not write programs that assume that parameters are evaluated from left to right or from right to left. For example, @example i = 5 j = atan2(i++, i *= 2) @end example If the order of evaluation is left to right, then @code{i} first becomes six, and then 12, and @code{atan2} is called with the two arguments six and 12. But if the order of evaluation is right to left, @code{i} first becomes 10, and then 11, and @code{atan2} is called with the two arguments 11 and 10. @node Numeric Functions, String Functions, Calling Built-in, Built-in @section Numeric Built-in Functions Here is a full list of built-in functions that work with numbers. Optional parameters are enclosed in square brackets (``['' and ``]''). @table @code @item int(@var{x}) @findex int This produces the nearest integer to @var{x}, located between @var{x} and zero, truncated toward zero. For example, @code{int(3)} is three, @code{int(3.9)} is three, @code{int(-3.9)} is @minus{}3, and @code{int(-3)} is @minus{}3 as well. @item sqrt(@var{x}) @findex sqrt This gives you the positive square root of @var{x}. It reports an error if @var{x} is negative. Thus, @code{sqrt(4)} is two. @item exp(@var{x}) @findex exp This gives you the exponential of @var{x} (@code{e ^ @var{x}}), or reports an error if @var{x} is out of range. The range of values @var{x} can have depends on your machine's floating point representation. @item log(@var{x}) @findex log This gives you the natural logarithm of @var{x}, if @var{x} is positive; otherwise, it reports an error. @item sin(@var{x}) @findex sin This gives you the sine of @var{x}, with @var{x} in radians. @item cos(@var{x}) @findex cos This gives you the cosine of @var{x}, with @var{x} in radians. @item atan2(@var{y}, @var{x}) @findex atan2 This gives you the arctangent of @code{@var{y} / @var{x}} in radians. @item rand() @findex rand This gives you a random number. The values of @code{rand} are uniformly-distributed between zero and one. The value is never zero and never one. Often you want random integers instead. Here is a user-defined function you can use to obtain a random non-negative integer less than @var{n}: @example function randint(n) @{ return int(n * rand()) @} @end example @noindent The multiplication produces a random real number greater than zero and less than @code{n}. We then make it an integer (using @code{int}) between zero and @code{n} @minus{} 1, inclusive. Here is an example where a similar function is used to produce random integers between one and @var{n}. This program prints a new random number for each input record. @example @group awk ' # Function to roll a simulated die. function roll(n) @{ return 1 + int(rand() * n) @} @end group @group # Roll 3 six-sided dice and # print total number of points. @{ printf("%d points\n", roll(6)+roll(6)+roll(6)) @}' @end group @end example @cindex seed for random numbers @cindex random numbers, seed of @comment MAWK uses a different seed each time. @strong{Caution:} In most @code{awk} implementations, including @code{gawk}, @code{rand} starts generating numbers from the same starting number, or @dfn{seed}, each time you run @code{awk}. Thus, a program will generate the same results each time you run it. The numbers are random within one @code{awk} run, but predictable from run to run. This is convenient for debugging, but if you want a program to do different things each time it is used, you must change the seed to a value that will be different in each run. To do this, use @code{srand}. @item srand(@r{[}@var{x}@r{]}) @findex srand The function @code{srand} sets the starting point, or seed, for generating random numbers to the value @var{x}. Each seed value leads to a particular sequence of random numbers.@footnote{Computer generated random numbers really are not truly random. They are technically known as ``pseudo-random.'' This means that while the numbers in a sequence appear to be random, you can in fact generate the same sequence of random numbers over and over again.} Thus, if you set the seed to the same value a second time, you will get the same sequence of random numbers again. If you omit the argument @var{x}, as in @code{srand()}, then the current date and time of day are used for a seed. This is the way to get random numbers that are truly unpredictable. The return value of @code{srand} is the previous seed. This makes it easy to keep track of the seeds for use in consistently reproducing sequences of random numbers. @end table @node String Functions, I/O Functions, Numeric Functions, Built-in @section Built-in Functions for String Manipulation The functions in this section look at or change the text of one or more strings. Optional parameters are enclosed in square brackets (``['' and ``]''). @table @code @item index(@var{in}, @var{find}) @findex index This searches the string @var{in} for the first occurrence of the string @var{find}, and returns the position in characters where that occurrence begins in the string @var{in}. For example: @example $ awk 'BEGIN @{ print index("peanut", "an") @}' @print{} 3 @end example @noindent If @var{find} is not found, @code{index} returns zero. (Remember that string indices in @code{awk} start at one.) @item length(@r{[}@var{string}@r{]}) @findex length This gives you the number of characters in @var{string}. If @var{string} is a number, the length of the digit string representing that number is returned. For example, @code{length("abcde")} is five. By contrast, @code{length(15 * 35)} works out to three. How? Well, 15 * 35 = 525, and 525 is then converted to the string @code{"525"}, which has three characters. If no argument is supplied, @code{length} returns the length of @code{$0}. @cindex historical features @cindex portability issues @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} In older versions of @code{awk}, you could call the @code{length} function without any parentheses. Doing so is marked as ``deprecated'' in the POSIX standard. This means that while you can do this in your programs, it is a feature that can eventually be removed from a future version of the standard. Therefore, for maximal portability of your @code{awk} programs, you should always supply the parentheses. @item match(@var{string}, @var{regexp}) @findex match The @code{match} function searches the string, @var{string}, for the longest, leftmost substring matched by the regular expression, @var{regexp}. It returns the character position, or @dfn{index}, of where that substring begins (one, if it starts at the beginning of @var{string}). If no match is found, it returns zero. @vindex RSTART @vindex RLENGTH The @code{match} function sets the built-in variable @code{RSTART} to the index. It also sets the built-in variable @code{RLENGTH} to the length in characters of the matched substring. If no match is found, @code{RSTART} is set to zero, and @code{RLENGTH} to @minus{}1. For example: @example @group @c file eg/misc/findpat.sh awk '@{ if ($1 == "FIND") regex = $2 else @{ where = match($0, regex) if (where != 0) print "Match of", regex, "found at", \ where, "in", $0 @} @}' @c endfile @end group @end example @noindent This program looks for lines that match the regular expression stored in the variable @code{regex}. This regular expression can be changed. If the first word on a line is @samp{FIND}, @code{regex} is changed to be the second word on that line. Therefore, given: @example @c file eg/misc/findpat.data FIND ru+n My program runs but not very quickly FIND Melvin JF+KM This line is property of Reality Engineering Co. Melvin was here. @c endfile @end example @noindent @code{awk} prints: @example Match of ru+n found at 12 in My program runs Match of Melvin found at 1 in Melvin was here. @end example @item split(@var{string}, @var{array} @r{[}, @var{fieldsep}@r{]}) @findex split This divides @var{string} into pieces separated by @var{fieldsep}, and stores the pieces in @var{array}. The first piece is stored in @code{@var{array}[1]}, the second piece in @code{@var{array}[2]}, and so forth. The string value of the third argument, @var{fieldsep}, is a regexp describing where to split @var{string} (much as @code{FS} can be a regexp describing where to split input records). If the @var{fieldsep} is omitted, the value of @code{FS} is used. @code{split} returns the number of elements created. The @code{split} function splits strings into pieces in a manner similar to the way input lines are split into fields. For example: @example split("cul-de-sac", a, "-") @end example @noindent splits the string @samp{cul-de-sac} into three fields using @samp{-} as the separator. It sets the contents of the array @code{a} as follows: @example a[1] = "cul" a[2] = "de" a[3] = "sac" @end example @noindent The value returned by this call to @code{split} is three. As with input field-splitting, when the value of @var{fieldsep} is @w{@code{" "}}, leading and trailing whitespace is ignored, and the elements are separated by runs of whitespace. @cindex differences between @code{gawk} and @code{awk} Also as with input field-splitting, if @var{fieldsep} is the null string, each individual character in the string is split into its own array element. (This is a @code{gawk}-specific extension.) @cindex dark corner Recent implementations of @code{awk}, including @code{gawk}, allow the third argument to be a regexp constant (@code{/abc/}), as well as a string (d.c.). The POSIX standard allows this as well. Before splitting the string, @code{split} deletes any previously existing elements in the array @var{array} (d.c.). If @var{string} does not match @var{fieldsep} at all, @var{array} will have one element. The value of that element will be the original @var{string}. @item sprintf(@var{format}, @var{expression1},@dots{}) @findex sprintf This returns (without printing) the string that @code{printf} would have printed out with the same arguments (@pxref{Printf, ,Using @code{printf} Statements for Fancier Printing}). For example: @example sprintf("pi = %.2f (approx.)", 22/7) @end example @noindent returns the string @w{@code{"pi = 3.14 (approx.)"}}. @ignore 2e: For sub, gsub, and gensub, either here or in the "how much matches" section, we need some explanation that it is possible to match the null string when using closures like *. E.g., $ echo abc | awk '{ gsub(/m*/, "X"); print }' @print{} XaXbXcX Although this makes a certain amount of sense, it can be very suprising. @end ignore @item sub(@var{regexp}, @var{replacement} @r{[}, @var{target}@r{]}) @findex sub The @code{sub} function alters the value of @var{target}. It searches this value, which is treated as a string, for the leftmost longest substring matched by the regular expression, @var{regexp}, extending this match as far as possible. Then the entire string is changed by replacing the matched text with @var{replacement}. The modified string becomes the new value of @var{target}. This function is peculiar because @var{target} is not simply used to compute a value, and not just any expression will do: it must be a variable, field or array element, so that @code{sub} can store a modified value there. If this argument is omitted, then the default is to use and alter @code{$0}. For example: @example str = "water, water, everywhere" sub(/at/, "ith", str) @end example @noindent sets @code{str} to @w{@code{"wither, water, everywhere"}}, by replacing the leftmost, longest occurrence of @samp{at} with @samp{ith}. The @code{sub} function returns the number of substitutions made (either one or zero). If the special character @samp{&} appears in @var{replacement}, it stands for the precise substring that was matched by @var{regexp}. (If the regexp can match more than one string, then this precise substring may vary.) For example: @example awk '@{ sub(/candidate/, "& and his wife"); print @}' @end example @noindent changes the first occurrence of @samp{candidate} to @samp{candidate and his wife} on each input line. Here is another example: @example awk 'BEGIN @{ str = "daabaaa" sub(/a*/, "c&c", str) print str @}' @print{} dcaacbaaa @end example @noindent This shows how @samp{&} can represent a non-constant string, and also illustrates the ``leftmost, longest'' rule in regexp matching (@pxref{Leftmost Longest, ,How Much Text Matches?}). The effect of this special character (@samp{&}) can be turned off by putting a backslash before it in the string. As usual, to insert one backslash in the string, you must write two backslashes. Therefore, write @samp{\\&} in a string constant to include a literal @samp{&} in the replacement. For example, here is how to replace the first @samp{|} on each line with an @samp{&}: @example awk '@{ sub(/\|/, "\\&"); print @}' @end example @cindex @code{sub}, third argument of @cindex @code{gsub}, third argument of @strong{Note:} As mentioned above, the third argument to @code{sub} must be a variable, field or array reference. Some versions of @code{awk} allow the third argument to be an expression which is not an lvalue. In such a case, @code{sub} would still search for the pattern and return zero or one, but the result of the substitution (if any) would be thrown away because there is no place to put it. Such versions of @code{awk} accept expressions like this: @example sub(/USA/, "United States", "the USA and Canada") @end example @noindent For historical compatibility, @code{gawk} will accept erroneous code, such as in the above example. However, using any other non-changeable object as the third parameter will cause a fatal error, and your program will not run. Finally, if the @var{regexp} is not a regexp constant, it is converted into a string and then the value of that string is treated as the regexp to match. @item gsub(@var{regexp}, @var{replacement} @r{[}, @var{target}@r{]}) @findex gsub This is similar to the @code{sub} function, except @code{gsub} replaces @emph{all} of the longest, leftmost, @emph{non-overlapping} matching substrings it can find. The @samp{g} in @code{gsub} stands for ``global,'' which means replace everywhere. For example: @example awk '@{ gsub(/Britain/, "United Kingdom"); print @}' @end example @noindent replaces all occurrences of the string @samp{Britain} with @samp{United Kingdom} for all input records. The @code{gsub} function returns the number of substitutions made. If the variable to be searched and altered, @var{target}, is omitted, then the entire input record, @code{$0}, is used. As in @code{sub}, the characters @samp{&} and @samp{\} are special, and the third argument must be an lvalue. @end table @table @code @item gensub(@var{regexp}, @var{replacement}, @var{how} @r{[}, @var{target}@r{]}) @findex gensub @code{gensub} is a general substitution function. Like @code{sub} and @code{gsub}, it searches the target string @var{target} for matches of the regular expression @var{regexp}. Unlike @code{sub} and @code{gsub}, the modified string is returned as the result of the function, and the original target string is @emph{not} changed. If @var{how} is a string beginning with @samp{g} or @samp{G}, then it replaces all matches of @var{regexp} with @var{replacement}. Otherwise, @var{how} is a number indicating which match of @var{regexp} to replace. If no @var{target} is supplied, @code{$0} is used instead. @code{gensub} provides an additional feature that is not available in @code{sub} or @code{gsub}: the ability to specify components of a regexp in the replacement text. This is done by using parentheses in the regexp to mark the components, and then specifying @samp{\@var{n}} in the replacement text, where @var{n} is a digit from one to nine. For example: @example @group $ gawk ' > BEGIN @{ > a = "abc def" > b = gensub(/(.+) (.+)/, "\\2 \\1", "g", a) > print b > @}' @print{} def abc @end group @end example @noindent As described above for @code{sub}, you must type two backslashes in order to get one into the string. In the replacement text, the sequence @samp{\0} represents the entire matched text, as does the character @samp{&}. This example shows how you can use the third argument to control which match of the regexp should be changed. @example $ echo a b c a b c | > gawk '@{ print gensub(/a/, "AA", 2) @}' @print{} a b c AA b c @end example In this case, @code{$0} is used as the default target string. @code{gensub} returns the new string as its result, which is passed directly to @code{print} for printing. If the @var{how} argument is a string that does not begin with @samp{g} or @samp{G}, or if it is a number that is less than zero, only one substitution is performed. If @var{regexp} does not match @var{target}, @code{gensub}'s return value is the original, unchanged value of @var{target}. @cindex differences between @code{gawk} and @code{awk} @code{gensub} is a @code{gawk} extension; it is not available in compatibility mode (@pxref{Options, ,Command Line Options}). @item substr(@var{string}, @var{start} @r{[}, @var{length}@r{]}) @findex substr This returns a @var{length}-character-long substring of @var{string}, starting at character number @var{start}. The first character of a string is character number one. For example, @code{substr("washington", 5, 3)} returns @code{"ing"}. If @var{length} is not present, this function returns the whole suffix of @var{string} that begins at character number @var{start}. For example, @code{substr("washington", 5)} returns @code{"ington"}. The whole suffix is also returned if @var{length} is greater than the number of characters remaining in the string, counting from character number @var{start}. @strong{Note:} The string returned by @code{substr} @emph{cannot} be assigned to. Thus, it is a mistake to attempt to change a portion of a string, like this: @example string = "abcdef" # try to get "abCDEf", won't work substr(string, 3, 3) = "CDE" @end example @noindent or to use @code{substr} as the third agument of @code{sub} or @code{gsub}: @example gsub(/xyz/, "pdq", substr($0, 5, 20)) # WRONG @end example @cindex case conversion @cindex conversion of case @item tolower(@var{string}) @findex tolower This returns a copy of @var{string}, with each upper-case character in the string replaced with its corresponding lower-case character. Non-alphabetic characters are left unchanged. For example, @code{tolower("MiXeD cAsE 123")} returns @code{"mixed case 123"}. @item toupper(@var{string}) @findex toupper This returns a copy of @var{string}, with each lower-case character in the string replaced with its corresponding upper-case character. Non-alphabetic characters are left unchanged. For example, @code{toupper("MiXeD cAsE 123")} returns @code{"MIXED CASE 123"}. @end table @c fakenode --- for prepinfo @subheading More About @samp{\} and @samp{&} with @code{sub}, @code{gsub} and @code{gensub} @cindex escape processing, @code{sub} et. al. When using @code{sub}, @code{gsub} or @code{gensub}, and trying to get literal backslashes and ampersands into the replacement text, you need to remember that there are several levels of @dfn{escape processing} going on. First, there is the @dfn{lexical} level, which is when @code{awk} reads your program, and builds an internal copy of your program that can be executed. Then there is the run-time level, when @code{awk} actually scans the replacement string to determine what to generate. At both levels, @code{awk} looks for a defined set of characters that can come after a backslash. At the lexical level, it looks for the escape sequences listed in @ref{Escape Sequences}. Thus, for every @samp{\} that @code{awk} will process at the run-time level, you type two @samp{\}s at the lexical level. When a character that is not valid for an escape sequence follows the @samp{\}, Unix @code{awk} and @code{gawk} both simply remove the initial @samp{\}, and put the following character into the string. Thus, for example, @code{"a\qb"} is treated as @code{"aqb"}. At the run-time level, the various functions handle sequences of @samp{\} and @samp{&} differently. The situation is (sadly) somewhat complex. Historically, the @code{sub} and @code{gsub} functions treated the two character sequence @samp{\&} specially; this sequence was replaced in the generated text with a single @samp{&}. Any other @samp{\} within the @var{replacement} string that did not precede an @samp{&} was passed through unchanged. To illustrate with a table: @c Thank to Karl Berry for help with the TeX stuff. @tex \vbox{\bigskip % This table has lots of &'s and \'s, so unspecialize them. \catcode`\& = \other \catcode`\\ = \other % But then we need character for escape and tab. @catcode`! = 4 @halign{@hfil#!@qquad@hfil#!@qquad#@hfil@cr You type!@code{sub} sees!@code{sub} generates@cr @hrulefill!@hrulefill!@hrulefill@cr @code{\&}! @code{&}!the matched text@cr @code{\\&}! @code{\&}!a literal @samp{&}@cr @code{\\\&}! @code{\&}!a literal @samp{&}@cr @code{\\\\&}! @code{\\&}!a literal @samp{\&}@cr @code{\\\\\&}! @code{\\&}!a literal @samp{\&}@cr @code{\\\\\\&}! @code{\\\&}!a literal @samp{\\&}@cr @code{\\q}! @code{\q}!a literal @samp{\q}@cr } @bigskip} @end tex @ifinfo @display You type @code{sub} sees @code{sub} generates -------- ---------- --------------- @code{\&} @code{&} the matched text @code{\\&} @code{\&} a literal @samp{&} @code{\\\&} @code{\&} a literal @samp{&} @code{\\\\&} @code{\\&} a literal @samp{\&} @code{\\\\\&} @code{\\&} a literal @samp{\&} @code{\\\\\\&} @code{\\\&} a literal @samp{\\&} @code{\\q} @code{\q} a literal @samp{\q} @end display @end ifinfo @noindent This table shows both the lexical level processing, where an odd number of backslashes becomes an even number at the run time level, and the run-time processing done by @code{sub}. (For the sake of simplicity, the rest of the tables below only show the case of even numbers of @samp{\}s entered at the lexical level.) The problem with the historical approach is that there is no way to get a literal @samp{\} followed by the matched text. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} The 1992 POSIX standard attempted to fix this problem. The standard says that @code{sub} and @code{gsub} look for either a @samp{\} or an @samp{&} after the @samp{\}. If either one follows a @samp{\}, that character is output literally. The interpretation of @samp{\} and @samp{&} then becomes like this: @c thanks to Karl Berry for formatting this table @tex \vbox{\bigskip % This table has lots of &'s and \'s, so unspecialize them. \catcode`\& = \other \catcode`\\ = \other % But then we need character for escape and tab. @catcode`! = 4 @halign{@hfil#!@qquad@hfil#!@qquad#@hfil@cr You type!@code{sub} sees!@code{sub} generates@cr @hrulefill!@hrulefill!@hrulefill@cr @code{&}! @code{&}!the matched text@cr @code{\\&}! @code{\&}!a literal @samp{&}@cr @code{\\\\&}! @code{\\&}!a literal @samp{\}, then the matched text@cr @code{\\\\\\&}! @code{\\\&}!a literal @samp{\&}@cr } @bigskip} @end tex @ifinfo @display You type @code{sub} sees @code{sub} generates -------- ---------- --------------- @code{&} @code{&} the matched text @code{\\&} @code{\&} a literal @samp{&} @code{\\\\&} @code{\\&} a literal @samp{\}, then the matched text @code{\\\\\\&} @code{\\\&} a literal @samp{\&} @end display @end ifinfo @noindent This would appear to solve the problem. Unfortunately, the phrasing of the standard is unusual. It says, in effect, that @samp{\} turns off the special meaning of any following character, but that for anything other than @samp{\} and @samp{&}, such special meaning is undefined. This wording leads to two problems. @enumerate @item Backslashes must now be doubled in the @var{replacement} string, breaking historical @code{awk} programs. @item To make sure that an @code{awk} program is portable, @emph{every} character in the @var{replacement} string must be preceded with a backslash.@footnote{This consequence was certainly unintended.} @c I can say that, 'cause I was involved in making this change @end enumerate The POSIX standard is under revision.@footnote{As of @value{UPDATE-MONTH}, with final approval and publication hopefully sometime in 1997.} Because of the above problems, proposed text for the revised standard reverts to rules that correspond more closely to the original existing practice. The proposed rules have special cases that make it possible to produce a @samp{\} preceding the matched text. @tex \vbox{\bigskip % This table has lots of &'s and \'s, so unspecialize them. \catcode`\& = \other \catcode`\\ = \other % But then we need character for escape and tab. @catcode`! = 4 @halign{@hfil#!@qquad@hfil#!@qquad#@hfil@cr You type!@code{sub} sees!@code{sub} generates@cr @hrulefill!@hrulefill!@hrulefill@cr @code{\\\\\\&}! @code{\\\&}!a literal @samp{\&}@cr @code{\\\\&}! @code{\\&}!a literal @samp{\}, followed by the matched text@cr @code{\\&}! @code{\&}!a literal @samp{&}@cr @code{\\q}! @code{\q}!a literal @samp{\q}@cr } @bigskip} @end tex @ifinfo @display You type @code{sub} sees @code{sub} generates -------- ---------- --------------- @code{\\\\\\&} @code{\\\&} a literal @samp{\&} @code{\\\\&} @code{\\&} a literal @samp{\}, followed by the matched text @code{\\&} @code{\&} a literal @samp{&} @code{\\q} @code{\q} a literal @samp{\q} @end display @end ifinfo In a nutshell, at the run-time level, there are now three special sequences of characters, @samp{\\\&}, @samp{\\&} and @samp{\&}, whereas historically, there was only one. However, as in the historical case, any @samp{\} that is not part of one of these three sequences is not special, and appears in the output literally. @code{gawk} 3.0 follows these proposed POSIX rules for @code{sub} and @code{gsub}. @c As much as we think it's a lousy idea. You win some, you lose some. Sigh. Whether these proposed rules will actually become codified into the standard is unknown at this point. Subsequent @code{gawk} releases will track the standard and implement whatever the final version specifies; this @value{DOCUMENT} will be updated as well. The rules for @code{gensub} are considerably simpler. At the run-time level, whenever @code{gawk} sees a @samp{\}, if the following character is a digit, then the text that matched the corresponding parenthesized subexpression is placed in the generated output. Otherwise, no matter what the character after the @samp{\} is, that character will appear in the generated text, and the @samp{\} will not. @tex \vbox{\bigskip % This table has lots of &'s and \'s, so unspecialize them. \catcode`\& = \other \catcode`\\ = \other % But then we need character for escape and tab. @catcode`! = 4 @halign{@hfil#!@qquad@hfil#!@qquad#@hfil@cr You type!@code{gensub} sees!@code{gensub} generates@cr @hrulefill!@hrulefill!@hrulefill@cr @code{&}! @code{&}!the matched text@cr @code{\\&}! @code{\&}!a literal @samp{&}@cr @code{\\\\}! @code{\\}!a literal @samp{\}@cr @code{\\\\&}! @code{\\&}!a literal @samp{\}, then the matched text@cr @code{\\\\\\&}! @code{\\\&}!a literal @samp{\&}@cr @code{\\q}! @code{\q}!a literal @samp{q}@cr } @bigskip} @end tex @ifinfo @display You type @code{gensub} sees @code{gensub} generates -------- ------------- ------------------ @code{&} @code{&} the matched text @code{\\&} @code{\&} a literal @samp{&} @code{\\\\} @code{\\} a literal @samp{\} @code{\\\\&} @code{\\&} a literal @samp{\}, then the matched text @code{\\\\\\&} @code{\\\&} a literal @samp{\&} @code{\\q} @code{\q} a literal @samp{q} @end display @end ifinfo Because of the complexity of the lexical and run-time level processing, and the special cases for @code{sub} and @code{gsub}, we recommend the use of @code{gawk} and @code{gensub} for when you have to do substitutions. @node I/O Functions, Time Functions, String Functions, Built-in @section Built-in Functions for Input/Output The following functions are related to Input/Output (I/O). Optional parameters are enclosed in square brackets (``['' and ``]''). @table @code @item close(@var{filename}) @findex close Close the file @var{filename}, for input or output. The argument may alternatively be a shell command that was used for redirecting to or from a pipe; then the pipe is closed. @xref{Close Files And Pipes, ,Closing Input and Output Files and Pipes}, for more information. @item fflush(@r{[}@var{filename}@r{]}) @findex fflush @cindex portability issues @cindex flushing buffers @cindex buffers, flushing @cindex buffering output @cindex output, buffering Flush any buffered output associated @var{filename}, which is either a file opened for writing, or a shell command for redirecting output to a pipe. Many utility programs will @dfn{buffer} their output; they save information to be written to a disk file or terminal in memory, until there is enough for it to be worthwhile to send the data to the ouput device. This is often more efficient than writing every little bit of information as soon as it is ready. However, sometimes it is necessary to force a program to @dfn{flush} its buffers; that is, write the information to its destination, even if a buffer is not full. This is the purpose of the @code{fflush} function; @code{gawk} too buffers its output, and the @code{fflush} function can be used to force @code{gawk} to flush its buffers. @code{fflush} is a recent (1994) addition to the Bell Labs research version of @code{awk}; it is not part of the POSIX standard, and will not be available if @samp{--posix} has been specified on the command line (@pxref{Options, ,Command Line Options}). @code{gawk} extends the @code{fflush} function in two ways. The first is to allow no argument at all. In this case, the buffer for the standard output is flushed. The second way is to allow the null string (@w{@code{""}}) as the argument. In this case, the buffers for @emph{all} open output files and pipes are flushed. @code{fflush} returns zero if the buffer was successfully flushed, and nonzero otherwise. @item system(@var{command}) @findex system @cindex interaction, @code{awk} and other programs The @code{system} function allows the user to execute operating system commands and then return to the @code{awk} program. The @code{system} function executes the command given by the string @var{command}. It returns, as its value, the status returned by the command that was executed. For example, if the following fragment of code is put in your @code{awk} program: @example END @{ system("date | mail -s 'awk run done' root") @} @end example @noindent the system administrator will be sent mail when the @code{awk} program finishes processing input and begins its end-of-input processing. Note that redirecting @code{print} or @code{printf} into a pipe is often enough to accomplish your task. If you need to run many commands, it will be more efficient to simply print them to a pipe to the shell: @example while (@var{more stuff to do}) print @var{command} | "/bin/sh" close("/bin/sh") @end example @noindent However, if your @code{awk} program is interactive, @code{system} is useful for cranking up large self-contained programs, such as a shell or an editor. Some operating systems cannot implement the @code{system} function. @code{system} causes a fatal error if it is not supported. @end table @c fakenode --- for prepinfo @subheading Interactive vs. Non-Interactive Buffering @cindex buffering, interactive vs. non-interactive @cindex buffering, non-interactive vs. interactive @cindex interactive buffering vs. non-interactive @cindex non-interactive buffering vs. interactive As a side point, buffering issues can be even more confusing depending upon whether or not your program is @dfn{interactive}, i.e., communicating with a user sitting at a keyboard.@footnote{A program is interactive if the standard output is connected to a terminal device.} Interactive programs generally @dfn{line buffer} their output; they write out every line. Non-interactive programs wait until they have a full buffer, which may be many lines of output. @c Thanks to Walter.Mecky@dresdnerbank.de for this example, and for @c motivating me to write this section. Here is an example of the difference. @example $ awk '@{ print $1 + $2 @}' 1 1 @print{} 2 2 3 @print{} 5 @kbd{Control-d} @end example @noindent Each line of output is printed immediately. Compare that behavior with this example. @example $ awk '@{ print $1 + $2 @}' | cat 1 1 2 3 @kbd{Control-d} @print{} 2 @print{} 5 @end example @noindent Here, no output is printed until after the @kbd{Control-d} is typed, since it is all buffered, and sent down the pipe to @code{cat} in one shot. @c fakenode --- for prepinfo @subheading Controlling Output Buffering with @code{system} @cindex flushing buffers @cindex buffers, flushing @cindex buffering output @cindex output, buffering The @code{fflush} function provides explicit control over output buffering for individual files and pipes. However, its use is not portable to many other @code{awk} implementations. An alternative method to flush output buffers is by calling @code{system} with a null string as its argument: @example system("") # flush output @end example @noindent @code{gawk} treats this use of the @code{system} function as a special case, and is smart enough not to run a shell (or other command interpreter) with the empty command. Therefore, with @code{gawk}, this idiom is not only useful, it is efficient. While this method should work with other @code{awk} implementations, it will not necessarily avoid starting an unnecessary shell. (Other implementations may only flush the buffer associated with the standard output, and not necessarily all buffered output.) If you think about what a programmer expects, it makes sense that @code{system} should flush any pending output. The following program: @example BEGIN @{ print "first print" system("echo system echo") print "second print" @} @end example @noindent must print @example first print system echo second print @end example @noindent and not @example system echo first print second print @end example If @code{awk} did not flush its buffers before calling @code{system}, the latter (undesirable) output is what you would see. @node Time Functions, , I/O Functions, Built-in @section Functions for Dealing with Time Stamps @cindex timestamps @cindex time of day A common use for @code{awk} programs is the processing of log files containing time stamp information, indicating when a particular log record was written. Many programs log their time stamp in the form returned by the @code{time} system call, which is the number of seconds since a particular epoch. On POSIX systems, it is the number of seconds since Midnight, January 1, 1970, UTC. In order to make it easier to process such log files, and to produce useful reports, @code{gawk} provides two functions for working with time stamps. Both of these are @code{gawk} extensions; they are not specified in the POSIX standard, nor are they in any other known version of @code{awk}. Optional parameters are enclosed in square brackets (``['' and ``]''). @table @code @item systime() @findex systime This function returns the current time as the number of seconds since the system epoch. On POSIX systems, this is the number of seconds since Midnight, January 1, 1970, UTC. It may be a different number on other systems. @item strftime(@r{[}@var{format} @r{[}, @var{timestamp}@r{]]}) @findex strftime This function returns a string. It is similar to the function of the same name in ANSI C. The time specified by @var{timestamp} is used to produce a string, based on the contents of the @var{format} string. The @var{timestamp} is in the same format as the value returned by the @code{systime} function. If no @var{timestamp} argument is supplied, @code{gawk} will use the current time of day as the time stamp. If no @var{format} argument is supplied, @code{strftime} uses @code{@w{"%a %b %d %H:%M:%S %Z %Y"}}. This format string produces output (almost) equivalent to that of the @code{date} utility. (Versions of @code{gawk} prior to 3.0 require the @var{format} argument.) @end table The @code{systime} function allows you to compare a time stamp from a log file with the current time of day. In particular, it is easy to determine how long ago a particular record was logged. It also allows you to produce log records using the ``seconds since the epoch'' format. The @code{strftime} function allows you to easily turn a time stamp into human-readable information. It is similar in nature to the @code{sprintf} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}), in that it copies non-format specification characters verbatim to the returned string, while substituting date and time values for format specifications in the @var{format} string. @code{strftime} is guaranteed by the ANSI C standard to support the following date format specifications: @table @code @item %a The locale's abbreviated weekday name. @item %A The locale's full weekday name. @item %b The locale's abbreviated month name. @item %B The locale's full month name. @item %c The locale's ``appropriate'' date and time representation. @item %d The day of the month as a decimal number (01--31). @item %H The hour (24-hour clock) as a decimal number (00--23). @item %I The hour (12-hour clock) as a decimal number (01--12). @item %j The day of the year as a decimal number (001--366). @item %m The month as a decimal number (01--12). @item %M The minute as a decimal number (00--59). @item %p The locale's equivalent of the AM/PM designations associated with a 12-hour clock. @item %S The second as a decimal number (00--60).@footnote{Occasionally there are minutes in a year with a leap second, which is why the seconds can go up to 60.} @item %U The week number of the year (the first Sunday as the first day of week one) as a decimal number (00--53). @item %w The weekday as a decimal number (0--6). Sunday is day zero. @item %W The week number of the year (the first Monday as the first day of week one) as a decimal number (00--53). @item %x The locale's ``appropriate'' date representation. @item %X The locale's ``appropriate'' time representation. @item %y The year without century as a decimal number (00--99). @item %Y The year with century as a decimal number (e.g., 1995). @item %Z The time zone name or abbreviation, or no characters if no time zone is determinable. @item %% A literal @samp{%}. @end table If a conversion specifier is not one of the above, the behavior is undefined.@footnote{This is because ANSI C leaves the behavior of the C version of @code{strftime} undefined, and @code{gawk} will use the system's version of @code{strftime} if it's there. Typically, the conversion specifier will either not appear in the returned string, or it will appear literally.} @cindex locale, definition of Informally, a @dfn{locale} is the geographic place in which a program is meant to run. For example, a common way to abbreviate the date September 4, 1991 in the United States would be ``9/4/91''. In many countries in Europe, however, it would be abbreviated ``4.9.91''. Thus, the @samp{%x} specification in a @code{"US"} locale might produce @samp{9/4/91}, while in a @code{"EUROPE"} locale, it might produce @samp{4.9.91}. The ANSI C standard defines a default @code{"C"} locale, which is an environment that is typical of what most C programmers are used to. A public-domain C version of @code{strftime} is supplied with @code{gawk} for systems that are not yet fully ANSI-compliant. If that version is used to compile @code{gawk} (@pxref{Installation, ,Installing @code{gawk}}), then the following additional format specifications are available: @table @code @item %D Equivalent to specifying @samp{%m/%d/%y}. @item %e The day of the month, padded with a space if it is only one digit. @item %h Equivalent to @samp{%b}, above. @item %n A newline character (ASCII LF). @item %r Equivalent to specifying @samp{%I:%M:%S %p}. @item %R Equivalent to specifying @samp{%H:%M}. @item %T Equivalent to specifying @samp{%H:%M:%S}. @item %t A tab character. @item %k The hour (24-hour clock) as a decimal number (0-23). Single digit numbers are padded with a space. @item %l The hour (12-hour clock) as a decimal number (1-12). Single digit numbers are padded with a space. @item %C The century, as a number between 00 and 99. @item %u The weekday as a decimal number [1 (Monday)--7]. @cindex ISO 8601 @item %V The week number of the year (the first Monday as the first day of week one) as a decimal number (01--53). The method for determining the week number is as specified by ISO 8601 (to wit: if the week containing January 1 has four or more days in the new year, then it is week one, otherwise it is week 53 of the previous year and the next week is week one). @item %G The year with century of the ISO week number, as a decimal number. For example, January 1, 1993, is in week 53 of 1992. Thus, the year of its ISO week number is 1992, even though its year is 1993. Similarly, December 31, 1973, is in week 1 of 1974. Thus, the year of its ISO week number is 1974, even though its year is 1973. @item %g The year without century of the ISO week number, as a decimal number (00--99). @item %Ec %EC %Ex %Ey %EY %Od %Oe %OH %OI @itemx %Om %OM %OS %Ou %OU %OV %Ow %OW %Oy These are ``alternate representations'' for the specifications that use only the second letter (@samp{%c}, @samp{%C}, and so on). They are recognized, but their normal representations are used.@footnote{If you don't understand any of this, don't worry about it; these facilities are meant to make it easier to ``internationalize'' programs.} (These facilitate compliance with the POSIX @code{date} utility.) @item %v The date in VMS format (e.g., 20-JUN-1991). @cindex RFC-822 @cindex RFC-1036 @item %z The timezone offset in a +HHMM format (e.g., the format necessary to produce RFC-822/RFC-1036 date headers). @end table This example is an @code{awk} implementation of the POSIX @code{date} utility. Normally, the @code{date} utility prints the current date and time of day in a well known format. However, if you provide an argument to it that begins with a @samp{+}, @code{date} will copy non-format specifier characters to the standard output, and will interpret the current time according to the format specifiers in the string. For example: @example $ date '+Today is %A, %B %d, %Y.' @print{} Today is Thursday, July 11, 1991. @end example Here is the @code{gawk} version of the @code{date} utility. It has a shell ``wrapper'', to handle the @samp{-u} option, which requires that @code{date} run as if the time zone was set to UTC. @example @group #! /bin/sh # # date --- approximate the P1003.2 'date' command case $1 in -u) TZ=GMT0 # use UTC export TZ shift ;; esac @end group @group gawk 'BEGIN @{ format = "%a %b %d %H:%M:%S %Z %Y" exitval = 0 @end group @group if (ARGC > 2) exitval = 1 else if (ARGC == 2) @{ format = ARGV[1] if (format ~ /^\+/) format = substr(format, 2) # remove leading + @} print strftime(format) exit exitval @}' "$@@" @end group @end example @node User-defined, Invoking Gawk, Built-in, Top @chapter User-defined Functions @cindex user-defined functions @cindex functions, user-defined Complicated @code{awk} programs can often be simplified by defining your own functions. User-defined functions can be called just like built-in ones (@pxref{Function Calls}), but it is up to you to define them---to tell @code{awk} what they should do. @menu * Definition Syntax:: How to write definitions and what they mean. * Function Example:: An example function definition and what it does. * Function Caveats:: Things to watch out for. * Return Statement:: Specifying the value a function returns. @end menu @node Definition Syntax, Function Example, User-defined, User-defined @section Function Definition Syntax @cindex defining functions @cindex function definition Definitions of functions can appear anywhere between the rules of an @code{awk} program. Thus, the general form of an @code{awk} program is extended to include sequences of rules @emph{and} user-defined function definitions. There is no need in @code{awk} to put the definition of a function before all uses of the function. This is because @code{awk} reads the entire program before starting to execute any of it. The definition of a function named @var{name} looks like this: @example function @var{name}(@var{parameter-list}) @{ @var{body-of-function} @} @end example @cindex names, use of @cindex namespaces @noindent @var{name} is the name of the function to be defined. A valid function name is like a valid variable name: a sequence of letters, digits and underscores, not starting with a digit. Within a single @code{awk} program, any particular name can only be used as a variable, array or function. @var{parameter-list} is a list of the function's arguments and local variable names, separated by commas. When the function is called, the argument names are used to hold the argument values given in the call. The local variables are initialized to the empty string. A function cannot have two parameters with the same name. The @var{body-of-function} consists of @code{awk} statements. It is the most important part of the definition, because it says what the function should actually @emph{do}. The argument names exist to give the body a way to talk about the arguments; local variables, to give the body places to keep temporary values. Argument names are not distinguished syntactically from local variable names; instead, the number of arguments supplied when the function is called determines how many argument variables there are. Thus, if three argument values are given, the first three names in @var{parameter-list} are arguments, and the rest are local variables. It follows that if the number of arguments is not the same in all calls to the function, some of the names in @var{parameter-list} may be arguments on some occasions and local variables on others. Another way to think of this is that omitted arguments default to the null string. Usually when you write a function you know how many names you intend to use for arguments and how many you intend to use as local variables. It is conventional to place some extra space between the arguments and the local variables, to document how your function is supposed to be used. @cindex variable shadowing During execution of the function body, the arguments and local variable values hide or @dfn{shadow} any variables of the same names used in the rest of the program. The shadowed variables are not accessible in the function definition, because there is no way to name them while their names have been taken away for the local variables. All other variables used in the @code{awk} program can be referenced or set normally in the function's body. The arguments and local variables last only as long as the function body is executing. Once the body finishes, you can once again access the variables that were shadowed while the function was running. @cindex recursive function @cindex function, recursive The function body can contain expressions which call functions. They can even call this function, either directly or by way of another function. When this happens, we say the function is @dfn{recursive}. @cindex @code{awk} language, POSIX version @cindex POSIX @code{awk} In many @code{awk} implementations, including @code{gawk}, the keyword @code{function} may be abbreviated @code{func}. However, POSIX only specifies the use of the keyword @code{function}. This actually has some practical implications. If @code{gawk} is in POSIX-compatibility mode (@pxref{Options, ,Command Line Options}), then the following statement will @emph{not} define a function: @example func foo() @{ a = sqrt($1) ; print a @} @end example @noindent Instead it defines a rule that, for each record, concatenates the value of the variable @samp{func} with the return value of the function @samp{foo}. If the resulting string is non-null, the action is executed. This is probably not what was desired. (@code{awk} accepts this input as syntactically valid, since functions may be used before they are defined in @code{awk} programs.) @cindex portability issues To ensure that your @code{awk} programs are portable, always use the keyword @code{function} when defining a function. @node Function Example, Function Caveats, Definition Syntax, User-defined @section Function Definition Examples Here is an example of a user-defined function, called @code{myprint}, that takes a number and prints it in a specific format. @example function myprint(num) @{ printf "%6.3g\n", num @} @end example @noindent To illustrate, here is an @code{awk} rule which uses our @code{myprint} function: @example $3 > 0 @{ myprint($3) @} @end example @noindent This program prints, in our special format, all the third fields that contain a positive number in our input. Therefore, when given: @example @group 1.2 3.4 5.6 7.8 9.10 11.12 -13.14 15.16 17.18 19.20 21.22 23.24 @end group @end example @noindent this program, using our function to format the results, prints: @example 5.6 21.2 @end example This function deletes all the elements in an array. @example function delarray(a, i) @{ for (i in a) delete a[i] @} @end example When working with arrays, it is often necessary to delete all the elements in an array and start over with a new list of elements (@pxref{Delete, ,The @code{delete} Statement}). Instead of having to repeat this loop everywhere in your program that you need to clear out an array, your program can just call @code{delarray}. Here is an example of a recursive function. It takes a string as an input parameter, and returns the string in backwards order. @example function rev(str, start) @{ if (start == 0) return "" return (substr(str, start, 1) rev(str, start - 1)) @} @end example If this function is in a file named @file{rev.awk}, we can test it this way: @example $ echo "Don't Panic!" | > gawk --source '@{ print rev($0, length($0)) @}' -f rev.awk @print{} !cinaP t'noD @end example Here is an example that uses the built-in function @code{strftime}. (@xref{Time Functions, ,Functions for Dealing with Time Stamps}, for more information on @code{strftime}.) The C @code{ctime} function takes a timestamp and returns it in a string, formatted in a well known fashion. Here is an @code{awk} version: @example @c file eg/lib/ctime.awk @group # ctime.awk # # awk version of C ctime(3) function function ctime(ts, format) @{ format = "%a %b %d %H:%M:%S %Z %Y" if (ts == 0) ts = systime() # use current time as default return strftime(format, ts) @} @c endfile @end group @end example @node Function Caveats, Return Statement, Function Example, User-defined @section Calling User-defined Functions @cindex call by value @cindex call by reference @cindex calling a function @cindex function call @dfn{Calling a function} means causing the function to run and do its job. A function call is an expression, and its value is the value returned by the function. A function call consists of the function name followed by the arguments in parentheses. What you write in the call for the arguments are @code{awk} expressions; each time the call is executed, these expressions are evaluated, and the values are the actual arguments. For example, here is a call to @code{foo} with three arguments (the first being a string concatenation): @example foo(x y, "lose", 4 * z) @end example @strong{Caution:} whitespace characters (spaces and tabs) are not allowed between the function name and the open-parenthesis of the argument list. If you write whitespace by mistake, @code{awk} might think that you mean to concatenate a variable with an expression in parentheses. However, it notices that you used a function name and not a variable name, and reports an error. @cindex call by value When a function is called, it is given a @emph{copy} of the values of its arguments. This is known as @dfn{call by value}. The caller may use a variable as the expression for the argument, but the called function does not know this: it only knows what value the argument had. For example, if you write this code: @example foo = "bar" z = myfunc(foo) @end example @noindent then you should not think of the argument to @code{myfunc} as being ``the variable @code{foo}.'' Instead, think of the argument as the string value, @code{"bar"}. If the function @code{myfunc} alters the values of its local variables, this has no effect on any other variables. Thus, if @code{myfunc} does this: @example @group function myfunc(str) @{ print str str = "zzz" print str @} @end group @end example @noindent to change its first argument variable @code{str}, this @emph{does not} change the value of @code{foo} in the caller. The role of @code{foo} in calling @code{myfunc} ended when its value, @code{"bar"}, was computed. If @code{str} also exists outside of @code{myfunc}, the function body cannot alter this outer value, because it is shadowed during the execution of @code{myfunc} and cannot be seen or changed from there. @cindex call by reference However, when arrays are the parameters to functions, they are @emph{not} copied. Instead, the array itself is made available for direct manipulation by the function. This is usually called @dfn{call by reference}. Changes made to an array parameter inside the body of a function @emph{are} visible outside that function. @ifinfo This can be @strong{very} dangerous if you do not watch what you are doing. For example: @end ifinfo @iftex @emph{This can be very dangerous if you do not watch what you are doing.} For example: @end iftex @example function changeit(array, ind, nvalue) @{ array[ind] = nvalue @} BEGIN @{ a[1] = 1; a[2] = 2; a[3] = 3 changeit(a, 2, "two") printf "a[1] = %s, a[2] = %s, a[3] = %s\n", a[1], a[2], a[3] @} @end example @noindent This program prints @samp{a[1] = 1, a[2] = two, a[3] = 3}, because @code{changeit} stores @code{"two"} in the second element of @code{a}. @cindex undefined functions @cindex functions, undefined Some @code{awk} implementations allow you to call a function that has not been defined, and only report a problem at run-time when the program actually tries to call the function. For example: @example @group BEGIN @{ if (0) foo() else bar() @} function bar() @{ @dots{} @} # note that `foo' is not defined @end group @end example @noindent Since the @samp{if} statement will never be true, it is not really a problem that @code{foo} has not been defined. Usually though, it is a problem if a program calls an undefined function. @ignore At one point, I had gawk dieing on this, but later decided that this might break old programs and/or test suites. @end ignore If @samp{--lint} has been specified (@pxref{Options, ,Command Line Options}), @code{gawk} will report about calls to undefined functions. Some @code{awk} implementations generate a run-time error if you use the @code{next} statement (@pxref{Next Statement, , The @code{next} Statement}) inside a user-defined function. @code{gawk} does not have this problem. @node Return Statement, , Function Caveats, User-defined @section The @code{return} Statement @cindex @code{return} statement The body of a user-defined function can contain a @code{return} statement. This statement returns control to the rest of the @code{awk} program. It can also be used to return a value for use in the rest of the @code{awk} program. It looks like this: @example return @r{[}@var{expression}@r{]} @end example The @var{expression} part is optional. If it is omitted, then the returned value is undefined and, therefore, unpredictable. A @code{return} statement with no value expression is assumed at the end of every function definition. So if control reaches the end of the function body, then the function returns an unpredictable value. @code{awk} will @emph{not} warn you if you use the return value of such a function. Sometimes, you want to write a function for what it does, not for what it returns. Such a function corresponds to a @code{void} function in C or to a @code{procedure} in Pascal. Thus, it may be appropriate to not return any value; you should simply bear in mind that if you use the return value of such a function, you do so at your own risk. Here is an example of a user-defined function that returns a value for the largest number among the elements of an array: @example @group function maxelt(vec, i, ret) @{ for (i in vec) @{ if (ret == "" || vec[i] > ret) ret = vec[i] @} return ret @} @end group @end example @noindent You call @code{maxelt} with one argument, which is an array name. The local variables @code{i} and @code{ret} are not intended to be arguments; while there is nothing to stop you from passing two or three arguments to @code{maxelt}, the results would be strange. The extra space before @code{i} in the function parameter list indicates that @code{i} and @code{ret} are not supposed to be arguments. This is a convention that you should follow when you define functions. Here is a program that uses our @code{maxelt} function. It loads an array, calls @code{maxelt}, and then reports the maximum number in that array: @example @group awk ' function maxelt(vec, i, ret) @{ for (i in vec) @{ if (ret == "" || vec[i] > ret) ret = vec[i] @} return ret @} @end group @group # Load all fields of each record into nums. @{ for(i = 1; i <= NF; i++) nums[NR, i] = $i @} END @{ print maxelt(nums) @}' @end group @end example Given the following input: @example @group 1 5 23 8 16 44 3 5 2 8 26 256 291 1396 2962 100 -6 467 998 1101 99385 11 0 225 @end group @end example @noindent our program tells us (predictably) that @code{99385} is the largest number in our array. @node Invoking Gawk, Library Functions, User-defined, Top @chapter Running @code{awk} @cindex command line @cindex invocation of @code{gawk} @cindex arguments, command line @cindex options, command line @cindex long options @cindex options, long There are two ways to run @code{awk}: with an explicit program, or with one or more program files. Here are templates for both of them; items enclosed in @samp{@r{[}@dots{}@r{]}} in these templates are optional. Besides traditional one-letter POSIX-style options, @code{gawk} also supports GNU long options. @example awk @r{[@var{options}]} -f progfile @r{[@code{--}]} @var{file} @dots{} awk @r{[@var{options}]} @r{[@code{--}]} '@var{program}' @var{file} @dots{} @end example @cindex empty program @cindex dark corner It is possible to invoke @code{awk} with an empty program: @example $ awk '' datafile1 datafile2 @end example @noindent Doing so makes little sense though; @code{awk} will simply exit silently when given an empty program (d.c.). If @samp{--lint} has been specified on the command line, @code{gawk} will issue a warning that the program is empty. @menu * Options:: Command line options and their meanings. * Other Arguments:: Input file names and variable assignments. * AWKPATH Variable:: Searching directories for @code{awk} programs. * Obsolete:: Obsolete Options and/or features. * Undocumented:: Undocumented Options and Features. * Known Bugs:: Known Bugs in @code{gawk}. @end menu @node Options, Other Arguments, Invoking Gawk, Invoking Gawk @section Command Line Options Options begin with a dash, and consist of a single character. GNU style long options consist of two dashes and a keyword. The keyword can be abbreviated, as long the abbreviation allows the option to be uniquely identified. If the option takes an argument, then the keyword is either immediately followed by an equals sign (@samp{=}) and the argument's value, or the keyword and the argument's value are separated by whitespace. For brevity, the discussion below only refers to the traditional short options; however the long and short options are interchangeable in all contexts. Each long option for @code{gawk} has a corresponding POSIX-style option. The options and their meanings are as follows: @table @code @item -F @var{fs} @itemx --field-separator @var{fs} @cindex @code{-F} option @cindex @code{--field-separator} option Sets the @code{FS} variable to @var{fs} (@pxref{Field Separators, ,Specifying How Fields are Separated}). @item -f @var{source-file} @itemx --file @var{source-file} @cindex @code{-f} option @cindex @code{--file} option Indicates that the @code{awk} program is to be found in @var{source-file} instead of in the first non-option argument. @item -v @var{var}=@var{val} @itemx --assign @var{var}=@var{val} @cindex @code{-v} option @cindex @code{--assign} option Sets the variable @var{var} to the value @var{val} @strong{before} execution of the program begins. Such variable values are available inside the @code{BEGIN} rule (@pxref{Other Arguments, ,Other Command Line Arguments}). The @samp{-v} option can only set one variable, but you can use it more than once, setting another variable each time, like this: @samp{awk @w{-v foo=1} @w{-v bar=2} @dots{}}. @item -mf @var{NNN} @itemx -mr @var{NNN} Set various memory limits to the value @var{NNN}. The @samp{f} flag sets the maximum number of fields, and the @samp{r} flag sets the maximum record size. These two flags and the @samp{-m} option are from the Bell Labs research version of Unix @code{awk}. They are provided for compatibility, but otherwise ignored by @code{gawk}, since @code{gawk} has no predefined limits. @item -W @var{gawk-opt} @cindex @code{-W} option Following the POSIX standard, options that are implementation specific are supplied as arguments to the @samp{-W} option. These options also have corresponding GNU style long options. See below. @item -- Signals the end of the command line options. The following arguments are not treated as options even if they begin with @samp{-}. This interpretation of @samp{--} follows the POSIX argument parsing conventions. This is useful if you have file names that start with @samp{-}, or in shell scripts, if you have file names that will be specified by the user which could start with @samp{-}. @end table The following @code{gawk}-specific options are available: @table @code @item -W traditional @itemx -W compat @itemx --traditional @itemx --compat @cindex @code{--compat} option @cindex @code{--traditional} option @cindex compatibility mode Specifies @dfn{compatibility mode}, in which the GNU extensions to the @code{awk} language are disabled, so that @code{gawk} behaves just like the Bell Labs research version of Unix @code{awk}. @samp{--traditional} is the preferred form of this option. @xref{POSIX/GNU, ,Extensions in @code{gawk} Not in POSIX @code{awk}}, which summarizes the extensions. Also see @ref{Compatibility Mode, ,Downward Compatibility and Debugging}. @item -W copyleft @itemx -W copyright @itemx --copyleft @itemx --copyright @cindex @code{--copyleft} option @cindex @code{--copyright} option Print the short version of the General Public License, and then exit. This option may disappear in a future version of @code{gawk}. @item -W help @itemx -W usage @itemx --help @itemx --usage @cindex @code{--help} option @cindex @code{--usage} option Print a ``usage'' message summarizing the short and long style options that @code{gawk} accepts, and then exit. @item -W lint @itemx --lint @cindex @code{--lint} option Warn about constructs that are dubious or non-portable to other @code{awk} implementations. Some warnings are issued when @code{gawk} first reads your program. Others are issued at run-time, as your program executes. @item -W lint-old @itemx --lint-old @cindex @code{--lint-old} option Warn about constructs that are not available in the original Version 7 Unix version of @code{awk} (@pxref{V7/SVR3.1, , Major Changes between V7 and SVR3.1}). @item -W posix @itemx --posix @cindex @code{--posix} option @cindex POSIX mode Operate in strict POSIX mode. This disables all @code{gawk} extensions (just like @samp{--traditional}), and adds the following additional restrictions: @c IMPORTANT! Keep this list in sync with the one in node POSIX @itemize @bullet @item @code{\x} escape sequences are not recognized (@pxref{Escape Sequences}). @item Newlines do not act as whitespace to separate fields when @code{FS} is equal to a single space. @item The synonym @code{func} for the keyword @code{function} is not recognized (@pxref{Definition Syntax, ,Function Definition Syntax}). @item The operators @samp{**} and @samp{**=} cannot be used in place of @samp{^} and @samp{^=} (@pxref{Arithmetic Ops, ,Arithmetic Operators}, and also @pxref{Assignment Ops, ,Assignment Expressions}). @item Specifying @samp{-Ft} on the command line does not set the value of @code{FS} to be a single tab character (@pxref{Field Separators, ,Specifying How Fields are Separated}). @item The @code{fflush} built-in function is not supported (@pxref{I/O Functions, , Built-in Functions for Input/Output}). @end itemize If you supply both @samp{--traditional} and @samp{--posix} on the command line, @samp{--posix} will take precedence. @code{gawk} will also issue a warning if both options are supplied. @item -W re-interval @itemx --re-interval Allow interval expressions (@pxref{Regexp Operators, , Regular Expression Operators}), in regexps. Because interval expressions were traditionally not available in @code{awk}, @code{gawk} does not provide them by default. This prevents old @code{awk} programs from breaking. @item -W source @var{program-text} @itemx --source @var{program-text} @cindex @code{--source} option Program source code is taken from the @var{program-text}. This option allows you to mix source code in files with source code that you enter on the command line. This is particularly useful when you have library functions that you wish to use from your command line programs (@pxref{AWKPATH Variable, ,The @code{AWKPATH} Environment Variable}). @item -W version @itemx --version @cindex @code{--version} option Prints version information for this particular copy of @code{gawk}. This allows you to determine if your copy of @code{gawk} is up to date with respect to whatever the Free Software Foundation is currently distributing. It is also useful for bug reports (@pxref{Bugs, , Reporting Problems and Bugs}). @end table Any other options are flagged as invalid with a warning message, but are otherwise ignored. In compatibility mode, as a special case, if the value of @var{fs} supplied to the @samp{-F} option is @samp{t}, then @code{FS} is set to the tab character (@code{"\t"}). This is only true for @samp{--traditional}, and not for @samp{--posix} (@pxref{Field Separators, ,Specifying How Fields are Separated}). The @samp{-f} option may be used more than once on the command line. If it is, @code{awk} reads its program source from all of the named files, as if they had been concatenated together into one big file. This is useful for creating libraries of @code{awk} functions. Useful functions can be written once, and then retrieved from a standard place, instead of having to be included into each individual program. You can type in a program at the terminal and still use library functions, by specifying @samp{-f /dev/tty}. @code{awk} will read a file from the terminal to use as part of the @code{awk} program. After typing your program, type @kbd{Control-d} (the end-of-file character) to terminate it. (You may also use @samp{-f -} to read program source from the standard input, but then you will not be able to also use the standard input as a source of data.) Because it is clumsy using the standard @code{awk} mechanisms to mix source file and command line @code{awk} programs, @code{gawk} provides the @samp{--source} option. This does not require you to pre-empt the standard input for your source code, and allows you to easily mix command line and library source code (@pxref{AWKPATH Variable, ,The @code{AWKPATH} Environment Variable}). If no @samp{-f} or @samp{--source} option is specified, then @code{gawk} will use the first non-option command line argument as the text of the program source code. @cindex @code{POSIXLY_CORRECT} environment variable @cindex environment variable, @code{POSIXLY_CORRECT} If the environment variable @code{POSIXLY_CORRECT} exists, then @code{gawk} will behave in strict POSIX mode, exactly as if you had supplied the @samp{--posix} command line option. Many GNU programs look for this environment variable to turn on strict POSIX mode. If you supply @samp{--lint} on the command line, and @code{gawk} turns on POSIX mode because of @code{POSIXLY_CORRECT}, then it will print a warning message indicating that POSIX mode is in effect. You would typically set this variable in your shell's startup file. For a Bourne compatible shell (such as Bash), you would add these lines to the @file{.profile} file in your home directory. @example @group POSIXLY_CORRECT=true export POSIXLY_CORRECT @end group @end example For a @code{csh} compatible shell,@footnote{Not recommended.} you would add this line to the @file{.login} file in your home directory. @example setenv POSIXLY_CORRECT true @end example @node Other Arguments, AWKPATH Variable, Options, Invoking Gawk @section Other Command Line Arguments Any additional arguments on the command line are normally treated as input files to be processed in the order specified. However, an argument that has the form @code{@var{var}=@var{value}}, assigns the value @var{value} to the variable @var{var}---it does not specify a file at all. @vindex ARGIND @vindex ARGV All these arguments are made available to your @code{awk} program in the @code{ARGV} array (@pxref{Built-in Variables}). Command line options and the program text (if present) are omitted from @code{ARGV}. All other arguments, including variable assignments, are included. As each element of @code{ARGV} is processed, @code{gawk} sets the variable @code{ARGIND} to the index in @code{ARGV} of the current element. The distinction between file name arguments and variable-assignment arguments is made when @code{awk} is about to open the next input file. At that point in execution, it checks the ``file name'' to see whether it is really a variable assignment; if so, @code{awk} sets the variable instead of reading a file. Therefore, the variables actually receive the given values after all previously specified files have been read. In particular, the values of variables assigned in this fashion are @emph{not} available inside a @code{BEGIN} rule (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}), since such rules are run before @code{awk} begins scanning the argument list. @cindex dark corner The variable values given on the command line are processed for escape sequences (d.c.) (@pxref{Escape Sequences}). In some earlier implementations of @code{awk}, when a variable assignment occurred before any file names, the assignment would happen @emph{before} the @code{BEGIN} rule was executed. @code{awk}'s behavior was thus inconsistent; some command line assignments were available inside the @code{BEGIN} rule, while others were not. However, some applications came to depend upon this ``feature.'' When @code{awk} was changed to be more consistent, the @samp{-v} option was added to accommodate applications that depended upon the old behavior. The variable assignment feature is most useful for assigning to variables such as @code{RS}, @code{OFS}, and @code{ORS}, which control input and output formats, before scanning the data files. It is also useful for controlling state if multiple passes are needed over a data file. For example: @cindex multiple passes over data @cindex passes, multiple @example awk 'pass == 1 @{ @var{pass 1 stuff} @} pass == 2 @{ @var{pass 2 stuff} @}' pass=1 mydata pass=2 mydata @end example Given the variable assignment feature, the @samp{-F} option for setting the value of @code{FS} is not strictly necessary. It remains for historical compatibility. @node AWKPATH Variable, Obsolete, Other Arguments, Invoking Gawk @section The @code{AWKPATH} Environment Variable @cindex @code{AWKPATH} environment variable @cindex environment variable, @code{AWKPATH} @cindex search path @cindex directory search @cindex path, search @cindex differences between @code{gawk} and @code{awk} The previous section described how @code{awk} program files can be named on the command line with the @samp{-f} option. In most @code{awk} implementations, you must supply a precise path name for each program file, unless the file is in the current directory. @cindex search path, for source files But in @code{gawk}, if the file name supplied to the @samp{-f} option does not contain a @samp{/}, then @code{gawk} searches a list of directories (called the @dfn{search path}), one by one, looking for a file with the specified name. The search path is a string consisting of directory names separated by colons. @code{gawk} gets its search path from the @code{AWKPATH} environment variable. If that variable does not exist, @code{gawk} uses a default path, which is @samp{.:/usr/local/share/awk}.@footnote{Your version of @code{gawk} may use a directory that is different than @file{/usr/local/share/awk}; it will depend upon how @code{gawk} was built and installed. The actual directory will be the value of @samp{$(datadir)} generated when @code{gawk} was configured. You probably don't need to worry about this though.} (Programs written for use by system administrators should use an @code{AWKPATH} variable that does not include the current directory, @file{.}.) The search path feature is particularly useful for building up libraries of useful @code{awk} functions. The library files can be placed in a standard directory that is in the default path, and then specified on the command line with a short file name. Otherwise, the full file name would have to be typed for each file. By using both the @samp{--source} and @samp{-f} options, your command line @code{awk} programs can use facilities in @code{awk} library files. @xref{Library Functions, , A Library of @code{awk} Functions}. Path searching is not done if @code{gawk} is in compatibility mode. This is true for both @samp{--traditional} and @samp{--posix}. @xref{Options, ,Command Line Options}. @strong{Note:} if you want files in the current directory to be found, you must include the current directory in the path, either by including @file{.} explicitly in the path, or by writing a null entry in the path. (A null entry is indicated by starting or ending the path with a colon, or by placing two colons next to each other (@samp{::}).) If the current directory is not included in the path, then files cannot be found in the current directory. This path search mechanism is identical to the shell's. @c someday, @cite{The Bourne Again Shell}.... Starting with version 3.0, if @code{AWKPATH} is not defined in the environment, @code{gawk} will place its default search path into @code{ENVIRON["AWKPATH"]}. This makes it easy to determine the actual search path @code{gawk} will use. @node Obsolete, Undocumented, AWKPATH Variable, Invoking Gawk @section Obsolete Options and/or Features @cindex deprecated options @cindex obsolete options @cindex deprecated features @cindex obsolete features This section describes features and/or command line options from previous releases of @code{gawk} that are either not available in the current version, or that are still supported but deprecated (meaning that they will @emph{not} be in the next release). @c update this section for each release! For version @value{VERSION}.@value{PATCHLEVEL} of @code{gawk}, there are no command line options or other deprecated features from the previous version of @code{gawk}. @iftex This section @end iftex @ifinfo This node @end ifinfo is thus essentially a place holder, in case some option becomes obsolete in a future version of @code{gawk}. @ignore @c This is pretty old news... The public-domain version of @code{strftime} that is distributed with @code{gawk} changed for the 2.14 release. The @samp{%V} conversion specifier that used to generate the date in VMS format was changed to @samp{%v}. This is because the POSIX standard for the @code{date} utility now specifies a @samp{%V} conversion specifier. @xref{Time Functions, ,Functions for Dealing with Time Stamps}, for details. @end ignore @node Undocumented, Known Bugs, Obsolete, Invoking Gawk @section Undocumented Options and Features @cindex undocumented features @display @i{Use the Source, Luke!} Obi-Wan @end display @sp 1 This section intentionally left blank. @c Read The Source, Luke! @ignore @c If these came out in the Info file or TeX document, then they wouldn't @c be undocumented, would they? @code{gawk} has one undocumented option: @table @code @item -W nostalgia @itemx --nostalgia Print the message @code{"awk: bailing out near line 1"} and dump core. This option was inspired by the common behavior of very early versions of Unix @code{awk}, and by a t--shirt. @end table Early versions of @code{awk} used to not require any separator (either a newline or @samp{;}) between the rules in @code{awk} programs. Thus, it was common to see one-line programs like: @example awk '@{ sum += $1 @} END @{ print sum @}' @end example @code{gawk} actually supports this, but it is purposely undocumented since it is considered bad style. The correct way to write such a program is either @example awk '@{ sum += $1 @} ; END @{ print sum @}' @end example @noindent or @example awk '@{ sum += $1 @} END @{ print sum @}' data @end example @noindent @xref{Statements/Lines, ,@code{awk} Statements Versus Lines}, for a fuller explanation. @end ignore @node Known Bugs, , Undocumented, Invoking Gawk @section Known Bugs in @code{gawk} @cindex bugs, known in @code{gawk} @cindex known bugs @itemize @bullet @item The @samp{-F} option for changing the value of @code{FS} (@pxref{Options, ,Command Line Options}) is not necessary given the command line variable assignment feature; it remains only for backwards compatibility. @item If your system actually has support for @file{/dev/fd} and the associated @file{/dev/stdin}, @file{/dev/stdout}, and @file{/dev/stderr} files, you may get different output from @code{gawk} than you would get on a system without those files. When @code{gawk} interprets these files internally, it synchronizes output to the standard output with output to @file{/dev/stdout}, while on a system with those files, the output is actually to different open files (@pxref{Special Files, ,Special File Names in @code{gawk}}). @item Syntactically invalid single character programs tend to overflow the parse stack, generating a rather unhelpful message. Such programs are surprisingly difficult to diagnose in the completely general case, and the effort to do so really is not worth it. @end itemize @node Library Functions, Sample Programs, Invoking Gawk, Top @chapter A Library of @code{awk} Functions @c 2e: USE TEXINFO-2 FUNCTION DEFINITION STUFF!!!!!!!!!!!!! This chapter presents a library of useful @code{awk} functions. The sample programs presented later (@pxref{Sample Programs, ,Practical @code{awk} Programs}) use these functions. The functions are presented here in a progression from simple to complex. @ref{Extract Program, ,Extracting Programs from Texinfo Source Files}, presents a program that you can use to extract the source code for these example library functions and programs from the Texinfo source for this @value{DOCUMENT}. (This has already been done as part of the @code{gawk} distribution.) If you have written one or more useful, general purpose @code{awk} functions, and would like to contribute them for a subsequent edition of this @value{DOCUMENT}, please contact the author. @xref{Bugs, ,Reporting Problems and Bugs}, for information on doing this. Don't just send code, as you will be required to either place your code in the public domain, publish it under the GPL (@pxref{Copying, ,GNU GENERAL PUBLIC LICENSE}), or assign the copyright in it to the Free Software Foundation. @menu * Portability Notes:: What to do if you don't have @code{gawk}. * Nextfile Function:: Two implementations of a @code{nextfile} function. * Assert Function:: A function for assertions in @code{awk} programs. * Round Function:: A function for rounding if @code{sprintf} does not do it correctly. * Ordinal Functions:: Functions for using characters as numbers and vice versa. * Join Function:: A function to join an array into a string. * Mktime Function:: A function to turn a date into a timestamp. * Gettimeofday Function:: A function to get formatted times. * Filetrans Function:: A function for handling data file transitions. * Getopt Function:: A function for processing command line arguments. * Passwd Functions:: Functions for getting user information. * Group Functions:: Functions for getting group information. * Library Names:: How to best name private global variables in library functions. @end menu @node Portability Notes, Nextfile Function, Library Functions, Library Functions @section Simulating @code{gawk}-specific Features @cindex portability issues The programs in this chapter and in @ref{Sample Programs, ,Practical @code{awk} Programs}, freely use features that are specific to @code{gawk}. This section briefly discusses how you can rewrite these programs for different implementations of @code{awk}. Diagnostic error messages are sent to @file{/dev/stderr}. Use @samp{| "cat 1>&2"} instead of @samp{> "/dev/stderr"}, if your system does not have a @file{/dev/stderr}, or if you cannot use @code{gawk}. A number of programs use @code{nextfile} (@pxref{Nextfile Statement, ,The @code{nextfile} Statement}), to skip any remaining input in the input file. @ref{Nextfile Function, ,Implementing @code{nextfile} as a Function}, shows you how to write a function that will do the same thing. Finally, some of the programs choose to ignore upper-case and lower-case distinctions in their input. They do this by assigning one to @code{IGNORECASE}. You can achieve the same effect by adding the following rule to the beginning of the program: @example # ignore case @{ $0 = tolower($0) @} @end example @noindent Also, verify that all regexp and string constants used in comparisons only use lower-case letters. @node Nextfile Function, Assert Function, Portability Notes, Library Functions @section Implementing @code{nextfile} as a Function @cindex skipping input files @cindex input files, skipping The @code{nextfile} statement presented in @ref{Nextfile Statement, ,The @code{nextfile} Statement}, is a @code{gawk}-specific extension. It is not available in other implementations of @code{awk}. This section shows two versions of a @code{nextfile} function that you can use to simulate @code{gawk}'s @code{nextfile} statement if you cannot use @code{gawk}. Here is a first attempt at writing a @code{nextfile} function. @example @group # nextfile --- skip remaining records in current file # this should be read in before the "main" awk program function nextfile() @{ _abandon_ = FILENAME; next @} _abandon_ == FILENAME @{ next @} @end group @end example This file should be included before the main program, because it supplies a rule that must be executed first. This rule compares the current data file's name (which is always in the @code{FILENAME} variable) to a private variable named @code{_abandon_}. If the file name matches, then the action part of the rule executes a @code{next} statement, to go on to the next record. (The use of @samp{_} in the variable name is a convention. It is discussed more fully in @ref{Library Names, , Naming Library Function Global Variables}.) The use of the @code{next} statement effectively creates a loop that reads all the records from the current data file. Eventually, the end of the file is reached, and a new data file is opened, changing the value of @code{FILENAME}. Once this happens, the comparison of @code{_abandon_} to @code{FILENAME} fails, and execution continues with the first rule of the ``real'' program. The @code{nextfile} function itself simply sets the value of @code{_abandon_} and then executes a @code{next} statement to start the loop going.@footnote{Some implementations of @code{awk} do not allow you to execute @code{next} from within a function body. Some other work-around will be necessary if you use such a version.} @c mawk is what we're talking about. This initial version has a subtle problem. What happens if the same data file is listed @emph{twice} on the command line, one right after the other, or even with just a variable assignment between the two occurrences of the file name? @c @findex nextfile @c do it this way, since all the indices are merged @cindex @code{nextfile} function In such a case, this code will skip right through the file, a second time, even though it should stop when it gets to the end of the first occurrence. Here is a second version of @code{nextfile} that remedies this problem. @example @group @c file eg/lib/nextfile.awk # nextfile --- skip remaining records in current file # correctly handle successive occurrences of the same file # Arnold Robbins, arnold@@gnu.org, Public Domain # May, 1993 # this should be read in before the "main" awk program function nextfile() @{ _abandon_ = FILENAME; next @} _abandon_ == FILENAME @{ if (FNR == 1) _abandon_ = "" else next @} @c endfile @end group @end example The @code{nextfile} function has not changed. It sets @code{_abandon_} equal to the current file name and then executes a @code{next} satement. The @code{next} statement reads the next record and increments @code{FNR}, so @code{FNR} is guaranteed to have a value of at least two. However, if @code{nextfile} is called for the last record in the file, then @code{awk} will close the current data file and move on to the next one. Upon doing so, @code{FILENAME} will be set to the name of the new file, and @code{FNR} will be reset to one. If this next file is the same as the previous one, @code{_abandon_} will still be equal to @code{FILENAME}. However, @code{FNR} will be equal to one, telling us that this is a new occurrence of the file, and not the one we were reading when the @code{nextfile} function was executed. In that case, @code{_abandon_} is reset to the empty string, so that further executions of this rule will fail (until the next time that @code{nextfile} is called). If @code{FNR} is not one, then we are still in the original data file, and the program executes a @code{next} statement to skip through it. An important question to ask at this point is: ``Given that the functionality of @code{nextfile} can be provided with a library file, why is it built into @code{gawk}?'' This is an important question. Adding features for little reason leads to larger, slower programs that are harder to maintain. The answer is that building @code{nextfile} into @code{gawk} provides significant gains in efficiency. If the @code{nextfile} function is executed at the beginning of a large data file, @code{awk} still has to scan the entire file, splitting it up into records, just to skip over it. The built-in @code{nextfile} can simply close the file immediately and proceed to the next one, saving a lot of time. This is particularly important in @code{awk}, since @code{awk} programs are generally I/O bound (i.e.@: they spend most of their time doing input and output, instead of performing computations). @node Assert Function, Round Function, Nextfile Function, Library Functions @section Assertions @cindex assertions @cindex @code{assert}, C version When writing large programs, it is often useful to be able to know that a condition or set of conditions is true. Before proceeding with a particular computation, you make a statement about what you believe to be the case. Such a statement is known as an ``assertion.'' The C language provides an @code{} header file and corresponding @code{assert} macro that the programmer can use to make assertions. If an assertion fails, the @code{assert} macro arranges to print a diagnostic message describing the condition that should have been true but was not, and then it kills the program. In C, using @code{assert} looks this: @example #include int myfunc(int a, double b) @{ assert(a <= 5 && b >= 17); @dots{} @} @end example If the assertion failed, the program would print a message similar to this: @example prog.c:5: assertion failed: a <= 5 && b >= 17 @end example @findex assert The ANSI C language makes it possible to turn the condition into a string for use in printing the diagnostic message. This is not possible in @code{awk}, so this @code{assert} function also requires a string version of the condition that is being tested. @example @c @group @c file eg/lib/assert.awk # assert --- assert that a condition is true. Otherwise exit. # Arnold Robbins, arnold@@gnu.org, Public Domain # May, 1993 function assert(condition, string) @{ if (! condition) @{ printf("%s:%d: assertion failed: %s\n", FILENAME, FNR, string) > "/dev/stderr" _assert_exit = 1 exit 1 @} @} END @{ if (_assert_exit) exit 1 @} @c endfile @c @end group @end example The @code{assert} function tests the @code{condition} parameter. If it is false, it prints a message to standard error, using the @code{string} parameter to describe the failed condition. It then sets the variable @code{_assert_exit} to one, and executes the @code{exit} statement. The @code{exit} statement jumps to the @code{END} rule. If the @code{END} rules finds @code{_assert_exit} to be true, then it exits immediately. The purpose of the @code{END} rule with its test is to keep any other @code{END} rules from running. When an assertion fails, the program should exit immediately. If no assertions fail, then @code{_assert_exit} will still be false when the @code{END} rule is run normally, and the rest of the program's @code{END} rules will execute. For all of this to work correctly, @file{assert.awk} must be the first source file read by @code{awk}. You would use this function in your programs this way: @example function myfunc(a, b) @{ assert(a <= 5 && b >= 17, "a <= 5 && b >= 17") @dots{} @} @end example @noindent If the assertion failed, you would see a message like this: @example mydata:1357: assertion failed: a <= 5 && b >= 17 @end example There is a problem with this version of @code{assert}, that it may not be possible to work around with standard @code{awk}. An @code{END} rule is automatically added to the program calling @code{assert}. Normally, if a program consists of just a @code{BEGIN} rule, the input files and/or standard input are not read. However, now that the program has an @code{END} rule, @code{awk} will attempt to read the input data files, or standard input (@pxref{Using BEGIN/END, , Startup and Cleanup Actions}), most likely causing the program to hang, waiting for input. @node Round Function, Ordinal Functions, Assert Function, Library Functions @section Rounding Numbers @cindex rounding The way @code{printf} and @code{sprintf} (@pxref{Printf, , Using @code{printf} Statements for Fancier Printing}) do rounding will often depend upon the system's C @code{sprintf} subroutine. On many machines, @code{sprintf} rounding is ``unbiased,'' which means it doesn't always round a trailing @samp{.5} up, contrary to naive expectations. In unbiased rounding, @samp{.5} rounds to even, rather than always up, so 1.5 rounds to 2 but 4.5 rounds to 4. The result is that if you are using a format that does rounding (e.g., @code{"%.0f"}) you should check what your system does. The following function does traditional rounding; it might be useful if your awk's @code{printf} does unbiased rounding. @findex round @example @c file eg/lib/round.awk # round --- do normal rounding # # Arnold Robbins, arnold@@gnu.org, August, 1996 # Public Domain function round(x, ival, aval, fraction) @{ ival = int(x) # integer part, int() truncates # see if fractional part if (ival == x) # no fraction return x if (x < 0) @{ aval = -x # absolute value ival = int(aval) fraction = aval - ival if (fraction >= .5) return int(x) - 1 # -2.5 --> -3 else return int(x) # -2.3 --> -2 @} else @{ fraction = x - ival if (fraction >= .5) return ival + 1 else return ival @} @} # test harness @{ print $0, round($0) @} @c endfile @end example @node Ordinal Functions, Join Function, Round Function, Library Functions @section Translating Between Characters and Numbers @cindex numeric character values @cindex values of characters as numbers One commercial implementation of @code{awk} supplies a built-in function, @code{ord}, which takes a character and returns the numeric value for that character in the machine's character set. If the string passed to @code{ord} has more than one character, only the first one is used. The inverse of this function is @code{chr} (from the function of the same name in Pascal), which takes a number and returns the corresponding character. Both functions can be written very nicely in @code{awk}; there is no real reason to build them into the @code{awk} interpreter. @findex ord @findex chr @example @group @c file eg/lib/ord.awk # ord.awk --- do ord and chr # # Global identifiers: # _ord_: numerical values indexed by characters # _ord_init: function to initialize _ord_ # # Arnold Robbins # arnold@@gnu.org # Public Domain # 16 January, 1992 # 20 July, 1992, revised BEGIN @{ _ord_init() @} @c endfile @end group @c @group @c file eg/lib/ord.awk function _ord_init( low, high, i, t) @{ low = sprintf("%c", 7) # BEL is ascii 7 if (low == "\a") @{ # regular ascii low = 0 high = 127 @} else if (sprintf("%c", 128 + 7) == "\a") @{ # ascii, mark parity low = 128 high = 255 @} else @{ # ebcdic(!) low = 0 high = 255 @} for (i = low; i <= high; i++) @{ t = sprintf("%c", i) _ord_[t] = i @} @} @c endfile @c @end group @end example @cindex character sets @cindex character encodings @cindex ASCII @cindex EBCDIC @cindex mark parity Some explanation of the numbers used by @code{chr} is worthwhile. The most prominent character set in use today is ASCII. Although an eight-bit byte can hold 256 distinct values (from zero to 255), ASCII only defines characters that use the values from zero to 127.@footnote{ASCII has been extended in many countries to use the values from 128 to 255 for country-specific characters. If your system uses these extensions, you can simplify @code{_ord_init} to simply loop from zero to 255.} At least one computer manufacturer that we know of @c Pr1me, blech uses ASCII, but with mark parity, meaning that the leftmost bit in the byte is always one. What this means is that on those systems, characters have numeric values from 128 to 255. Finally, large mainframe systems use the EBCDIC character set, which uses all 256 values. While there are other character sets in use on some older systems, they are not really worth worrying about. @example @group @c file eg/lib/ord.awk function ord(str, c) @{ # only first character is of interest c = substr(str, 1, 1) return _ord_[c] @} @c endfile @end group @group @c file eg/lib/ord.awk function chr(c) @{ # force c to be numeric by adding 0 return sprintf("%c", c + 0) @} @c endfile @end group @c @group @c file eg/lib/ord.awk #### test code #### # BEGIN \ # @{ # for (;;) @{ # printf("enter a character: ") # if (getline var <= 0) # break # printf("ord(%s) = %d\n", var, ord(var)) # @} # @} @c endfile @c @end group @end example An obvious improvement to these functions would be to move the code for the @code{@w{_ord_init}} function into the body of the @code{BEGIN} rule. It was written this way initially for ease of development. There is a ``test program'' in a @code{BEGIN} rule, for testing the function. It is commented out for production use. @node Join Function, Mktime Function, Ordinal Functions, Library Functions @section Merging an Array Into a String @cindex merging strings When doing string processing, it is often useful to be able to join all the strings in an array into one long string. The following function, @code{join}, accomplishes this task. It is used later in several of the application programs (@pxref{Sample Programs, ,Practical @code{awk} Programs}). Good function design is important; this function needs to be general, but it should also have a reasonable default behavior. It is called with an array and the beginning and ending indices of the elements in the array to be merged. This assumes that the array indices are numeric---a reasonable assumption since the array was likely created with @code{split} (@pxref{String Functions, ,Built-in Functions for String Manipulation}). @findex join @example @group @c file eg/lib/join.awk # join.awk --- join an array into a string # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 function join(array, start, end, sep, result, i) @{ if (sep == "") sep = " " else if (sep == SUBSEP) # magic value sep = "" result = array[start] for (i = start + 1; i <= end; i++) result = result sep array[i] return result @} @c endfile @end group @end example An optional additional argument is the separator to use when joining the strings back together. If the caller supplies a non-empty value, @code{join} uses it. If it is not supplied, it will have a null value. In this case, @code{join} uses a single blank as a default separator for the strings. If the value is equal to @code{SUBSEP}, then @code{join} joins the strings with no separator between them. @code{SUBSEP} serves as a ``magic'' value to indicate that there should be no separation between the component strings. It would be nice if @code{awk} had an assignment operator for concatenation. The lack of an explicit operator for concatenation makes string operations more difficult than they really need to be. @node Mktime Function, Gettimeofday Function, Join Function, Library Functions @section Turning Dates Into Timestamps The @code{systime} function built in to @code{gawk} returns the current time of day as a timestamp in ``seconds since the Epoch.'' This timestamp can be converted into a printable date of almost infinitely variable format using the built-in @code{strftime} function. (For more information on @code{systime} and @code{strftime}, @pxref{Time Functions, ,Functions for Dealing with Time Stamps}.) @cindex converting dates to timestamps @cindex dates, converting to timestamps @cindex timestamps, converting from dates An interesting but difficult problem is to convert a readable representation of a date back into a timestamp. The ANSI C library provides a @code{mktime} function that does the basic job, converting a canonical representation of a date into a timestamp. It would appear at first glance that @code{gawk} would have to supply a @code{mktime} built-in function that was simply a ``hook'' to the C language version. In fact though, @code{mktime} can be implemented entirely in @code{awk}. Here is a version of @code{mktime} for @code{awk}. It takes a simple representation of the date and time, and converts it into a timestamp. The code is presented here intermixed with explanatory prose. In @ref{Extract Program, ,Extracting Programs from Texinfo Source Files}, you will see how the Texinfo source file for this @value{DOCUMENT} can be processed to extract the code into a single source file. The program begins with a descriptive comment and a @code{BEGIN} rule that initializes a table @code{_tm_months}. This table is a two-dimensional array that has the lengths of the months. The first index is zero for regular years, and one for leap years. The values are the same for all the months in both kinds of years, except for February; thus the use of multiple assignment. @example @c @group @c file eg/lib/mktime.awk # mktime.awk --- convert a canonical date representation # into a timestamp # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 BEGIN \ @{ # Initialize table of month lengths _tm_months[0,1] = _tm_months[1,1] = 31 _tm_months[0,2] = 28; _tm_months[1,2] = 29 _tm_months[0,3] = _tm_months[1,3] = 31 _tm_months[0,4] = _tm_months[1,4] = 30 _tm_months[0,5] = _tm_months[1,5] = 31 _tm_months[0,6] = _tm_months[1,6] = 30 _tm_months[0,7] = _tm_months[1,7] = 31 _tm_months[0,8] = _tm_months[1,8] = 31 _tm_months[0,9] = _tm_months[1,9] = 30 _tm_months[0,10] = _tm_months[1,10] = 31 _tm_months[0,11] = _tm_months[1,11] = 30 _tm_months[0,12] = _tm_months[1,12] = 31 @} @c endfile @c @end group @end example The benefit of merging multiple @code{BEGIN} rules (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}) is particularly clear when writing library files. Functions in library files can cleanly initialize their own private data and also provide clean-up actions in private @code{END} rules. The next function is a simple one that computes whether a given year is or is not a leap year. If a year is evenly divisible by four, but not evenly divisible by 100, or if it is evenly divisible by 400, then it is a leap year. Thus, 1904 was a leap year, 1900 was not, but 2000 will be. @c Change this after the year 2000 to ``2000 was'' (:-) @findex _tm_isleap @example @group @c file eg/lib/mktime.awk # decide if a year is a leap year function _tm_isleap(year, ret) @{ ret = (year % 4 == 0 && year % 100 != 0) || (year % 400 == 0) return ret @} @c endfile @end group @end example This function is only used a few times in this file, and its computation could have been written @dfn{in-line} (at the point where it's used). Making it a separate function made the original development easier, and also avoids the possibility of typing errors when duplicating the code in multiple places. The next function is more interesting. It does most of the work of generating a timestamp, which is converting a date and time into some number of seconds since the Epoch. The caller passes an array (rather imaginatively named @code{a}) containing six values: the year including century, the month as a number between one and 12, the day of the month, the hour as a number between zero and 23, the minute in the hour, and the seconds within the minute. The function uses several local variables to precompute the number of seconds in an hour, seconds in a day, and seconds in a year. Often, similar C code simply writes out the expression in-line, expecting the compiler to do @dfn{constant folding}. E.g., most C compilers would turn @samp{60 * 60} into @samp{3600} at compile time, instead of recomputing it every time at run time. Precomputing these values makes the function more efficient. @findex _tm_addup @example @c @group @c file eg/lib/mktime.awk # convert a date into seconds function _tm_addup(a, total, yearsecs, daysecs, hoursecs, i, j) @{ hoursecs = 60 * 60 daysecs = 24 * hoursecs yearsecs = 365 * daysecs total = (a[1] - 1970) * yearsecs @group # extra day for leap years for (i = 1970; i < a[1]; i++) if (_tm_isleap(i)) total += daysecs @end group @group j = _tm_isleap(a[1]) for (i = 1; i < a[2]; i++) total += _tm_months[j, i] * daysecs @end group total += (a[3] - 1) * daysecs total += a[4] * hoursecs total += a[5] * 60 total += a[6] return total @} @c endfile @c @end group @end example The function starts with a first approximation of all the seconds between Midnight, January 1, 1970,@footnote{This is the Epoch on POSIX systems. It may be different on other systems.} and the beginning of the current year. It then goes through all those years, and for every leap year, adds an additional day's worth of seconds. The variable @code{j} holds either one or zero, if the current year is or is not a leap year. For every month in the current year prior to the current month, it adds the number of seconds in the month, using the appropriate entry in the @code{_tm_months} array. Finally, it adds in the seconds for the number of days prior to the current day, and the number of hours, minutes, and seconds in the current day. The result is a count of seconds since January 1, 1970. This value is not yet what is needed though. The reason why is described shortly. The main @code{mktime} function takes a single character string argument. This string is a representation of a date and time in a ``canonical'' (fixed) form. This string should be @code{"@var{year} @var{month} @var{day} @var{hour} @var{minute} @var{second}"}. @findex mktime @example @c @group @c file eg/lib/mktime.awk # mktime --- convert a date into seconds, # compensate for time zone function mktime(str, res1, res2, a, b, i, j, t, diff) @{ i = split(str, a, " ") # don't rely on FS if (i != 6) return -1 # force numeric for (j in a) a[j] += 0 @group # validate if (a[1] < 1970 || a[2] < 1 || a[2] > 12 || a[3] < 1 || a[3] > 31 || a[4] < 0 || a[4] > 23 || a[5] < 0 || a[5] > 59 || a[6] < 0 || a[6] > 60 ) return -1 @end group res1 = _tm_addup(a) t = strftime("%Y %m %d %H %M %S", res1) if (_tm_debug) printf("(%s) -> (%s)\n", str, t) > "/dev/stderr" split(t, b, " ") res2 = _tm_addup(b) diff = res1 - res2 if (_tm_debug) printf("diff = %d seconds\n", diff) > "/dev/stderr" res1 += diff return res1 @} @c endfile @c @end group @end example The function first splits the string into an array, using spaces and tabs as separators. If there are not six elements in the array, it returns an error, signaled as the value @minus{}1. Next, it forces each element of the array to be numeric, by adding zero to it. The following @samp{if} statement then makes sure that each element is within an allowable range. (This checking could be extended further, e.g., to make sure that the day of the month is within the correct range for the particular month supplied.) All of this is essentially preliminary set-up and error checking. Recall that @code{_tm_addup} generated a value in seconds since Midnight, January 1, 1970. This value is not directly usable as the result we want, @emph{since the calculation does not account for the local timezone}. In other words, the value represents the count in seconds since the Epoch, but only for UTC (Universal Coordinated Time). If the local timezone is east or west of UTC, then some number of hours should be either added to, or subtracted from the resulting timestamp. For example, 6:23 p.m. in Atlanta, Georgia (USA), is normally five hours west of (behind) UTC. It is only four hours behind UTC if daylight savings time is in effect. If you are calling @code{mktime} in Atlanta, with the argument @code{@w{"1993 5 23 18 23 12"}}, the result from @code{_tm_addup} will be for 6:23 p.m. UTC, which is only 2:23 p.m. in Atlanta. It is necessary to add another four hours worth of seconds to the result. How can @code{mktime} determine how far away it is from UTC? This is surprisingly easy. The returned timestamp represents the time passed to @code{mktime} @emph{as UTC}. This timestamp can be fed back to @code{strftime}, which will format it as a @emph{local} time; i.e.@: as if it already had the UTC difference added in to it. This is done by giving @code{@w{"%Y %m %d %H %M %S"}} to @code{strftime} as the format argument. It returns the computed timestamp in the original string format. The result represents a time that accounts for the UTC difference. When the new time is converted back to a timestamp, the difference between the two timestamps is the difference (in seconds) between the local timezone and UTC. This difference is then added back to the original result. An example demonstrating this is presented below. Finally, there is a ``main'' program for testing the function. @example @c @group @c file eg/lib/mktime.awk BEGIN @{ if (_tm_test) @{ printf "Enter date as yyyy mm dd hh mm ss: " getline _tm_test_date t = mktime(_tm_test_date) r = strftime("%Y %m %d %H %M %S", t) printf "Got back (%s)\n", r @} @} @c endfile @c @end group @end example The entire program uses two variables that can be set on the command line to control debugging output and to enable the test in the final @code{BEGIN} rule. Here is the result of a test run. (Note that debugging output is to standard error, and test output is to standard output.) @example @c @group $ gawk -f mktime.awk -v _tm_test=1 -v _tm_debug=1 @print{} Enter date as yyyy mm dd hh mm ss: 1993 5 23 15 35 10 @error{} (1993 5 23 15 35 10) -> (1993 05 23 11 35 10) @error{} diff = 14400 seconds @print{} Got back (1993 05 23 15 35 10) @c @end group @end example The time entered was 3:35 p.m. (15:35 on a 24-hour clock), on May 23, 1993. The first line of debugging output shows the resulting time as UTC---four hours ahead of the local time zone. The second line shows that the difference is 14400 seconds, which is four hours. (The difference is only four hours, since daylight savings time is in effect during May.) The final line of test output shows that the timezone compensation algorithm works; the returned time is the same as the entered time. This program does not solve the general problem of turning an arbitrary date representation into a timestamp. That problem is very involved. However, the @code{mktime} function provides a foundation upon which to build. Other software can convert month names into numeric months, and AM/PM times into 24-hour clocks, to generate the ``canonical'' format that @code{mktime} requires. @node Gettimeofday Function, Filetrans Function, Mktime Function, Library Functions @section Managing the Time of Day @cindex formatted timestamps @cindex timestamps, formatted The @code{systime} and @code{strftime} functions described in @ref{Time Functions, ,Functions for Dealing with Time Stamps}, provide the minimum functionality necessary for dealing with the time of day in human readable form. While @code{strftime} is extensive, the control formats are not necessarily easy to remember or intuitively obvious when reading a program. The following function, @code{gettimeofday}, populates a user-supplied array with pre-formatted time information. It returns a string with the current time formatted in the same way as the @code{date} utility. @findex gettimeofday @example @c @group @c file eg/lib/gettime.awk # gettimeofday --- get the time of day in a usable format # Arnold Robbins, arnold@@gnu.org, Public Domain, May 1993 # # Returns a string in the format of output of date(1) # Populates the array argument time with individual values: # time["second"] -- seconds (0 - 59) # time["minute"] -- minutes (0 - 59) # time["hour"] -- hours (0 - 23) # time["althour"] -- hours (0 - 12) # time["monthday"] -- day of month (1 - 31) # time["month"] -- month of year (1 - 12) # time["monthname"] -- name of the month # time["shortmonth"] -- short name of the month # time["year"] -- year within century (0 - 99) # time["fullyear"] -- year with century (19xx or 20xx) # time["weekday"] -- day of week (Sunday = 0) # time["altweekday"] -- day of week (Monday = 0) # time["weeknum"] -- week number, Sunday first day # time["altweeknum"] -- week number, Monday first day # time["dayname"] -- name of weekday # time["shortdayname"] -- short name of weekday # time["yearday"] -- day of year (0 - 365) # time["timezone"] -- abbreviation of timezone name # time["ampm"] -- AM or PM designation @group function gettimeofday(time, ret, now, i) @{ # get time once, avoids unnecessary system calls now = systime() # return date(1)-style output ret = strftime("%a %b %d %H:%M:%S %Z %Y", now) # clear out target array for (i in time) delete time[i] @end group @group # fill in values, force numeric values to be # numeric by adding 0 time["second"] = strftime("%S", now) + 0 time["minute"] = strftime("%M", now) + 0 time["hour"] = strftime("%H", now) + 0 time["althour"] = strftime("%I", now) + 0 time["monthday"] = strftime("%d", now) + 0 time["month"] = strftime("%m", now) + 0 time["monthname"] = strftime("%B", now) time["shortmonth"] = strftime("%b", now) time["year"] = strftime("%y", now) + 0 time["fullyear"] = strftime("%Y", now) + 0 time["weekday"] = strftime("%w", now) + 0 time["altweekday"] = strftime("%u", now) + 0 time["dayname"] = strftime("%A", now) time["shortdayname"] = strftime("%a", now) time["yearday"] = strftime("%j", now) + 0 time["timezone"] = strftime("%Z", now) time["ampm"] = strftime("%p", now) time["weeknum"] = strftime("%U", now) + 0 time["altweeknum"] = strftime("%W", now) + 0 return ret @} @end group @c endfile @end example The string indices are easier to use and read than the various formats required by @code{strftime}. The @code{alarm} program presented in @ref{Alarm Program, ,An Alarm Clock Program}, uses this function. @c exercise!!! The @code{gettimeofday} function is presented above as it was written. A more general design for this function would have allowed the user to supply an optional timestamp value that would have been used instead of the current time. @node Filetrans Function, Getopt Function, Gettimeofday Function, Library Functions @section Noting Data File Boundaries @cindex per file initialization and clean-up The @code{BEGIN} and @code{END} rules are each executed exactly once, at the beginning and end respectively of your @code{awk} program (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}). We (the @code{gawk} authors) once had a user who mistakenly thought that the @code{BEGIN} rule was executed at the beginning of each data file and the @code{END} rule was executed at the end of each data file. When informed that this was not the case, the user requested that we add new special patterns to @code{gawk}, named @code{BEGIN_FILE} and @code{END_FILE}, that would have the desired behavior. He even supplied us the code to do so. However, after a little thought, I came up with the following library program. It arranges to call two user-supplied functions, @code{beginfile} and @code{endfile}, at the beginning and end of each data file. Besides solving the problem in only nine(!) lines of code, it does so @emph{portably}; this will work with any implementation of @code{awk}. @example @c @group # transfile.awk # # Give the user a hook for filename transitions # # The user must supply functions beginfile() and endfile() # that each take the name of the file being started or # finished, respectively. # # Arnold Robbins, arnold@@gnu.org, January 1992 # Public Domain FILENAME != _oldfilename \ @{ if (_oldfilename != "") endfile(_oldfilename) _oldfilename = FILENAME beginfile(FILENAME) @} END @{ endfile(FILENAME) @} @c @end group @end example This file must be loaded before the user's ``main'' program, so that the rule it supplies will be executed first. This rule relies on @code{awk}'s @code{FILENAME} variable that automatically changes for each new data file. The current file name is saved in a private variable, @code{_oldfilename}. If @code{FILENAME} does not equal @code{_oldfilename}, then a new data file is being processed, and it is necessary to call @code{endfile} for the old file. Since @code{endfile} should only be called if a file has been processed, the program first checks to make sure that @code{_oldfilename} is not the null string. The program then assigns the current file name to @code{_oldfilename}, and calls @code{beginfile} for the file. Since, like all @code{awk} variables, @code{_oldfilename} will be initialized to the null string, this rule executes correctly even for the first data file. The program also supplies an @code{END} rule, to do the final processing for the last file. Since this @code{END} rule comes before any @code{END} rules supplied in the ``main'' program, @code{endfile} will be called first. Once again the value of multiple @code{BEGIN} and @code{END} rules should be clear. @findex beginfile @findex endfile This version has same problem as the first version of @code{nextfile} (@pxref{Nextfile Function, ,Implementing @code{nextfile} as a Function}). If the same data file occurs twice in a row on command line, then @code{endfile} and @code{beginfile} will not be executed at the end of the first pass and at the beginning of the second pass. This version solves the problem. @example @c @group @c file eg/lib/ftrans.awk # ftrans.awk --- handle data file transitions # # user supplies beginfile() and endfile() functions # # Arnold Robbins, arnold@@gnu.org, November 1992 # Public Domain FNR == 1 @{ if (_filename_ != "") endfile(_filename_) _filename_ = FILENAME beginfile(FILENAME) @} END @{ endfile(_filename_) @} @c endfile @c @end group @end example In @ref{Wc Program, ,Counting Things}, you will see how this library function can be used, and how it simplifies writing the main program. @node Getopt Function, Passwd Functions, Filetrans Function, Library Functions @section Processing Command Line Options @cindex @code{getopt}, C version @cindex processing arguments @cindex argument processing Most utilities on POSIX compatible systems take options or ``switches'' on the command line that can be used to change the way a program behaves. @code{awk} is an example of such a program (@pxref{Options, ,Command Line Options}). Often, options take @dfn{arguments}, data that the program needs to correctly obey the command line option. For example, @code{awk}'s @samp{-F} option requires a string to use as the field separator. The first occurrence on the command line of either @samp{--} or a string that does not begin with @samp{-} ends the options. Most Unix systems provide a C function named @code{getopt} for processing command line arguments. The programmer provides a string describing the one letter options. If an option requires an argument, it is followed in the string with a colon. @code{getopt} is also passed the count and values of the command line arguments, and is called in a loop. @code{getopt} processes the command line arguments for option letters. Each time around the loop, it returns a single character representing the next option letter that it found, or @samp{?} if it found an invalid option. When it returns @minus{}1, there are no options left on the command line. When using @code{getopt}, options that do not take arguments can be grouped together. Furthermore, options that take arguments require that the argument be present. The argument can immediately follow the option letter, or it can be a separate command line argument. Given a hypothetical program that takes three command line options, @samp{-a}, @samp{-b}, and @samp{-c}, and @samp{-b} requires an argument, all of the following are valid ways of invoking the program: @example @c @group prog -a -b foo -c data1 data2 data3 prog -ac -bfoo -- data1 data2 data3 prog -acbfoo data1 data2 data3 @c @end group @end example Notice that when the argument is grouped with its option, the rest of the command line argument is considered to be the option's argument. In the above example, @samp{-acbfoo} indicates that all of the @samp{-a}, @samp{-b}, and @samp{-c} options were supplied, and that @samp{foo} is the argument to the @samp{-b} option. @code{getopt} provides four external variables that the programmer can use. @table @code @item optind The index in the argument value array (@code{argv}) where the first non-option command line argument can be found. @item optarg The string value of the argument to an option. @item opterr Usually @code{getopt} prints an error message when it finds an invalid option. Setting @code{opterr} to zero disables this feature. (An application might wish to print its own error message.) @item optopt The letter representing the command line option. While not usually documented, most versions supply this variable. @end table The following C fragment shows how @code{getopt} might process command line arguments for @code{awk}. @example @group int main(int argc, char *argv[]) @{ @dots{} /* print our own message */ opterr = 0; @end group @group while ((c = getopt(argc, argv, "v:f:F:W:")) != -1) @{ switch (c) @{ case 'f': /* file */ @dots{} break; case 'F': /* field separator */ @dots{} break; case 'v': /* variable assignment */ @dots{} break; case 'W': /* extension */ @dots{} break; case '?': default: usage(); break; @} @} @dots{} @} @end group @end example As a side point, @code{gawk} actually uses the GNU @code{getopt_long} function to process both normal and GNU-style long options (@pxref{Options, ,Command Line Options}). The abstraction provided by @code{getopt} is very useful, and would be quite handy in @code{awk} programs as well. Here is an @code{awk} version of @code{getopt}. This function highlights one of the greatest weaknesses in @code{awk}, which is that it is very poor at manipulating single characters. Repeated calls to @code{substr} are necessary for accessing individual characters (@pxref{String Functions, ,Built-in Functions for String Manipulation}). The discussion walks through the code a bit at a time. @example @c @group @c file eg/lib/getopt.awk # getopt --- do C library getopt(3) function in awk # # arnold@@gnu.org # Public domain # # Initial version: March, 1991 # Revised: May, 1993 @group # External variables: # Optind -- index of ARGV for first non-option argument # Optarg -- string value of argument to current option # Opterr -- if non-zero, print our own diagnostic # Optopt -- current option letter @end group # Returns # -1 at end of options # ? for unrecognized option # a character representing the current option # Private Data # _opti index in multi-flag option, e.g., -abc @c endfile @c @end group @end example The function starts out with some documentation: who wrote the code, and when it was revised, followed by a list of the global variables it uses, what the return values are and what they mean, and any global variables that are ``private'' to this library function. Such documentation is essential for any program, and particularly for library functions. @findex getopt @example @c @group @c file eg/lib/getopt.awk function getopt(argc, argv, options, optl, thisopt, i) @{ optl = length(options) if (optl == 0) # no options given return -1 if (argv[Optind] == "--") @{ # all done Optind++ _opti = 0 return -1 @} else if (argv[Optind] !~ /^-[^: \t\n\f\r\v\b]/) @{ _opti = 0 return -1 @} @c endfile @c @end group @end example The function first checks that it was indeed called with a string of options (the @code{options} parameter). If @code{options} has a zero length, @code{getopt} immediately returns @minus{}1. The next thing to check for is the end of the options. A @samp{--} ends the command line options, as does any command line argument that does not begin with a @samp{-}. @code{Optind} is used to step through the array of command line arguments; it retains its value across calls to @code{getopt}, since it is a global variable. The regexp used, @code{@w{/^-[^: \t\n\f\r\v\b]/}}, is perhaps a bit of overkill; it checks for a @samp{-} followed by anything that is not whitespace and not a colon. If the current command line argument does not match this pattern, it is not an option, and it ends option processing. @example @group @c file eg/lib/getopt.awk if (_opti == 0) _opti = 2 thisopt = substr(argv[Optind], _opti, 1) Optopt = thisopt i = index(options, thisopt) if (i == 0) @{ if (Opterr) printf("%c -- invalid option\n", thisopt) > "/dev/stderr" if (_opti >= length(argv[Optind])) @{ Optind++ _opti = 0 @} else _opti++ return "?" @} @c endfile @end group @end example The @code{_opti} variable tracks the position in the current command line argument (@code{argv[Optind]}). In the case that multiple options were grouped together with one @samp{-} (e.g., @samp{-abx}), it is necessary to return them to the user one at a time. If @code{_opti} is equal to zero, it is set to two, the index in the string of the next character to look at (we skip the @samp{-}, which is at position one). The variable @code{thisopt} holds the character, obtained with @code{substr}. It is saved in @code{Optopt} for the main program to use. If @code{thisopt} is not in the @code{options} string, then it is an invalid option. If @code{Opterr} is non-zero, @code{getopt} prints an error message on the standard error that is similar to the message from the C version of @code{getopt}. Since the option is invalid, it is necessary to skip it and move on to the next option character. If @code{_opti} is greater than or equal to the length of the current command line argument, then it is necessary to move on to the next one, so @code{Optind} is incremented and @code{_opti} is reset to zero. Otherwise, @code{Optind} is left alone and @code{_opti} is merely incremented. In any case, since the option was invalid, @code{getopt} returns @samp{?}. The main program can examine @code{Optopt} if it needs to know what the invalid option letter actually was. @example @group @c file eg/lib/getopt.awk if (substr(options, i + 1, 1) == ":") @{ # get option argument if (length(substr(argv[Optind], _opti + 1)) > 0) Optarg = substr(argv[Optind], _opti + 1) else Optarg = argv[++Optind] _opti = 0 @} else Optarg = "" @c endfile @end group @end example If the option requires an argument, the option letter is followed by a colon in the @code{options} string. If there are remaining characters in the current command line argument (@code{argv[Optind]}), then the rest of that string is assigned to @code{Optarg}. Otherwise, the next command line argument is used (@samp{-xFOO} vs. @samp{@w{-x FOO}}). In either case, @code{_opti} is reset to zero, since there are no more characters left to examine in the current command line argument. @example @c @group @c file eg/lib/getopt.awk if (_opti == 0 || _opti >= length(argv[Optind])) @{ Optind++ _opti = 0 @} else _opti++ return thisopt @} @c endfile @c @end group @end example Finally, if @code{_opti} is either zero or greater than the length of the current command line argument, it means this element in @code{argv} is through being processed, so @code{Optind} is incremented to point to the next element in @code{argv}. If neither condition is true, then only @code{_opti} is incremented, so that the next option letter can be processed on the next call to @code{getopt}. @example @c @group @c file eg/lib/getopt.awk BEGIN @{ Opterr = 1 # default is to diagnose Optind = 1 # skip ARGV[0] # test program if (_getopt_test) @{ while ((_go_c = getopt(ARGC, ARGV, "ab:cd")) != -1) printf("c = <%c>, optarg = <%s>\n", _go_c, Optarg) printf("non-option arguments:\n") for (; Optind < ARGC; Optind++) printf("\tARGV[%d] = <%s>\n", Optind, ARGV[Optind]) @} @} @c endfile @c @end group @end example The @code{BEGIN} rule initializes both @code{Opterr} and @code{Optind} to one. @code{Opterr} is set to one, since the default behavior is for @code{getopt} to print a diagnostic message upon seeing an invalid option. @code{Optind} is set to one, since there's no reason to look at the program name, which is in @code{ARGV[0]}. The rest of the @code{BEGIN} rule is a simple test program. Here is the result of two sample runs of the test program. @example @group $ awk -f getopt.awk -v _getopt_test=1 -- -a -cbARG bax -x @print{} c = , optarg = <> @print{} c = , optarg = <> @print{} c = , optarg = @print{} non-option arguments: @print{} ARGV[3] = @print{} ARGV[4] = <-x> @end group @group $ awk -f getopt.awk -v _getopt_test=1 -- -a -x -- xyz abc @print{} c = , optarg = <> @error{} x -- invalid option @print{} c = , optarg = <> @print{} non-option arguments: @print{} ARGV[4] = @print{} ARGV[5] = @end group @end example The first @samp{--} terminates the arguments to @code{awk}, so that it does not try to interpret the @samp{-a} etc. as its own options. Several of the sample programs presented in @ref{Sample Programs, ,Practical @code{awk} Programs}, use @code{getopt} to process their arguments. @node Passwd Functions, Group Functions, Getopt Function, Library Functions @section Reading the User Database @cindex @file{/dev/user} The @file{/dev/user} special file (@pxref{Special Files, ,Special File Names in @code{gawk}}) provides access to the current user's real and effective user and group id numbers, and if available, the user's supplementary group set. However, since these are numbers, they do not provide very useful information to the average user. There needs to be some way to find the user information associated with the user and group numbers. This section presents a suite of functions for retrieving information from the user database. @xref{Group Functions, ,Reading the Group Database}, for a similar suite that retrieves information from the group database. @cindex @code{getpwent}, C version @cindex user information @cindex login information @cindex account information @cindex password file The POSIX standard does not define the file where user information is kept. Instead, it provides the @code{} header file and several C language subroutines for obtaining user information. The primary function is @code{getpwent}, for ``get password entry.'' The ``password'' comes from the original user database file, @file{/etc/passwd}, which kept user information, along with the encrypted passwords (hence the name). While an @code{awk} program could simply read @file{/etc/passwd} directly (the format is well known), because of the way password files are handled on networked systems, this file may not contain complete information about the system's set of users. @cindex @code{pwcat} program To be sure of being able to produce a readable, complete version of the user database, it is necessary to write a small C program that calls @code{getpwent}. @code{getpwent} is defined to return a pointer to a @code{struct passwd}. Each time it is called, it returns the next entry in the database. When there are no more entries, it returns @code{NULL}, the null pointer. When this happens, the C program should call @code{endpwent} to close the database. Here is @code{pwcat}, a C program that ``cats'' the password database. @findex pwcat.c @example @c @group @c file eg/lib/pwcat.c /* * pwcat.c * * Generate a printable version of the password database * * Arnold Robbins * arnold@@gnu.org * May 1993 * Public Domain */ #include #include int main(argc, argv) int argc; char **argv; @{ struct passwd *p; while ((p = getpwent()) != NULL) printf("%s:%s:%d:%d:%s:%s:%s\n", p->pw_name, p->pw_passwd, p->pw_uid, p->pw_gid, p->pw_gecos, p->pw_dir, p->pw_shell); endpwent(); exit(0); @} @c endfile @c @end group @end example If you don't understand C, don't worry about it. The output from @code{pwcat} is the user database, in the traditional @file{/etc/passwd} format of colon-separated fields. The fields are: @table @asis @item Login name The user's login name. @item Encrypted password The user's encrypted password. This may not be available on some systems. @item User-ID The user's numeric user-id number. @item Group-ID The user's numeric group-id number. @item Full name The user's full name, and perhaps other information associated with the user. @item Home directory The user's login, or ``home'' directory (familiar to shell programmers as @code{$HOME}). @item Login shell The program that will be run when the user logs in. This is usually a shell, such as Bash (the Gnu Bourne-Again shell). @end table Here are a few lines representative of @code{pwcat}'s output. @example @c @group $ pwcat @print{} root:3Ov02d5VaUPB6:0:1:Operator:/:/bin/sh @print{} nobody:*:65534:65534::/: @print{} daemon:*:1:1::/: @print{} sys:*:2:2::/:/bin/csh @print{} bin:*:3:3::/bin: @print{} arnold:xyzzy:2076:10:Arnold Robbins:/home/arnold:/bin/sh @print{} miriam:yxaay:112:10:Miriam Robbins:/home/miriam:/bin/sh @print{} andy:abcca2:113:10:Andy Jacobs:/home/andy:/bin/sh @dots{} @c @end group @end example With that introduction, here is a group of functions for getting user information. There are several functions here, corresponding to the C functions of the same name. @findex _pw_init @example @c file eg/lib/passwdawk.in @group # passwd.awk --- access password file information # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 BEGIN @{ # tailor this to suit your system _pw_awklib = "/usr/local/libexec/awk/" @} @end group @group function _pw_init( oldfs, oldrs, olddol0, pwcat) @{ if (_pw_inited) return oldfs = FS oldrs = RS olddol0 = $0 FS = ":" RS = "\n" pwcat = _pw_awklib "pwcat" while ((pwcat | getline) > 0) @{ _pw_byname[$1] = $0 _pw_byuid[$3] = $0 _pw_bycount[++_pw_total] = $0 @} close(pwcat) _pw_count = 0 _pw_inited = 1 FS = oldfs RS = oldrs $0 = olddol0 @} @c endfile @end group @end example The @code{BEGIN} rule sets a private variable to the directory where @code{pwcat} is stored. Since it is used to help out an @code{awk} library routine, we have chosen to put it in @file{/usr/local/libexec/awk}. You might want it to be in a different directory on your system. The function @code{_pw_init} keeps three copies of the user information in three associative arrays. The arrays are indexed by user name (@code{_pw_byname}), by user-id number (@code{_pw_byuid}), and by order of occurrence (@code{_pw_bycount}). The variable @code{_pw_inited} is used for efficiency; @code{_pw_init} only needs to be called once. Since this function uses @code{getline} to read information from @code{pwcat}, it first saves the values of @code{FS}, @code{RS}, and @code{$0}. Doing so is necessary, since these functions could be called from anywhere within a user's program, and the user may have his or her own values for @code{FS} and @code{RS}. @ignore Problem, what if FIELDWIDTHS is in use? Sigh. @end ignore The main part of the function uses a loop to read database lines, split the line into fields, and then store the line into each array as necessary. When the loop is done, @code{@w{_pw_init}} cleans up by closing the pipeline, setting @code{@w{_pw_inited}} to one, and restoring @code{FS}, @code{RS}, and @code{$0}. The use of @code{@w{_pw_count}} will be explained below. @findex getpwnam @example @group @c file eg/lib/passwdawk.in function getpwnam(name) @{ _pw_init() if (name in _pw_byname) return _pw_byname[name] return "" @} @c endfile @end group @end example The @code{getpwnam} function takes a user name as a string argument. If that user is in the database, it returns the appropriate line. Otherwise it returns the null string. @findex getpwuid @example @group @c file eg/lib/passwdawk.in function getpwuid(uid) @{ _pw_init() if (uid in _pw_byuid) return _pw_byuid[uid] return "" @} @c endfile @end group @end example Similarly, the @code{getpwuid} function takes a user-id number argument. If that user number is in the database, it returns the appropriate line. Otherwise it returns the null string. @findex getpwent @example @c @group @c file eg/lib/passwdawk.in function getpwent() @{ _pw_init() if (_pw_count < _pw_total) return _pw_bycount[++_pw_count] return "" @} @c endfile @c @end group @end example The @code{getpwent} function simply steps through the database, one entry at a time. It uses @code{_pw_count} to track its current position in the @code{_pw_bycount} array. @findex endpwent @example @c @group @c file eg/lib/passwdawk.in function endpwent() @{ _pw_count = 0 @} @c endfile @c @end group @end example The @code{@w{endpwent}} function resets @code{@w{_pw_count}} to zero, so that subsequent calls to @code{getpwent} will start over again. A conscious design decision in this suite is that each subroutine calls @code{@w{_pw_init}} to initialize the database arrays. The overhead of running a separate process to generate the user database, and the I/O to scan it, will only be incurred if the user's main program actually calls one of these functions. If this library file is loaded along with a user's program, but none of the routines are ever called, then there is no extra run-time overhead. (The alternative would be to move the body of @code{@w{_pw_init}} into a @code{BEGIN} rule, which would always run @code{pwcat}. This simplifies the code but runs an extra process that may never be needed.) In turn, calling @code{_pw_init} is not too expensive, since the @code{_pw_inited} variable keeps the program from reading the data more than once. If you are worried about squeezing every last cycle out of your @code{awk} program, the check of @code{_pw_inited} could be moved out of @code{_pw_init} and duplicated in all the other functions. In practice, this is not necessary, since most @code{awk} programs are I/O bound, and it would clutter up the code. The @code{id} program in @ref{Id Program, ,Printing Out User Information}, uses these functions. @node Group Functions, Library Names, Passwd Functions, Library Functions @section Reading the Group Database @cindex @code{getgrent}, C version @cindex group information @cindex account information @cindex group file Much of the discussion presented in @ref{Passwd Functions, ,Reading the User Database}, applies to the group database as well. Although there has traditionally been a well known file, @file{/etc/group}, in a well known format, the POSIX standard only provides a set of C library routines (@code{} and @code{getgrent}) for accessing the information. Even though this file may exist, it likely does not have complete information. Therefore, as with the user database, it is necessary to have a small C program that generates the group database as its output. @cindex @code{grcat} program Here is @code{grcat}, a C program that ``cats'' the group database. @findex grcat.c @example @c @group @c file eg/lib/grcat.c /* * grcat.c * * Generate a printable version of the group database * * Arnold Robbins, arnold@@gnu.org * May 1993 * Public Domain */ #include #include @group int main(argc, argv) int argc; char **argv; @{ struct group *g; int i; @end group while ((g = getgrent()) != NULL) @{ printf("%s:%s:%d:", g->gr_name, g->gr_passwd, g->gr_gid); for (i = 0; g->gr_mem[i] != NULL; i++) @{ printf("%s", g->gr_mem[i]); if (g->gr_mem[i+1] != NULL) putchar(','); @} putchar('\n'); @} endgrent(); exit(0); @} @c endfile @c @end group @end example Each line in the group database represent one group. The fields are separated with colons, and represent the following information. @table @asis @item Group Name The name of the group. @item Group Password The encrypted group password. In practice, this field is never used. It is usually empty, or set to @samp{*}. @item Group ID Number The numeric group-id number. This number should be unique within the file. @item Group Member List A comma-separated list of user names. These users are members of the group. Most Unix systems allow users to be members of several groups simultaneously. If your system does, then reading @file{/dev/user} will return those group-id numbers in @code{$5} through @code{$NF}. (Note that @file{/dev/user} is a @code{gawk} extension; @pxref{Special Files, ,Special File Names in @code{gawk}}.) @end table Here is what running @code{grcat} might produce: @example @group $ grcat @print{} wheel:*:0:arnold @print{} nogroup:*:65534: @print{} daemon:*:1: @print{} kmem:*:2: @print{} staff:*:10:arnold,miriam,andy @print{} other:*:20: @dots{} @end group @end example Here are the functions for obtaining information from the group database. There are several, modeled after the C library functions of the same names. @findex _gr_init @example @group @c file eg/lib/groupawk.in # group.awk --- functions for dealing with the group file # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 BEGIN \ @{ # Change to suit your system _gr_awklib = "/usr/local/libexec/awk/" @} @c endfile @end group @group @c file eg/lib/groupawk.in function _gr_init( oldfs, oldrs, olddol0, grcat, n, a, i) @{ if (_gr_inited) return @end group @group oldfs = FS oldrs = RS olddol0 = $0 FS = ":" RS = "\n" @end group @group grcat = _gr_awklib "grcat" while ((grcat | getline) > 0) @{ if ($1 in _gr_byname) _gr_byname[$1] = _gr_byname[$1] "," $4 else _gr_byname[$1] = $0 if ($3 in _gr_bygid) _gr_bygid[$3] = _gr_bygid[$3] "," $4 else _gr_bygid[$3] = $0 n = split($4, a, "[ \t]*,[ \t]*") @end group @group for (i = 1; i <= n; i++) if (a[i] in _gr_groupsbyuser) _gr_groupsbyuser[a[i]] = \ _gr_groupsbyuser[a[i]] " " $1 else _gr_groupsbyuser[a[i]] = $1 @end group @group _gr_bycount[++_gr_count] = $0 @} @end group @group close(grcat) _gr_count = 0 _gr_inited++ FS = oldfs RS = oldrs $0 = olddol0 @} @c endfile @end group @end example The @code{BEGIN} rule sets a private variable to the directory where @code{grcat} is stored. Since it is used to help out an @code{awk} library routine, we have chosen to put it in @file{/usr/local/libexec/awk}. You might want it to be in a different directory on your system. These routines follow the same general outline as the user database routines (@pxref{Passwd Functions, ,Reading the User Database}). The @code{@w{_gr_inited}} variable is used to ensure that the database is scanned no more than once. The @code{@w{_gr_init}} function first saves @code{FS}, @code{RS}, and @code{$0}, and then sets @code{FS} and @code{RS} to the correct values for scanning the group information. The group information is stored is several associative arrays. The arrays are indexed by group name (@code{@w{_gr_byname}}), by group-id number (@code{@w{_gr_bygid}}), and by position in the database (@code{@w{_gr_bycount}}). There is an additional array indexed by user name (@code{@w{_gr_groupsbyuser}}), that is a space separated list of groups that each user belongs to. Unlike the user database, it is possible to have multiple records in the database for the same group. This is common when a group has a large number of members. Such a pair of entries might look like: @example tvpeople:*:101:johny,jay,arsenio tvpeople:*:101:david,conan,tom,joan @end example For this reason, @code{_gr_init} looks to see if a group name or group-id number has already been seen. If it has, then the user names are simply concatenated onto the previous list of users. (There is actually a subtle problem with the code presented above. Suppose that the first time there were no names. This code adds the names with a leading comma. It also doesn't check that there is a @code{$4}.) Finally, @code{_gr_init} closes the pipeline to @code{grcat}, restores @code{FS}, @code{RS}, and @code{$0}, initializes @code{_gr_count} to zero (it is used later), and makes @code{_gr_inited} non-zero. @findex getgrnam @example @c @group @c file eg/lib/groupawk.in function getgrnam(group) @{ _gr_init() if (group in _gr_byname) return _gr_byname[group] return "" @} @c endfile @c @end group @end example The @code{getgrnam} function takes a group name as its argument, and if that group exists, it is returned. Otherwise, @code{getgrnam} returns the null string. @findex getgrgid @example @c @group @c file eg/lib/groupawk.in function getgrgid(gid) @{ _gr_init() if (gid in _gr_bygid) return _gr_bygid[gid] return "" @} @c endfile @c @end group @end example The @code{getgrgid} function is similar, it takes a numeric group-id, and looks up the information associated with that group-id. @findex getgruser @example @group @c file eg/lib/groupawk.in function getgruser(user) @{ _gr_init() if (user in _gr_groupsbyuser) return _gr_groupsbyuser[user] return "" @} @c endfile @end group @end example The @code{getgruser} function does not have a C counterpart. It takes a user name, and returns the list of groups that have the user as a member. @findex getgrent @example @c @group @c file eg/lib/groupawk.in function getgrent() @{ _gr_init() if (++_gr_count in _gr_bycount) return _gr_bycount[_gr_count] return "" @} @c endfile @c @end group @end example The @code{getgrent} function steps through the database one entry at a time. It uses @code{_gr_count} to track its position in the list. @findex endgrent @example @group @c file eg/lib/groupawk.in function endgrent() @{ _gr_count = 0 @} @c endfile @end group @end example @code{endgrent} resets @code{_gr_count} to zero so that @code{getgrent} can start over again. As with the user database routines, each function calls @code{_gr_init} to initialize the arrays. Doing so only incurs the extra overhead of running @code{grcat} if these functions are used (as opposed to moving the body of @code{_gr_init} into a @code{BEGIN} rule). Most of the work is in scanning the database and building the various associative arrays. The functions that the user calls are themselves very simple, relying on @code{awk}'s associative arrays to do work. The @code{id} program in @ref{Id Program, ,Printing Out User Information}, uses these functions. @node Library Names, , Group Functions, Library Functions @section Naming Library Function Global Variables @cindex namespace issues in @code{awk} @cindex documenting @code{awk} programs @cindex programs, documenting Due to the way the @code{awk} language evolved, variables are either @dfn{global} (usable by the entire program), or @dfn{local} (usable just by a specific function). There is no intermediate state analogous to @code{static} variables in C. Library functions often need to have global variables that they can use to preserve state information between calls to the function. For example, @code{getopt}'s variable @code{_opti} (@pxref{Getopt Function, ,Processing Command Line Options}), and the @code{_tm_months} array used by @code{mktime} (@pxref{Mktime Function, ,Turning Dates Into Timestamps}). Such variables are called @dfn{private}, since the only functions that need to use them are the ones in the library. When writing a library function, you should try to choose names for your private variables so that they will not conflict with any variables used by either another library function or a user's main program. For example, a name like @samp{i} or @samp{j} is not a good choice, since user programs often use variable names like these for their own purposes. The example programs shown in this chapter all start the names of their private variables with an underscore (@samp{_}). Users generally don't use leading underscores in their variable names, so this convention immediately decreases the chances that the variable name will be accidentally shared with the user's program. In addition, several of the library functions use a prefix that helps indicate what function or set of functions uses the variables. For example, @code{_tm_months} in @code{mktime} (@pxref{Mktime Function, ,Turning Dates Into Timestamps}), and @code{_pw_byname} in the user data base routines (@pxref{Passwd Functions, ,Reading the User Database}). This convention is recommended, since it even further decreases the chance of inadvertent conflict among variable names. Note that this convention can be used equally well both for variable names and for private function names too. While I could have re-written all the library routines to use this convention, I did not do so, in order to show how my own @code{awk} programming style has evolved, and to provide some basis for this discussion. As a final note on variable naming, if a function makes global variables available for use by a main program, it is a good convention to start that variable's name with a capital letter. For example, @code{getopt}'s @code{Opterr} and @code{Optind} variables (@pxref{Getopt Function, ,Processing Command Line Options}). The leading capital letter indicates that it is global, while the fact that the variable name is not all capital letters indicates that the variable is not one of @code{awk}'s built-in variables, like @code{FS}. It is also important that @emph{all} variables in library functions that do not need to save state are in fact declared local. If this is not done, the variable could accidentally be used in the user's program, leading to bugs that are very difficult to track down. @example function lib_func(x, y, l1, l2) @{ @dots{} @var{use variable} some_var # some_var could be local @dots{} # but is not by oversight @} @end example @cindex Tcl A different convention, common in the Tcl community, is to use a single associative array to hold the values needed by the library function(s), or ``package.'' This significantly decreases the number of actual global names in use. For example, the functions described in @ref{Passwd Functions, , Reading the User Database}, might have used @code{@w{PW_data["inited"]}}, @code{@w{PW_data["total"]}}, @code{@w{PW_data["count"]}} and @code{@w{PW_data["awklib"]}}, instead of @code{@w{_pw_inited}}, @code{@w{_pw_awklib}}, @code{@w{_pw_total}}, and @code{@w{_pw_count}}. The conventions presented in this section are exactly that, conventions. You are not required to write your programs this way, we merely recommend that you do so. @node Sample Programs, Language History, Library Functions, Top @chapter Practical @code{awk} Programs This chapter presents a potpourri of @code{awk} programs for your reading enjoyment. @iftex There are two sections. The first presents @code{awk} versions of several common POSIX utilities. The second is a grab-bag of interesting programs. @end iftex Many of these programs use the library functions presented in @ref{Library Functions, ,A Library of @code{awk} Functions}. @menu * Clones:: Clones of common utilities. * Miscellaneous Programs:: Some interesting @code{awk} programs. @end menu @node Clones, Miscellaneous Programs, Sample Programs, Sample Programs @section Re-inventing Wheels for Fun and Profit This section presents a number of POSIX utilities that are implemented in @code{awk}. Re-inventing these programs in @code{awk} is often enjoyable, since the algorithms can be very clearly expressed, and usually the code is very concise and simple. This is true because @code{awk} does so much for you. It should be noted that these programs are not necessarily intended to replace the installed versions on your system. Instead, their purpose is to illustrate @code{awk} language programming for ``real world'' tasks. The programs are presented in alphabetical order. @menu * Cut Program:: The @code{cut} utility. * Egrep Program:: The @code{egrep} utility. * Id Program:: The @code{id} utility. * Split Program:: The @code{split} utility. * Tee Program:: The @code{tee} utility. * Uniq Program:: The @code{uniq} utility. * Wc Program:: The @code{wc} utility. @end menu @node Cut Program, Egrep Program, Clones, Clones @subsection Cutting Out Fields and Columns @cindex @code{cut} utility The @code{cut} utility selects, or ``cuts,'' either characters or fields from its standard input and sends them to its standard output. @code{cut} can cut out either a list of characters, or a list of fields. By default, fields are separated by tabs, but you may supply a command line option to change the field @dfn{delimiter}, i.e.@: the field separator character. @code{cut}'s definition of fields is less general than @code{awk}'s. A common use of @code{cut} might be to pull out just the login name of logged-on users from the output of @code{who}. For example, the following pipeline generates a sorted, unique list of the logged on users: @example who | cut -c1-8 | sort | uniq @end example The options for @code{cut} are: @table @code @item -c @var{list} Use @var{list} as the list of characters to cut out. Items within the list may be separated by commas, and ranges of characters can be separated with dashes. The list @samp{1-8,15,22-35} specifies characters one through eight, 15, and 22 through 35. @item -f @var{list} Use @var{list} as the list of fields to cut out. @item -d @var{delim} Use @var{delim} as the field separator character instead of the tab character. @item -s Suppress printing of lines that do not contain the field delimiter. @end table The @code{awk} implementation of @code{cut} uses the @code{getopt} library function (@pxref{Getopt Function, ,Processing Command Line Options}), and the @code{join} library function (@pxref{Join Function, ,Merging an Array Into a String}). The program begins with a comment describing the options and a @code{usage} function which prints out a usage message and exits. @code{usage} is called if invalid arguments are supplied. @findex cut.awk @example @c @group @c file eg/prog/cut.awk # cut.awk --- implement cut in awk # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # Options: # -f list Cut fields # -d c Field delimiter character # -c list Cut characters # # -s Suppress lines without the delimiter character function usage( e1, e2) @{ e1 = "usage: cut [-f list] [-d c] [-s] [files...]" e2 = "usage: cut [-c list] [files...]" print e1 > "/dev/stderr" print e2 > "/dev/stderr" exit 1 @} @c endfile @c @end group @end example @noindent The variables @code{e1} and @code{e2} are used so that the function fits nicely on the @iftex page. @end iftex @ifinfo screen. @end ifinfo Next comes a @code{BEGIN} rule that parses the command line options. It sets @code{FS} to a single tab character, since that is @code{cut}'s default field separator. The output field separator is also set to be the same as the input field separator. Then @code{getopt} is used to step through the command line options. One or the other of the variables @code{by_fields} or @code{by_chars} is set to true, to indicate that processing should be done by fields or by characters respectively. When cutting by characters, the output field separator is set to the null string. @example @c @group @c file eg/prog/cut.awk BEGIN \ @{ FS = "\t" # default OFS = FS while ((c = getopt(ARGC, ARGV, "sf:c:d:")) != -1) @{ if (c == "f") @{ by_fields = 1 fieldlist = Optarg @group @} else if (c == "c") @{ by_chars = 1 fieldlist = Optarg OFS = "" @} else if (c == "d") @{ if (length(Optarg) > 1) @{ printf("Using first character of %s" \ " for delimiter\n", Optarg) > "/dev/stderr" Optarg = substr(Optarg, 1, 1) @} FS = Optarg OFS = FS if (FS == " ") # defeat awk semantics FS = "[ ]" @} else if (c == "s") suppress++ else usage() @} @end group for (i = 1; i < Optind; i++) ARGV[i] = "" @c endfile @c @end group @end example Special care is taken when the field delimiter is a space. Using @code{@w{" "}} (a single space) for the value of @code{FS} is incorrect---@code{awk} would separate fields with runs of spaces, tabs and/or newlines, and we want them to be separated with individual spaces. Also, note that after @code{getopt} is through, we have to clear out all the elements of @code{ARGV} from one to @code{Optind}, so that @code{awk} will not try to process the command line options as file names. After dealing with the command line options, the program verifies that the options make sense. Only one or the other of @samp{-c} and @samp{-f} should be used, and both require a field list. Then either @code{set_fieldlist} or @code{set_charlist} is called to pull apart the list of fields or characters. @example @c @group @c file eg/prog/cut.awk if (by_fields && by_chars) usage() if (by_fields == 0 && by_chars == 0) by_fields = 1 # default if (fieldlist == "") @{ print "cut: needs list for -c or -f" > "/dev/stderr" exit 1 @} @group if (by_fields) set_fieldlist() else set_charlist() @} @c endfile @end group @end example Here is @code{set_fieldlist}. It first splits the field list apart at the commas, into an array. Then, for each element of the array, it looks to see if it is actually a range, and if so splits it apart. The range is verified to make sure the first number is smaller than the second. Each number in the list is added to the @code{flist} array, which simply lists the fields that will be printed. Normal field splitting is used. The program lets @code{awk} handle the job of doing the field splitting. @example @c @group @c file eg/prog/cut.awk function set_fieldlist( n, m, i, j, k, f, g) @{ n = split(fieldlist, f, ",") j = 1 # index in flist for (i = 1; i <= n; i++) @{ if (index(f[i], "-") != 0) @{ # a range m = split(f[i], g, "-") if (m != 2 || g[1] >= g[2]) @{ printf("bad field list: %s\n", f[i]) > "/dev/stderr" exit 1 @} for (k = g[1]; k <= g[2]; k++) flist[j++] = k @} else flist[j++] = f[i] @} nfields = j - 1 @} @c endfile @c @end group @end example The @code{set_charlist} function is more complicated than @code{set_fieldlist}. The idea here is to use @code{gawk}'s @code{FIELDWIDTHS} variable (@pxref{Constant Size, ,Reading Fixed-width Data}), which describes constant width input. When using a character list, that is exactly what we have. Setting up @code{FIELDWIDTHS} is more complicated than simply listing the fields that need to be printed. We have to keep track of the fields to be printed, and also the intervening characters that have to be skipped. For example, suppose you wanted characters one through eight, 15, and 22 through 35. You would use @samp{-c 1-8,15,22-35}. The necessary value for @code{FIELDWIDTHS} would be @code{@w{"8 6 1 6 14"}}. This gives us five fields, and what should be printed are @code{$1}, @code{$3}, and @code{$5}. The intermediate fields are ``filler,'' stuff in between the desired data. @code{flist} lists the fields to be printed, and @code{t} tracks the complete field list, including filler fields. @example @c @group @c file eg/prog/cut.awk function set_charlist( field, i, j, f, g, t, filler, last, len) @{ field = 1 # count total fields n = split(fieldlist, f, ",") j = 1 # index in flist for (i = 1; i <= n; i++) @{ if (index(f[i], "-") != 0) @{ # range m = split(f[i], g, "-") if (m != 2 || g[1] >= g[2]) @{ printf("bad character list: %s\n", f[i]) > "/dev/stderr" exit 1 @} len = g[2] - g[1] + 1 if (g[1] > 1) # compute length of filler filler = g[1] - last - 1 else filler = 0 if (filler) t[field++] = filler t[field++] = len # length of field last = g[2] flist[j++] = field - 1 @} else @{ if (f[i] > 1) filler = f[i] - last - 1 else filler = 0 if (filler) t[field++] = filler t[field++] = 1 last = f[i] flist[j++] = field - 1 @} @} @group FIELDWIDTHS = join(t, 1, field - 1) nfields = j - 1 @} @end group @c endfile @end example Here is the rule that actually processes the data. If the @samp{-s} option was given, then @code{suppress} will be true. The first @code{if} statement makes sure that the input record does have the field separator. If @code{cut} is processing fields, @code{suppress} is true, and the field separator character is not in the record, then the record is skipped. If the record is valid, then at this point, @code{gawk} has split the data into fields, either using the character in @code{FS} or using fixed-length fields and @code{FIELDWIDTHS}. The loop goes through the list of fields that should be printed. If the corresponding field has data in it, it is printed. If the next field also has data, then the separator character is written out in between the fields. @c 2e: Could use `index($0, FS) != 0' instead of `$0 !~ FS', below @example @c @group @c file eg/prog/cut.awk @{ if (by_fields && suppress && $0 !~ FS) next for (i = 1; i <= nfields; i++) @{ if ($flist[i] != "") @{ printf "%s", $flist[i] if (i < nfields && $flist[i+1] != "") printf "%s", OFS @} @} print "" @} @c endfile @c @end group @end example This version of @code{cut} relies on @code{gawk}'s @code{FIELDWIDTHS} variable to do the character-based cutting. While it would be possible in other @code{awk} implementations to use @code{substr} (@pxref{String Functions, ,Built-in Functions for String Manipulation}), it would also be extremely painful to do so. The @code{FIELDWIDTHS} variable supplies an elegant solution to the problem of picking the input line apart by characters. @node Egrep Program, Id Program, Cut Program, Clones @subsection Searching for Regular Expressions in Files @cindex @code{egrep} utility The @code{egrep} utility searches files for patterns. It uses regular expressions that are almost identical to those available in @code{awk} (@pxref{Regexp Constants, ,Regular Expression Constants}). It is used this way: @example egrep @r{[} @var{options} @r{]} '@var{pattern}' @var{files} @dots{} @end example The @var{pattern} is a regexp. In typical usage, the regexp is quoted to prevent the shell from expanding any of the special characters as file name wildcards. Normally, @code{egrep} prints the lines that matched. If multiple file names are provided on the command line, each output line is preceded by the name of the file and a colon. @c NEEDED @page The options are: @table @code @item -c Print out a count of the lines that matched the pattern, instead of the lines themselves. @item -s Be silent. No output is produced, and the exit value indicates whether or not the pattern was matched. @item -v Invert the sense of the test. @code{egrep} prints the lines that do @emph{not} match the pattern, and exits successfully if the pattern was not matched. @item -i Ignore case distinctions in both the pattern and the input data. @item -l Only print the names of the files that matched, not the lines that matched. @item -e @var{pattern} Use @var{pattern} as the regexp to match. The purpose of the @samp{-e} option is to allow patterns that start with a @samp{-}. @end table This version uses the @code{getopt} library function (@pxref{Getopt Function, ,Processing Command Line Options}), and the file transition library program (@pxref{Filetrans Function, ,Noting Data File Boundaries}). The program begins with a descriptive comment, and then a @code{BEGIN} rule that processes the command line arguments with @code{getopt}. The @samp{-i} (ignore case) option is particularly easy with @code{gawk}; we just use the @code{IGNORECASE} built in variable (@pxref{Built-in Variables}). @findex egrep.awk @example @c @group @c file eg/prog/egrep.awk # egrep.awk --- simulate egrep in awk # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # Options: # -c count of lines # -s silent - use exit value # -v invert test, success if no match # -i ignore case # -l print filenames only # -e argument is pattern BEGIN @{ while ((c = getopt(ARGC, ARGV, "ce:svil")) != -1) @{ if (c == "c") count_only++ else if (c == "s") no_print++ else if (c == "v") invert++ else if (c == "i") IGNORECASE = 1 else if (c == "l") filenames_only++ else if (c == "e") pattern = Optarg else usage() @} @c endfile @c @end group @end example Next comes the code that handles the @code{egrep} specific behavior. If no pattern was supplied with @samp{-e}, the first non-option on the command line is used. The @code{awk} command line arguments up to @code{ARGV[Optind]} are cleared, so that @code{awk} won't try to process them as files. If no files were specified, the standard input is used, and if multiple files were specified, we make sure to note this so that the file names can precede the matched lines in the output. The last two lines are commented out, since they are not needed in @code{gawk}. They should be uncommented if you have to use another version of @code{awk}. @example @c @group @c file eg/prog/egrep.awk if (pattern == "") pattern = ARGV[Optind++] for (i = 1; i < Optind; i++) ARGV[i] = "" if (Optind >= ARGC) @{ ARGV[1] = "-" ARGC = 2 @} else if (ARGC - Optind > 1) do_filenames++ # if (IGNORECASE) # pattern = tolower(pattern) @} @c endfile @c @end group @end example The next set of lines should be uncommented if you are not using @code{gawk}. This rule translates all the characters in the input line into lower-case if the @samp{-i} option was specified. The rule is commented out since it is not necessary with @code{gawk}. @c bug: if a match happens, we output the translated line, not the original @example @c @group @c file eg/prog/egrep.awk #@{ # if (IGNORECASE) # $0 = tolower($0) #@} @c endfile @c @end group @end example The @code{beginfile} function is called by the rule in @file{ftrans.awk} when each new file is processed. In this case, it is very simple; all it does is initialize a variable @code{fcount} to zero. @code{fcount} tracks how many lines in the current file matched the pattern. @example @group @c file eg/prog/egrep.awk function beginfile(junk) @{ fcount = 0 @} @c endfile @end group @end example The @code{endfile} function is called after each file has been processed. It is used only when the user wants a count of the number of lines that matched. @code{no_print} will be true only if the exit status is desired. @code{count_only} will be true if line counts are desired. @code{egrep} will therefore only print line counts if printing and counting are enabled. The output format must be adjusted depending upon the number of files to be processed. Finally, @code{fcount} is added to @code{total}, so that we know how many lines altogether matched the pattern. @example @c @group @c file eg/prog/egrep.awk function endfile(file) @{ if (! no_print && count_only) if (do_filenames) print file ":" fcount else print fcount total += fcount @} @c endfile @c @end group @end example This rule does most of the work of matching lines. The variable @code{matches} will be true if the line matched the pattern. If the user wants lines that did not match, the sense of the @code{matches} is inverted using the @samp{!} operator. @code{fcount} is incremented with the value of @code{matches}, which will be either one or zero, depending upon a successful or unsuccessful match. If the line did not match, the @code{next} statement just moves on to the next record. There are several optimizations for performance in the following few lines of code. If the user only wants exit status (@code{no_print} is true), and we don't have to count lines, then it is enough to know that one line in this file matched, and we can skip on to the next file with @code{nextfile}. Along similar lines, if we are only printing file names, and we don't need to count lines, we can print the file name, and then skip to the next file with @code{nextfile}. Finally, each line is printed, with a leading filename and colon if necessary. @ignore 2e: note, probably better to recode the last few lines as if (! count_only) @{ if (no_print) nextfile if (filenames_only) @{ print FILENAME nextfile @} if (do_filenames) print FILENAME ":" $0 else print @} @end ignore @example @c @group @c file eg/prog/egrep.awk @{ matches = ($0 ~ pattern) if (invert) matches = ! matches fcount += matches # 1 or 0 @group if (! matches) next @end group if (no_print && ! count_only) nextfile if (filenames_only && ! count_only) @{ print FILENAME nextfile @} if (do_filenames && ! count_only) print FILENAME ":" $0 else if (! count_only) print @} @c endfile @c @end group @end example @c @strong{Exercise}: rearrange the code inside @samp{if (! count_only)}. The @code{END} rule takes care of producing the correct exit status. If there were no matches, the exit status is one, otherwise it is zero. @example @c @group @c file eg/prog/egrep.awk END \ @{ if (total == 0) exit 1 exit 0 @} @c endfile @c @end group @end example The @code{usage} function prints a usage message in case of invalid options and then exits. @example @c @group @c file eg/prog/egrep.awk function usage( e) @{ e = "Usage: egrep [-csvil] [-e pat] [files ...]" print e > "/dev/stderr" exit 1 @} @c endfile @c @end group @end example The variable @code{e} is used so that the function fits nicely on the printed page. @cindex backslash continuation Just a note on programming style. You may have noticed that the @code{END} rule uses backslash continuation, with the open brace on a line by itself. This is so that it more closely resembles the way functions are written. Many of the examples @iftex in this chapter @end iftex use this style. You can decide for yourself if you like writing your @code{BEGIN} and @code{END} rules this way, or not. @node Id Program, Split Program, Egrep Program, Clones @subsection Printing Out User Information @cindex @code{id} utility The @code{id} utility lists a user's real and effective user-id numbers, real and effective group-id numbers, and the user's group set, if any. @code{id} will only print the effective user-id and group-id if they are different from the real ones. If possible, @code{id} will also supply the corresponding user and group names. The output might look like this: @example $ id @print{} uid=2076(arnold) gid=10(staff) groups=10(staff),4(tty) @end example This information is exactly what is provided by @code{gawk}'s @file{/dev/user} special file (@pxref{Special Files, ,Special File Names in @code{gawk}}). However, the @code{id} utility provides a more palatable output than just a string of numbers. Here is a simple version of @code{id} written in @code{awk}. It uses the user database library functions (@pxref{Passwd Functions, ,Reading the User Database}), and the group database library functions (@pxref{Group Functions, ,Reading the Group Database}). The program is fairly straightforward. All the work is done in the @code{BEGIN} rule. The user and group id numbers are obtained from @file{/dev/user}. If there is no support for @file{/dev/user}, the program gives up. The code is repetitive. The entry in the user database for the real user-id number is split into parts at the @samp{:}. The name is the first field. Similar code is used for the effective user-id number, and the group numbers. @findex id.awk @example @c @group @c file eg/prog/id.awk # id.awk --- implement id in awk # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # output is: # uid=12(foo) euid=34(bar) gid=3(baz) \ # egid=5(blat) groups=9(nine),2(two),1(one) BEGIN \ @{ if ((getline < "/dev/user") < 0) @{ err = "id: no /dev/user support - cannot run" print err > "/dev/stderr" exit 1 @} close("/dev/user") uid = $1 euid = $2 gid = $3 egid = $4 printf("uid=%d", uid) pw = getpwuid(uid) @group if (pw != "") @{ split(pw, a, ":") printf("(%s)", a[1]) @} @end group if (euid != uid) @{ printf(" euid=%d", euid) pw = getpwuid(euid) if (pw != "") @{ split(pw, a, ":") printf("(%s)", a[1]) @} @} printf(" gid=%d", gid) pw = getgrgid(gid) if (pw != "") @{ split(pw, a, ":") printf("(%s)", a[1]) @} if (egid != gid) @{ printf(" egid=%d", egid) pw = getgrgid(egid) if (pw != "") @{ split(pw, a, ":") printf("(%s)", a[1]) @} @} if (NF > 4) @{ printf(" groups="); for (i = 5; i <= NF; i++) @{ printf("%d", $i) pw = getgrgid($i) if (pw != "") @{ split(pw, a, ":") printf("(%s)", a[1]) @} @group if (i < NF) printf(",") @end group @} @} print "" @} @c endfile @c @end group @end example @c exercise!!! @ignore The POSIX version of @code{id} takes arguments that control which information is printed. Modify this version to accept the same arguments and perform in the same way. @end ignore @node Split Program, Tee Program, Id Program, Clones @subsection Splitting a Large File Into Pieces @cindex @code{split} utility The @code{split} program splits large text files into smaller pieces. By default, the output files are named @file{xaa}, @file{xab}, and so on. Each file has 1000 lines in it, with the likely exception of the last file. To change the number of lines in each file, you supply a number on the command line preceded with a minus, e.g., @samp{-500} for files with 500 lines in them instead of 1000. To change the name of the output files to something like @file{myfileaa}, @file{myfileab}, and so on, you supply an additional argument that specifies the filename. Here is a version of @code{split} in @code{awk}. It uses the @code{ord} and @code{chr} functions presented in @ref{Ordinal Functions, ,Translating Between Characters and Numbers}. The program first sets its defaults, and then tests to make sure there are not too many arguments. It then looks at each argument in turn. The first argument could be a minus followed by a number. If it is, this happens to look like a negative number, so it is made positive, and that is the count of lines. The data file name is skipped over, and the final argument is used as the prefix for the output file names. @findex split.awk @example @c @group @c file eg/prog/split.awk # split.awk --- do split in awk # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # usage: split [-num] [file] [outname] BEGIN @{ outfile = "x" # default count = 1000 if (ARGC > 4) usage() i = 1 if (ARGV[i] ~ /^-[0-9]+$/) @{ count = -ARGV[i] ARGV[i] = "" i++ @} # test argv in case reading from stdin instead of file if (i in ARGV) i++ # skip data file name if (i in ARGV) @{ outfile = ARGV[i] ARGV[i] = "" @} s1 = s2 = "a" out = (outfile s1 s2) @} @c endfile @c @end group @end example The next rule does most of the work. @code{tcount} (temporary count) tracks how many lines have been printed to the output file so far. If it is greater than @code{count}, it is time to close the current file and start a new one. @code{s1} and @code{s2} track the current suffixes for the file name. If they are both @samp{z}, the file is just too big. Otherwise, @code{s1} moves to the next letter in the alphabet and @code{s2} starts over again at @samp{a}. @example @c @group @c file eg/prog/split.awk @{ if (++tcount > count) @{ close(out) if (s2 == "z") @{ if (s1 == "z") @{ printf("split: %s is too large to split\n", \ FILENAME) > "/dev/stderr" exit 1 @} s1 = chr(ord(s1) + 1) s2 = "a" @} else s2 = chr(ord(s2) + 1) out = (outfile s1 s2) tcount = 1 @} print > out @} @c endfile @c @end group @end example The @code{usage} function simply prints an error message and exits. @example @c @group @c file eg/prog/split.awk function usage( e) @{ e = "usage: split [-num] [file] [outname]" print e > "/dev/stderr" exit 1 @} @c endfile @c @end group @end example @noindent The variable @code{e} is used so that the function fits nicely on the @iftex page. @end iftex @ifinfo screen. @end ifinfo This program is a bit sloppy; it relies on @code{awk} to close the last file for it automatically, instead of doing it in an @code{END} rule. @node Tee Program, Uniq Program, Split Program, Clones @subsection Duplicating Output Into Multiple Files @cindex @code{tee} utility The @code{tee} program is known as a ``pipe fitting.'' @code{tee} copies its standard input to its standard output, and also duplicates it to the files named on the command line. Its usage is: @example tee @r{[}-a@r{]} file @dots{} @end example The @samp{-a} option tells @code{tee} to append to the named files, instead of truncating them and starting over. The @code{BEGIN} rule first makes a copy of all the command line arguments, into an array named @code{copy}. @code{ARGV[0]} is not copied, since it is not needed. @code{tee} cannot use @code{ARGV} directly, since @code{awk} will attempt to process each file named in @code{ARGV} as input data. If the first argument is @samp{-a}, then the flag variable @code{append} is set to true, and both @code{ARGV[1]} and @code{copy[1]} are deleted. If @code{ARGC} is less than two, then no file names were supplied, and @code{tee} prints a usage message and exits. Finally, @code{awk} is forced to read the standard input by setting @code{ARGV[1]} to @code{"-"}, and @code{ARGC} to two. @c 2e: the `ARGC--' in the `if (ARGV[1] == "-a")' isn't needed. @findex tee.awk @example @group @c file eg/prog/tee.awk # tee.awk --- tee in awk # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # Revised December 1995 @end group @group BEGIN \ @{ for (i = 1; i < ARGC; i++) copy[i] = ARGV[i] @end group @group if (ARGV[1] == "-a") @{ append = 1 delete ARGV[1] delete copy[1] ARGC-- @} @end group @group if (ARGC < 2) @{ print "usage: tee [-a] file ..." > "/dev/stderr" exit 1 @} @end group @group ARGV[1] = "-" ARGC = 2 @} @c endfile @end group @end example The single rule does all the work. Since there is no pattern, it is executed for each line of input. The body of the rule simply prints the line into each file on the command line, and then to the standard output. @example @group @c file eg/prog/tee.awk @{ # moving the if outside the loop makes it run faster if (append) for (i in copy) print >> copy[i] else for (i in copy) print > copy[i] print @} @c endfile @end group @end example It would have been possible to code the loop this way: @example for (i in copy) if (append) print >> copy[i] else print > copy[i] @end example @noindent This is more concise, but it is also less efficient. The @samp{if} is tested for each record and for each output file. By duplicating the loop body, the @samp{if} is only tested once for each input record. If there are @var{N} input records and @var{M} input files, the first method only executes @var{N} @samp{if} statements, while the second would execute @var{N}@code{*}@var{M} @samp{if} statements. Finally, the @code{END} rule cleans up, by closing all the output files. @example @c @group @c file eg/prog/tee.awk END \ @{ for (i in copy) close(copy[i]) @} @c endfile @c @end group @end example @node Uniq Program, Wc Program, Tee Program, Clones @subsection Printing Non-duplicated Lines of Text @cindex @code{uniq} utility The @code{uniq} utility reads sorted lines of data on its standard input, and (by default) removes duplicate lines. In other words, only unique lines are printed, hence the name. @code{uniq} has a number of options. The usage is: @example uniq @r{[}-udc @r{[}-@var{n}@r{]]} @r{[}+@var{n}@r{]} @r{[} @var{input file} @r{[} @var{output file} @r{]]} @end example The option meanings are: @table @code @item -d Only print repeated lines. @item -u Only print non-repeated lines. @item -c Count lines. This option overrides @samp{-d} and @samp{-u}. Both repeated and non-repeated lines are counted. @item -@var{n} Skip @var{n} fields before comparing lines. The definition of fields is similar to @code{awk}'s default: non-whitespace characters separated by runs of spaces and/or tabs. @item +@var{n} Skip @var{n} characters before comparing lines. Any fields specified with @samp{-@var{n}} are skipped first. @item @var{input file} Data is read from the input file named on the command line, instead of from the standard input. @item @var{output file} The generated output is sent to the named output file, instead of to the standard output. @end table Normally @code{uniq} behaves as if both the @samp{-d} and @samp{-u} options had been provided. Here is an @code{awk} implementation of @code{uniq}. It uses the @code{getopt} library function (@pxref{Getopt Function, ,Processing Command Line Options}), and the @code{join} library function (@pxref{Join Function, ,Merging an Array Into a String}). The program begins with a @code{usage} function and then a brief outline of the options and their meanings in a comment. The @code{BEGIN} rule deals with the command line arguments and options. It uses a trick to get @code{getopt} to handle options of the form @samp{-25}, treating such an option as the option letter @samp{2} with an argument of @samp{5}. If indeed two or more digits were supplied (@code{Optarg} looks like a number), @code{Optarg} is concatenated with the option digit, and then result is added to zero to make it into a number. If there is only one digit in the option, then @code{Optarg} is not needed, and @code{Optind} must be decremented so that @code{getopt} will process it next time. This code is admittedly a bit tricky. If no options were supplied, then the default is taken, to print both repeated and non-repeated lines. The output file, if provided, is assigned to @code{outputfile}. Earlier, @code{outputfile} was initialized to the standard output, @file{/dev/stdout}. @findex uniq.awk @example @c @group @c file eg/prog/uniq.awk # uniq.awk --- do uniq in awk # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 @group function usage( e) @{ e = "Usage: uniq [-udc [-n]] [+n] [ in [ out ]]" print e > "/dev/stderr" exit 1 @} @end group @group # -c count lines. overrides -d and -u # -d only repeated lines # -u only non-repeated lines # -n skip n fields # +n skip n characters, skip fields first @end group BEGIN \ @{ count = 1 outputfile = "/dev/stdout" opts = "udc0:1:2:3:4:5:6:7:8:9:" while ((c = getopt(ARGC, ARGV, opts)) != -1) @{ if (c == "u") non_repeated_only++ else if (c == "d") repeated_only++ else if (c == "c") do_count++ else if (index("0123456789", c) != 0) @{ # getopt requires args to options # this messes us up for things like -5 if (Optarg ~ /^[0-9]+$/) fcount = (c Optarg) + 0 else @{ fcount = c + 0 Optind-- @} @} else usage() @} if (ARGV[Optind] ~ /^\+[0-9]+$/) @{ charcount = substr(ARGV[Optind], 2) + 0 Optind++ @} for (i = 1; i < Optind; i++) ARGV[i] = "" if (repeated_only == 0 && non_repeated_only == 0) repeated_only = non_repeated_only = 1 @group if (ARGC - Optind == 2) @{ outputfile = ARGV[ARGC - 1] ARGV[ARGC - 1] = "" @} @} @c endfile @end group @end example The following function, @code{are_equal}, compares the current line, @code{$0}, to the previous line, @code{last}. It handles skipping fields and characters. If no field count and no character count were specified, @code{are_equal} simply returns one or zero depending upon the result of a simple string comparison of @code{last} and @code{$0}. Otherwise, things get more complicated. If fields have to be skipped, each line is broken into an array using @code{split} (@pxref{String Functions, ,Built-in Functions for String Manipulation}), and then the desired fields are joined back into a line using @code{join}. The joined lines are stored in @code{clast} and @code{cline}. If no fields are skipped, @code{clast} and @code{cline} are set to @code{last} and @code{$0} respectively. Finally, if characters are skipped, @code{substr} is used to strip off the leading @code{charcount} characters in @code{clast} and @code{cline}. The two strings are then compared, and @code{are_equal} returns the result. @example @c @group @c file eg/prog/uniq.awk function are_equal( n, m, clast, cline, alast, aline) @{ if (fcount == 0 && charcount == 0) return (last == $0) if (fcount > 0) @{ n = split(last, alast) m = split($0, aline) clast = join(alast, fcount+1, n) cline = join(aline, fcount+1, m) @} else @{ clast = last cline = $0 @} if (charcount) @{ clast = substr(clast, charcount + 1) cline = substr(cline, charcount + 1) @} return (clast == cline) @} @c endfile @c @end group @end example The following two rules are the body of the program. The first one is executed only for the very first line of data. It sets @code{last} equal to @code{$0}, so that subsequent lines of text have something to be compared to. The second rule does the work. The variable @code{equal} will be one or zero depending upon the results of @code{are_equal}'s comparison. If @code{uniq} is counting repeated lines, then the @code{count} variable is incremented if the lines are equal. Otherwise the line is printed and @code{count} is reset, since the two lines are not equal. If @code{uniq} is not counting, @code{count} is incremented if the lines are equal. Otherwise, if @code{uniq} is counting repeated lines, and more than one line has been seen, or if @code{uniq} is counting non-repeated lines, and only one line has been seen, then the line is printed, and @code{count} is reset. Finally, similar logic is used in the @code{END} rule to print the final line of input data. @example @c @group @c file eg/prog/uniq.awk @group NR == 1 @{ last = $0 next @} @end group @{ equal = are_equal() if (do_count) @{ # overrides -d and -u if (equal) count++ else @{ printf("%4d %s\n", count, last) > outputfile last = $0 count = 1 # reset @} next @} if (equal) count++ else @{ if ((repeated_only && count > 1) || (non_repeated_only && count == 1)) print last > outputfile last = $0 count = 1 @} @} @group END @{ if (do_count) printf("%4d %s\n", count, last) > outputfile else if ((repeated_only && count > 1) || (non_repeated_only && count == 1)) print last > outputfile @} @end group @c endfile @c @end group @end example @node Wc Program, , Uniq Program, Clones @subsection Counting Things @cindex @code{wc} utility The @code{wc} (word count) utility counts lines, words, and characters in one or more input files. Its usage is: @example wc @r{[}-lwc@r{]} @r{[} @var{files} @dots{} @r{]} @end example If no files are specified on the command line, @code{wc} reads its standard input. If there are multiple files, it will also print total counts for all the files. The options and their meanings are: @table @code @item -l Only count lines. @item -w Only count words. A ``word'' is a contiguous sequence of non-whitespace characters, separated by spaces and/or tabs. Happily, this is the normal way @code{awk} separates fields in its input data. @item -c Only count characters. @end table Implementing @code{wc} in @code{awk} is particularly elegant, since @code{awk} does a lot of the work for us; it splits lines into words (i.e.@: fields) and counts them, it counts lines (i.e.@: records) for us, and it can easily tell us how long a line is. This version uses the @code{getopt} library function (@pxref{Getopt Function, ,Processing Command Line Options}), and the file transition functions (@pxref{Filetrans Function, ,Noting Data File Boundaries}). This version has one major difference from traditional versions of @code{wc}. Our version always prints the counts in the order lines, words, and characters. Traditional versions note the order of the @samp{-l}, @samp{-w}, and @samp{-c} options on the command line, and print the counts in that order. The @code{BEGIN} rule does the argument processing. The variable @code{print_total} will be true if more than one file was named on the command line. @findex wc.awk @example @c @group @c file eg/prog/wc.awk # wc.awk --- count lines, words, characters # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # Options: # -l only count lines # -w only count words # -c only count characters # # Default is to count lines, words, characters BEGIN @{ # let getopt print a message about # invalid options. we ignore them while ((c = getopt(ARGC, ARGV, "lwc")) != -1) @{ if (c == "l") do_lines = 1 else if (c == "w") do_words = 1 else if (c == "c") do_chars = 1 @} for (i = 1; i < Optind; i++) ARGV[i] = "" # if no options, do all if (! do_lines && ! do_words && ! do_chars) do_lines = do_words = do_chars = 1 print_total = (ARGC - i > 2) @} @c endfile @c @end group @end example The @code{beginfile} function is simple; it just resets the counts of lines, words, and characters to zero, and saves the current file name in @code{fname}. The @code{endfile} function adds the current file's numbers to the running totals of lines, words, and characters. It then prints out those numbers for the file that was just read. It relies on @code{beginfile} to reset the numbers for the following data file. @example @c @group @c file eg/prog/wc.awk function beginfile(file) @{ chars = lines = words = 0 fname = FILENAME @} function endfile(file) @{ tchars += chars tlines += lines twords += words @group if (do_lines) printf "\t%d", lines @end group if (do_words) printf "\t%d", words if (do_chars) printf "\t%d", chars printf "\t%s\n", fname @} @c endfile @c @end group @end example There is one rule that is executed for each line. It adds the length of the record to @code{chars}. It has to add one, since the newline character separating records (the value of @code{RS}) is not part of the record itself. @code{lines} is incremented for each line read, and @code{words} is incremented by the value of @code{NF}, the number of ``words'' on this line.@footnote{Examine the code in @ref{Filetrans Function, ,Noting Data File Boundaries}. Why must @code{wc} use a separate @code{lines} variable, instead of using the value of @code{FNR} in @code{endfile}?} Finally, the @code{END} rule simply prints the totals for all the files. @example @c @group @c file eg/prog/wc.awk # do per line @{ chars += length($0) + 1 # get newline lines++ words += NF @} END @{ if (print_total) @{ if (do_lines) printf "\t%d", tlines if (do_words) printf "\t%d", twords if (do_chars) printf "\t%d", tchars print "\ttotal" @} @} @c endfile @c @end group @end example @node Miscellaneous Programs, , Clones, Sample Programs @section A Grab Bag of @code{awk} Programs This section is a large ``grab bag'' of miscellaneous programs. We hope you find them both interesting and enjoyable. @menu * Dupword Program:: Finding duplicated words in a document. * Alarm Program:: An alarm clock. * Translate Program:: A program similar to the @code{tr} utility. * Labels Program:: Printing mailing labels. * Word Sorting:: A program to produce a word usage count. * History Sorting:: Eliminating duplicate entries from a history file. * Extract Program:: Pulling out programs from Texinfo source files. * Simple Sed:: A Simple Stream Editor. * Igawk Program:: A wrapper for @code{awk} that includes files. @end menu @node Dupword Program, Alarm Program, Miscellaneous Programs, Miscellaneous Programs @subsection Finding Duplicated Words in a Document A common error when writing large amounts of prose is to accidentally duplicate words. Often you will see this in text as something like ``the the program does the following @dots{}.'' When the text is on-line, often the duplicated words occur at the end of one line and the beginning of another, making them very difficult to spot. @c as here! This program, @file{dupword.awk}, scans through a file one line at a time, and looks for adjacent occurrences of the same word. It also saves the last word on a line (in the variable @code{prev}) for comparison with the first word on the next line. The first two statements make sure that the line is all lower-case, so that, for example, ``The'' and ``the'' compare equal to each other. The second statement removes all non-alphanumeric and non-whitespace characters from the line, so that punctuation does not affect the comparison either. This sometimes leads to reports of duplicated words that really are different, but this is unusual. @c FIXME: add check for $i != "" @findex dupword.awk @example @group @c file eg/prog/dupword.awk # dupword --- find duplicate words in text # Arnold Robbins, arnold@@gnu.org, Public Domain # December 1991 @{ $0 = tolower($0) gsub(/[^A-Za-z0-9 \t]/, ""); if ($1 == prev) printf("%s:%d: duplicate %s\n", FILENAME, FNR, $1) for (i = 2; i <= NF; i++) if ($i == $(i-1)) printf("%s:%d: duplicate %s\n", FILENAME, FNR, $i) prev = $NF @} @c endfile @end group @end example @node Alarm Program, Translate Program, Dupword Program, Miscellaneous Programs @subsection An Alarm Clock Program The following program is a simple ``alarm clock'' program. You give it a time of day, and an optional message. At the given time, it prints the message on the standard output. In addition, you can give it the number of times to repeat the message, and also a delay between repetitions. This program uses the @code{gettimeofday} function from @ref{Gettimeofday Function, ,Managing the Time of Day}. All the work is done in the @code{BEGIN} rule. The first part is argument checking and setting of defaults; the delay, the count, and the message to print. If the user supplied a message, but it does not contain the ASCII BEL character (known as the ``alert'' character, @samp{\a}), then it is added to the message. (On many systems, printing the ASCII BEL generates some sort of audible alert. Thus, when the alarm goes off, the system calls attention to itself, in case the user is not looking at their computer or terminal.) @findex alarm.awk @example @c @group @c file eg/prog/alarm.awk # alarm --- set an alarm # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # usage: alarm time [ "message" [ count [ delay ] ] ] BEGIN \ @{ # Initial argument sanity checking usage1 = "usage: alarm time ['message' [count [delay]]]" usage2 = sprintf("\t(%s) time ::= hh:mm", ARGV[1]) if (ARGC < 2) @{ print usage > "/dev/stderr" exit 1 @} else if (ARGC == 5) @{ delay = ARGV[4] + 0 count = ARGV[3] + 0 message = ARGV[2] @} else if (ARGC == 4) @{ count = ARGV[3] + 0 message = ARGV[2] @} else if (ARGC == 3) @{ message = ARGV[2] @} else if (ARGV[1] !~ /[0-9]?[0-9]:[0-9][0-9]/) @{ print usage1 > "/dev/stderr" print usage2 > "/dev/stderr" exit 1 @} # set defaults for once we reach the desired time if (delay == 0) delay = 180 # 3 minutes if (count == 0) count = 5 @group if (message == "") message = sprintf("\aIt is now %s!\a", ARGV[1]) else if (index(message, "\a") == 0) message = "\a" message "\a" @end group @c endfile @end example The next section of code turns the alarm time into hours and minutes, and converts it if necessary to a 24-hour clock. Then it turns that time into a count of the seconds since midnight. Next it turns the current time into a count of seconds since midnight. The difference between the two is how long to wait before setting off the alarm. @example @c @group @c file eg/prog/alarm.awk # split up dest time split(ARGV[1], atime, ":") hour = atime[1] + 0 # force numeric minute = atime[2] + 0 # force numeric # get current broken down time gettimeofday(now) # if time given is 12-hour hours and it's after that # hour, e.g., `alarm 5:30' at 9 a.m. means 5:30 p.m., # then add 12 to real hour if (hour < 12 && now["hour"] > hour) hour += 12 # set target time in seconds since midnight target = (hour * 60 * 60) + (minute * 60) # get current time in seconds since midnight current = (now["hour"] * 60 * 60) + \ (now["minute"] * 60) + now["second"] # how long to sleep for naptime = target - current if (naptime <= 0) @{ print "time is in the past!" > "/dev/stderr" exit 1 @} @c endfile @c @end group @end example Finally, the program uses the @code{system} function (@pxref{I/O Functions, ,Built-in Functions for Input/Output}) to call the @code{sleep} utility. The @code{sleep} utility simply pauses for the given number of seconds. If the exit status is not zero, the program assumes that @code{sleep} was interrupted, and exits. If @code{sleep} exited with an OK status (zero), then the program prints the message in a loop, again using @code{sleep} to delay for however many seconds are necessary. @example @c @group @c file eg/prog/alarm.awk # zzzzzz..... go away if interrupted if (system(sprintf("sleep %d", naptime)) != 0) exit 1 # time to notify! command = sprintf("sleep %d", delay) for (i = 1; i <= count; i++) @{ print message # if sleep command interrupted, go away if (system(command) != 0) break @} exit 0 @} @c endfile @c @end group @end example @node Translate Program, Labels Program, Alarm Program, Miscellaneous Programs @subsection Transliterating Characters The system @code{tr} utility transliterates characters. For example, it is often used to map upper-case letters into lower-case, for further processing. @example @var{generate data} | tr '[A-Z]' '[a-z]' | @var{process data} @dots{} @end example You give @code{tr} two lists of characters enclosed in square brackets. Usually, the lists are quoted to keep the shell from attempting to do a filename expansion.@footnote{On older, non-POSIX systems, @code{tr} often does not require that the lists be enclosed in square brackets and quoted. This is a feature.} When processing the input, the first character in the first list is replaced with the first character in the second list, the second character in the first list is replaced with the second character in the second list, and so on. If there are more characters in the ``from'' list than in the ``to'' list, the last character of the ``to'' list is used for the remaining characters in the ``from'' list. Some time ago, @c early or mid-1989! a user proposed to us that we add a transliteration function to @code{gawk}. Being opposed to ``creeping featurism,'' I wrote the following program to prove that character transliteration could be done with a user-level function. This program is not as complete as the system @code{tr} utility, but it will do most of the job. The @code{translate} program demonstrates one of the few weaknesses of standard @code{awk}: dealing with individual characters is very painful, requiring repeated use of the @code{substr}, @code{index}, and @code{gsub} built-in functions (@pxref{String Functions, ,Built-in Functions for String Manipulation}).@footnote{This program was written before @code{gawk} acquired the ability to split each character in a string into separate array elements. How might this ability simplify the program?} There are two functions. The first, @code{stranslate}, takes three arguments. @table @code @item from A list of characters to translate from. @item to A list of characters to translate to. @item target The string to do the translation on. @end table Associative arrays make the translation part fairly easy. @code{t_ar} holds the ``to'' characters, indexed by the ``from'' characters. Then a simple loop goes through @code{from}, one character at a time. For each character in @code{from}, if the character appears in @code{target}, @code{gsub} is used to change it to the corresponding @code{to} character. The @code{translate} function simply calls @code{stranslate} using @code{$0} as the target. The main program sets two global variables, @code{FROM} and @code{TO}, from the command line, and then changes @code{ARGV} so that @code{awk} will read from the standard input. Finally, the processing rule simply calls @code{translate} for each record. @findex translate.awk @example @c @group @c file eg/prog/translate.awk # translate --- do tr like stuff # Arnold Robbins, arnold@@gnu.org, Public Domain # August 1989 # bugs: does not handle things like: tr A-Z a-z, it has # to be spelled out. However, if `to' is shorter than `from', # the last character in `to' is used for the rest of `from'. function stranslate(from, to, target, lf, lt, t_ar, i, c) @{ lf = length(from) lt = length(to) for (i = 1; i <= lt; i++) t_ar[substr(from, i, 1)] = substr(to, i, 1) if (lt < lf) for (; i <= lf; i++) t_ar[substr(from, i, 1)] = substr(to, lt, 1) for (i = 1; i <= lf; i++) @{ c = substr(from, i, 1) if (index(target, c) > 0) gsub(c, t_ar[c], target) @} return target @} @group function translate(from, to) @{ return $0 = stranslate(from, to, $0) @} @end group # main program BEGIN @{ if (ARGC < 3) @{ print "usage: translate from to" > "/dev/stderr" exit @} FROM = ARGV[1] TO = ARGV[2] ARGC = 2 ARGV[1] = "-" @} @{ translate(FROM, TO) print @} @c endfile @c @end group @end example While it is possible to do character transliteration in a user-level function, it is not necessarily efficient, and we started to consider adding a built-in function. However, shortly after writing this program, we learned that the System V Release 4 @code{awk} had added the @code{toupper} and @code{tolower} functions. These functions handle the vast majority of the cases where character transliteration is necessary, and so we chose to simply add those functions to @code{gawk} as well, and then leave well enough alone. An obvious improvement to this program would be to set up the @code{t_ar} array only once, in a @code{BEGIN} rule. However, this assumes that the ``from'' and ``to'' lists will never change throughout the lifetime of the program. @node Labels Program, Word Sorting, Translate Program, Miscellaneous Programs @subsection Printing Mailing Labels Here is a ``real world''@footnote{``Real world'' is defined as ``a program actually used to get something done.''} program. This script reads lists of names and addresses, and generates mailing labels. Each page of labels has 20 labels on it, two across and ten down. The addresses are guaranteed to be no more than five lines of data. Each address is separated from the next by a blank line. The basic idea is to read 20 labels worth of data. Each line of each label is stored in the @code{line} array. The single rule takes care of filling the @code{line} array and printing the page when 20 labels have been read. The @code{BEGIN} rule simply sets @code{RS} to the empty string, so that @code{awk} will split records at blank lines (@pxref{Records, ,How Input is Split into Records}). It sets @code{MAXLINES} to 100, since @code{MAXLINE} is the maximum number of lines on the page (20 * 5 = 100). Most of the work is done in the @code{printpage} function. The label lines are stored sequentially in the @code{line} array. But they have to be printed horizontally; @code{line[1]} next to @code{line[6]}, @code{line[2]} next to @code{line[7]}, and so on. Two loops are used to accomplish this. The outer loop, controlled by @code{i}, steps through every 10 lines of data; this is each row of labels. The inner loop, controlled by @code{j}, goes through the lines within the row. As @code{j} goes from zero to four, @samp{i+j} is the @code{j}'th line in the row, and @samp{i+j+5} is the entry next to it. The output ends up looking something like this: @example line 1 line 6 line 2 line 7 line 3 line 8 line 4 line 9 line 5 line 10 @end example As a final note, at lines 21 and 61, an extra blank line is printed, to keep the output lined up on the labels. This is dependent on the particular brand of labels in use when the program was written. You will also note that there are two blank lines at the top and two blank lines at the bottom. The @code{END} rule arranges to flush the final page of labels; there may not have been an even multiple of 20 labels in the data. @findex labels.awk @example @c @group @c file eg/prog/labels.awk # labels.awk # Arnold Robbins, arnold@@gnu.org, Public Domain # June 1992 # Program to print labels. Each label is 5 lines of data # that may have blank lines. The label sheets have 2 # blank lines at the top and 2 at the bottom. BEGIN @{ RS = "" ; MAXLINES = 100 @} function printpage( i, j) @{ if (Nlines <= 0) return printf "\n\n" # header for (i = 1; i <= Nlines; i += 10) @{ if (i == 21 || i == 61) print "" for (j = 0; j < 5; j++) @{ if (i + j > MAXLINES) break printf " %-41s %s\n", line[i+j], line[i+j+5] @} print "" @} printf "\n\n" # footer for (i in line) line[i] = "" @} # main rule @{ if (Count >= 20) @{ printpage() Count = 0 Nlines = 0 @} n = split($0, a, "\n") for (i = 1; i <= n; i++) line[++Nlines] = a[i] for (; i <= 5; i++) line[++Nlines] = "" Count++ @} END \ @{ printpage() @} @c endfile @c @end group @end example @node Word Sorting, History Sorting, Labels Program, Miscellaneous Programs @subsection Generating Word Usage Counts The following @code{awk} program prints the number of occurrences of each word in its input. It illustrates the associative nature of @code{awk} arrays by using strings as subscripts. It also demonstrates the @samp{for @var{x} in @var{array}} construction. Finally, it shows how @code{awk} can be used in conjunction with other utility programs to do a useful task of some complexity with a minimum of effort. Some explanations follow the program listing. @example awk ' # Print list of word frequencies @{ for (i = 1; i <= NF; i++) freq[$i]++ @} END @{ for (word in freq) printf "%s\t%d\n", word, freq[word] @}' @end example The first thing to notice about this program is that it has two rules. The first rule, because it has an empty pattern, is executed on every line of the input. It uses @code{awk}'s field-accessing mechanism (@pxref{Fields, ,Examining Fields}) to pick out the individual words from the line, and the built-in variable @code{NF} (@pxref{Built-in Variables}) to know how many fields are available. For each input word, an element of the array @code{freq} is incremented to reflect that the word has been seen an additional time. The second rule, because it has the pattern @code{END}, is not executed until the input has been exhausted. It prints out the contents of the @code{freq} table that has been built up inside the first action. This program has several problems that would prevent it from being useful by itself on real text files: @itemize @bullet @item Words are detected using the @code{awk} convention that fields are separated by whitespace and that other characters in the input (except newlines) don't have any special meaning to @code{awk}. This means that punctuation characters count as part of words. @item The @code{awk} language considers upper- and lower-case characters to be distinct. Therefore, @samp{bartender} and @samp{Bartender} are not treated as the same word. This is undesirable since, in normal text, words are capitalized if they begin sentences, and a frequency analyzer should not be sensitive to capitalization. @item The output does not come out in any useful order. You're more likely to be interested in which words occur most frequently, or having an alphabetized table of how frequently each word occurs. @end itemize The way to solve these problems is to use some of the more advanced features of the @code{awk} language. First, we use @code{tolower} to remove case distinctions. Next, we use @code{gsub} to remove punctuation characters. Finally, we use the system @code{sort} utility to process the output of the @code{awk} script. Here is the new version of the program: @findex wordfreq.sh @example @c file eg/prog/wordfreq.awk # Print list of word frequencies @{ $0 = tolower($0) # remove case distinctions gsub(/[^a-z0-9_ \t]/, "", $0) # remove punctuation for (i = 1; i <= NF; i++) freq[$i]++ @} @c endfile END @{ for (word in freq) printf "%s\t%d\n", word, freq[word] @} @end example Assuming we have saved this program in a file named @file{wordfreq.awk}, and that the data is in @file{file1}, the following pipeline @example awk -f wordfreq.awk file1 | sort +1 -nr @end example @noindent produces a table of the words appearing in @file{file1} in order of decreasing frequency. The @code{awk} program suitably massages the data and produces a word frequency table, which is not ordered. The @code{awk} script's output is then sorted by the @code{sort} utility and printed on the terminal. The options given to @code{sort} in this example specify to sort using the second field of each input line (skipping one field), that the sort keys should be treated as numeric quantities (otherwise @samp{15} would come before @samp{5}), and that the sorting should be done in descending (reverse) order. We could have even done the @code{sort} from within the program, by changing the @code{END} action to: @example @c file eg/prog/wordfreq.awk END @{ sort = "sort +1 -nr" for (word in freq) printf "%s\t%d\n", word, freq[word] | sort close(sort) @} @c endfile @end example You would have to use this way of sorting on systems that do not have true pipes. See the general operating system documentation for more information on how to use the @code{sort} program. @node History Sorting, Extract Program, Word Sorting, Miscellaneous Programs @subsection Removing Duplicates from Unsorted Text The @code{uniq} program (@pxref{Uniq Program, ,Printing Non-duplicated Lines of Text}), removes duplicate lines from @emph{sorted} data. Suppose, however, you need to remove duplicate lines from a data file, but that you wish to preserve the order the lines are in? A good example of this might be a shell history file. The history file keeps a copy of all the commands you have entered, and it is not unusual to repeat a command several times in a row. Occasionally you might wish to compact the history by removing duplicate entries. Yet it is desirable to maintain the order of the original commands. This simple program does the job. It uses two arrays. The @code{data} array is indexed by the text of each line. For each line, @code{data[$0]} is incremented. If a particular line has not been seen before, then @code{data[$0]} will be zero. In that case, the text of the line is stored in @code{lines[count]}. Each element of @code{lines} is a unique command, and the indices of @code{lines} indicate the order in which those lines were encountered. The @code{END} rule simply prints out the lines, in order. @cindex Rakitzis, Byron @findex histsort.awk @example @group @c file eg/prog/histsort.awk # histsort.awk --- compact a shell history file # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 # Thanks to Byron Rakitzis for the general idea @{ if (data[$0]++ == 0) lines[++count] = $0 @} END @{ for (i = 1; i <= count; i++) print lines[i] @} @c endfile @end group @end example This program also provides a foundation for generating other useful information. For example, using the following @code{print} satement in the @code{END} rule would indicate how often a particular command was used. @example print data[lines[i]], lines[i] @end example This works because @code{data[$0]} was incremented each time a line was seen. @node Extract Program, Simple Sed, History Sorting, Miscellaneous Programs @subsection Extracting Programs from Texinfo Source Files @iftex Both this chapter and the previous chapter (@ref{Library Functions, ,A Library of @code{awk} Functions}), present a large number of @code{awk} programs. @end iftex @ifinfo The nodes @ref{Library Functions, ,A Library of @code{awk} Functions}, and @ref{Sample Programs, ,Practical @code{awk} Programs}, are the top level nodes for a large number of @code{awk} programs. @end ifinfo If you wish to experiment with these programs, it is tedious to have to type them in by hand. Here we present a program that can extract parts of a Texinfo input file into separate files. This @value{DOCUMENT} is written in Texinfo, the GNU project's document formatting language. A single Texinfo source file can be used to produce both printed and on-line documentation. @iftex Texinfo is fully documented in @cite{Texinfo---The GNU Documentation Format}, available from the Free Software Foundation. @end iftex @ifinfo The Texinfo language is described fully, starting with @ref{Top, , Introduction, texi, Texinfo---The GNU Documentation Format}. @end ifinfo For our purposes, it is enough to know three things about Texinfo input files. @itemize @bullet @item The ``at'' symbol, @samp{@@}, is special in Texinfo, much like @samp{\} in C or @code{awk}. Literal @samp{@@} symbols are represented in Texinfo source files as @samp{@@@@}. @item Comments start with either @samp{@@c} or @samp{@@comment}. The file extraction program will work by using special comments that start at the beginning of a line. @item Example text that should not be split across a page boundary is bracketed between lines containing @samp{@@group} and @samp{@@end group} commands. @end itemize The following program, @file{extract.awk}, reads through a Texinfo source file, and does two things, based on the special comments. Upon seeing @samp{@w{@@c system @dots{}}}, it runs a command, by extracting the command text from the control line and passing it on to the @code{system} function (@pxref{I/O Functions, ,Built-in Functions for Input/Output}). Upon seeing @samp{@@c file @var{filename}}, each subsequent line is sent to the file @var{filename}, until @samp{@@c endfile} is encountered. The rules in @file{extract.awk} will match either @samp{@@c} or @samp{@@comment} by letting the @samp{omment} part be optional. Lines containing @samp{@@group} and @samp{@@end group} are simply removed. @file{extract.awk} uses the @code{join} library function (@pxref{Join Function, ,Merging an Array Into a String}). The example programs in the on-line Texinfo source for @cite{@value{TITLE}} (@file{gawk.texi}) have all been bracketed inside @samp{file}, and @samp{endfile} lines. The @code{gawk} distribution uses a copy of @file{extract.awk} to extract the sample programs and install many of them in a standard directory, where @code{gawk} can find them. The Texinfo file looks something like this: @example @dots{} This program has a @@code@{BEGIN@} block, which prints a nice message: @@example @@c file examples/messages.awk BEGIN @@@{ print "Don't panic!" @@@} @@c end file @@end example It also prints some final advice: @@example @@c file examples/messages.awk END @@@{ print "Always avoid bored archeologists!" @@@} @@c end file @@end example @dots{} @end example @file{extract.awk} begins by setting @code{IGNORECASE} to one, so that mixed upper-case and lower-case letters in the directives won't matter. The first rule handles calling @code{system}, checking that a command was given (@code{NF} is at least three), and also checking that the command exited with a zero exit status, signifying OK. @findex extract.awk @example @c @group @c file eg/prog/extract.awk # extract.awk --- extract files and run programs # from texinfo files # Arnold Robbins, arnold@@gnu.org, Public Domain # May 1993 BEGIN @{ IGNORECASE = 1 @} @group /^@@c(omment)?[ \t]+system/ \ @{ if (NF < 3) @{ e = (FILENAME ":" FNR) e = (e ": badly formed `system' line") print e > "/dev/stderr" next @} $1 = "" $2 = "" stat = system($0) if (stat != 0) @{ e = (FILENAME ":" FNR) e = (e ": warning: system returned " stat) print e > "/dev/stderr" @} @} @end group @c endfile @end example @noindent The variable @code{e} is used so that the function fits nicely on the @iftex page. @end iftex @ifinfo screen. @end ifinfo The second rule handles moving data into files. It verifies that a file name was given in the directive. If the file named is not the current file, then the current file is closed. This means that an @samp{@@c endfile} was not given for that file. (We should probably print a diagnostic in this case, although at the moment we do not.) The @samp{for} loop does the work. It reads lines using @code{getline} (@pxref{Getline, ,Explicit Input with @code{getline}}). For an unexpected end of file, it calls the @code{@w{unexpected_eof}} function. If the line is an ``endfile'' line, then it breaks out of the loop. If the line is an @samp{@@group} or @samp{@@end group} line, then it ignores it, and goes on to the next line. (These Texinfo control lines keep blocks of code together on one page; unfortunately, @TeX{} isn't always smart enough to do things exactly right, and we have to give it some advice.) Most of the work is in the following few lines. If the line has no @samp{@@} symbols, it can be printed directly. Otherwise, each leading @samp{@@} must be stripped off. To remove the @samp{@@} symbols, the line is split into separate elements of the array @code{a}, using the @code{split} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). Each element of @code{a} that is empty indicates two successive @samp{@@} symbols in the original line. For each two empty elements (@samp{@@@@} in the original file), we have to add back in a single @samp{@@} symbol. When the processing of the array is finished, @code{join} is called with the value of @code{SUBSEP}, to rejoin the pieces back into a single line. That line is then printed to the output file. @example @c @group @c file eg/prog/extract.awk @group /^@@c(omment)?[ \t]+file/ \ @{ if (NF != 3) @{ e = (FILENAME ":" FNR ": badly formed `file' line") print e > "/dev/stderr" next @} @end group if ($3 != curfile) @{ if (curfile != "") close(curfile) curfile = $3 @} for (;;) @{ if ((getline line) <= 0) unexpected_eof() if (line ~ /^@@c(omment)?[ \t]+endfile/) break else if (line ~ /^@@(end[ \t]+)?group/) continue if (index(line, "@@") == 0) @{ print line > curfile continue @} n = split(line, a, "@@") @group # if a[1] == "", means leading @@, # don't add one back in. @end group for (i = 2; i <= n; i++) @{ if (a[i] == "") @{ # was an @@@@ a[i] = "@@" if (a[i+1] == "") i++ @} @} print join(a, 1, n, SUBSEP) > curfile @} @} @c endfile @c @end group @end example An important thing to note is the use of the @samp{>} redirection. Output done with @samp{>} only opens the file once; it stays open and subsequent output is appended to the file (@pxref{Redirection, , Redirecting Output of @code{print} and @code{printf}}). This allows us to easily mix program text and explanatory prose for the same sample source file (as has been done here!) without any hassle. The file is only closed when a new data file name is encountered, or at the end of the input file. Finally, the function @code{@w{unexpected_eof}} prints an appropriate error message and then exits. The @code{END} rule handles the final cleanup, closing the open file. @example @c file eg/prog/extract.awk @group function unexpected_eof() @{ printf("%s:%d: unexpected EOF or error\n", \ FILENAME, FNR) > "/dev/stderr" exit 1 @} @end group END @{ if (curfile) close(curfile) @} @c endfile @end example @node Simple Sed, Igawk Program, Extract Program, Miscellaneous Programs @subsection A Simple Stream Editor @cindex @code{sed} utility The @code{sed} utility is a ``stream editor,'' a program that reads a stream of data, makes changes to it, and passes the modified data on. It is often used to make global changes to a large file, or to a stream of data generated by a pipeline of commands. While @code{sed} is a complicated program in its own right, its most common use is to perform global substitutions in the middle of a pipeline: @example command1 < orig.data | sed 's/old/new/g' | command2 > result @end example Here, the @samp{s/old/new/g} tells @code{sed} to look for the regexp @samp{old} on each input line, and replace it with the text @samp{new}, globally (i.e.@: all the occurrences on a line). This is similar to @code{awk}'s @code{gsub} function (@pxref{String Functions, , Built-in Functions for String Manipulation}). The following program, @file{awksed.awk}, accepts at least two command line arguments; the pattern to look for and the text to replace it with. Any additional arguments are treated as data file names to process. If none are provided, the standard input is used. @cindex Brennan, Michael @cindex @code{awksed} @cindex simple stream editor @cindex stream editor, simple @example @c @group @c file eg/prog/awksed.awk # awksed.awk --- do s/foo/bar/g using just print # Thanks to Michael Brennan for the idea # Arnold Robbins, arnold@@gnu.org, Public Domain # August 1995 @group function usage() @{ print "usage: awksed pat repl [files...]" > "/dev/stderr" exit 1 @} @end group BEGIN @{ # validate arguments if (ARGC < 3) usage() RS = ARGV[1] ORS = ARGV[2] # don't use arguments as files ARGV[1] = ARGV[2] = "" @} # look ma, no hands! @{ if (RT == "") printf "%s", $0 else print @} @c endfile @c @end group @end example The program relies on @code{gawk}'s ability to have @code{RS} be a regexp and on the setting of @code{RT} to the actual text that terminated the record (@pxref{Records, ,How Input is Split into Records}). The idea is to have @code{RS} be the pattern to look for. @code{gawk} will automatically set @code{$0} to the text between matches of the pattern. This is text that we wish to keep, unmodified. Then, by setting @code{ORS} to the replacement text, a simple @code{print} statement will output the text we wish to keep, followed by the replacement text. There is one wrinkle to this scheme, which is what to do if the last record doesn't end with text that matches @code{RS}? Using a @code{print} statement unconditionally prints the replacement text, which is not correct. However, if the file did not end in text that matches @code{RS}, @code{RT} will be set to the null string. In this case, we can print @code{$0} using @code{printf} (@pxref{Printf, ,Using @code{printf} Statements for Fancier Printing}). The @code{BEGIN} rule handles the setup, checking for the right number of arguments, and calling @code{usage} if there is a problem. Then it sets @code{RS} and @code{ORS} from the command line arguments, and sets @code{ARGV[1]} and @code{ARGV[2]} to the null string, so that they will not be treated as file names (@pxref{ARGC and ARGV, , Using @code{ARGC} and @code{ARGV}}). The @code{usage} function prints an error message and exits. Finally, the single rule handles the printing scheme outlined above, using @code{print} or @code{printf} as appropriate, depending upon the value of @code{RT}. @ignore Exercise, compare the performance of this version with the more straightforward: BEGIN { pat = ARGV[1] repl = ARGV[2] ARGV[1] = ARGV[2] = "" } { gsub(pat, repl); print } Exercise: what are the advantages and disadvantages of this version vs. sed? Advantage: egrep regexps speed (?) Disadvantage: no & in replacement text Others? @end ignore @node Igawk Program, , Simple Sed, Miscellaneous Programs @subsection An Easy Way to Use Library Functions Using library functions in @code{awk} can be very beneficial. It encourages code re-use and the writing of general functions. Programs are smaller, and therefore clearer. However, using library functions is only easy when writing @code{awk} programs; it is painful when running them, requiring multiple @samp{-f} options. If @code{gawk} is unavailable, then so too is the @code{AWKPATH} environment variable and the ability to put @code{awk} functions into a library directory (@pxref{Options, ,Command Line Options}). It would be nice to be able to write programs like so: @example # library functions @@include getopt.awk @@include join.awk @dots{} # main program BEGIN @{ while ((c = getopt(ARGC, ARGV, "a:b:cde")) != -1) @dots{} @dots{} @} @end example The following program, @file{igawk.sh}, provides this service. It simulates @code{gawk}'s searching of the @code{AWKPATH} variable, and also allows @dfn{nested} includes; i.e.@: a file that has been included with @samp{@@include} can contain further @samp{@@include} statements. @code{igawk} will make an effort to only include files once, so that nested includes don't accidentally include a library function twice. @code{igawk} should behave externally just like @code{gawk}. This means it should accept all of @code{gawk}'s command line arguments, including the ability to have multiple source files specified via @samp{-f}, and the ability to mix command line and library source files. The program is written using the POSIX Shell (@code{sh}) command language. The way the program works is as follows: @enumerate @item Loop through the arguments, saving anything that doesn't represent @code{awk} source code for later, when the expanded program is run. @item For any arguments that do represent @code{awk} text, put the arguments into a temporary file that will be expanded. There are two cases. @enumerate a @item Literal text, provided with @samp{--source} or @samp{--source=}. This text is just echoed directly. The @code{echo} program will automatically supply a trailing newline. @item File names provided with @samp{-f}. We use a neat trick, and echo @samp{@@include @var{filename}} into the temporary file. Since the file inclusion program will work the way @code{gawk} does, this will get the text of the file included into the program at the correct point. @end enumerate @item Run an @code{awk} program (naturally) over the temporary file to expand @samp{@@include} statements. The expanded program is placed in a second temporary file. @item Run the expanded program with @code{gawk} and any other original command line arguments that the user supplied (such as the data file names). @end enumerate The initial part of the program turns on shell tracing if the first argument was @samp{debug}. Otherwise, a shell @code{trap} statement arranges to clean up any temporary files on program exit or upon an interrupt. @c 2e: For the temp file handling, go with Darrel's ig=${TMP:-/tmp}/igs.$$ @c 2e: or something as similar as possible. The next part loops through all the command line arguments. There are several cases of interest. @table @code @item -- This ends the arguments to @code{igawk}. Anything else should be passed on to the user's @code{awk} program without being evaluated. @item -W This indicates that the next option is specific to @code{gawk}. To make argument processing easier, the @samp{-W} is appended to the front of the remaining arguments and the loop continues. (This is an @code{sh} programming trick. Don't worry about it if you are not familiar with @code{sh}.) @item -v @itemx -F These are saved and passed on to @code{gawk}. @item -f @itemx --file @itemx --file= @itemx -Wfile= The file name is saved to the temporary file @file{/tmp/ig.s.$$} with an @samp{@@include} statement. The @code{sed} utility is used to remove the leading option part of the argument (e.g., @samp{--file=}). @item --source @itemx --source= @itemx -Wsource= The source text is echoed into @file{/tmp/ig.s.$$}. @item --version @itemx --version @itemx -Wversion @code{igawk} prints its version number, and runs @samp{gawk --version} to get the @code{gawk} version information, and then exits. @end table If none of @samp{-f}, @samp{--file}, @samp{-Wfile}, @samp{--source}, or @samp{-Wsource}, were supplied, then the first non-option argument should be the @code{awk} program. If there are no command line arguments left, @code{igawk} prints an error message and exits. Otherwise, the first argument is echoed into @file{/tmp/ig.s.$$}. In any case, after the arguments have been processed, @file{/tmp/ig.s.$$} contains the complete text of the original @code{awk} program. The @samp{$$} in @code{sh} represents the current process ID number. It is often used in shell programs to generate unique temporary file names. This allows multiple users to run @code{igawk} without worrying that the temporary file names will clash. @cindex @code{sed} utility Here's the program: @findex igawk.sh @example @c @group @c file eg/prog/igawk.sh #! /bin/sh # igawk --- like gawk but do @@include processing # Arnold Robbins, arnold@@gnu.org, Public Domain # July 1993 if [ "$1" = debug ] then set -x shift else # cleanup on exit, hangup, interrupt, quit, termination trap 'rm -f /tmp/ig.[se].$$' 0 1 2 3 15 fi while [ $# -ne 0 ] # loop over arguments do case $1 in --) shift; break;; -W) shift set -- -W"$@@" continue;; -[vF]) opts="$opts $1 '$2'" shift;; -[vF]*) opts="$opts '$1'" ;; -f) echo @@include "$2" >> /tmp/ig.s.$$ shift;; @group -f*) f=`echo "$1" | sed 's/-f//'` echo @@include "$f" >> /tmp/ig.s.$$ ;; @end group -?file=*) # -Wfile or --file f=`echo "$1" | sed 's/-.file=//'` echo @@include "$f" >> /tmp/ig.s.$$ ;; -?file) # get arg, $2 echo @@include "$2" >> /tmp/ig.s.$$ shift;; -?source=*) # -Wsource or --source t=`echo "$1" | sed 's/-.source=//'` echo "$t" >> /tmp/ig.s.$$ ;; -?source) # get arg, $2 echo "$2" >> /tmp/ig.s.$$ shift;; -?version) echo igawk: version 1.0 1>&2 gawk --version exit 0 ;; -[W-]*) opts="$opts '$1'" ;; *) break;; esac shift done if [ ! -s /tmp/ig.s.$$ ] then if [ -z "$1" ] then echo igawk: no program! 1>&2 exit 1 else echo "$1" > /tmp/ig.s.$$ shift fi fi # at this point, /tmp/ig.s.$$ has the program @c endfile @c @end group @end example The @code{awk} program to process @samp{@@include} directives reads through the program, one line at a time using @code{getline} (@pxref{Getline, ,Explicit Input with @code{getline}}). The input file names and @samp{@@include} statements are managed using a stack. As each @samp{@@include} is encountered, the current file name is ``pushed'' onto the stack, and the file named in the @samp{@@include} directive becomes the current file name. As each file is finished, the stack is ``popped,'' and the previous input file becomes the current input file again. The process is started by making the original file the first one on the stack. The @code{pathto} function does the work of finding the full path to a file. It simulates @code{gawk}'s behavior when searching the @code{AWKPATH} environment variable (@pxref{AWKPATH Variable, ,The @code{AWKPATH} Environment Variable}). If a file name has a @samp{/} in it, no path search is done. Otherwise, the file name is concatenated with the name of each directory in the path, and an attempt is made to open the generated file name. The only way in @code{awk} to test if a file can be read is to go ahead and try to read it with @code{getline}; that is what @code{pathto} does.@footnote{On some very old versions of @code{awk}, the test @samp{getline junk < t} can loop forever if the file exists but is empty. Caveat Emptor.} If the file can be read, it is closed, and the file name is returned. @ignore An alternative way to test for the file's existence would be to call @samp{system("test -r " t)}, which uses the @code{test} utility to see if the file exists and is readable. The disadvantage to this method is that it requires creating an extra process, and can thus be slightly slower. @end ignore @example @c @group @c file eg/prog/igawk.sh gawk -- ' # process @@include directives function pathto(file, i, t, junk) @{ if (index(file, "/") != 0) return file for (i = 1; i <= ndirs; i++) @{ t = (pathlist[i] "/" file) if ((getline junk < t) > 0) @{ # found it close(t) return t @} @} return "" @} @c endfile @c @end group @end example The main program is contained inside one @code{BEGIN} rule. The first thing it does is set up the @code{pathlist} array that @code{pathto} uses. After splitting the path on @samp{:}, null elements are replaced with @code{"."}, which represents the current directory. @example @group @c file eg/prog/igawk.sh BEGIN @{ path = ENVIRON["AWKPATH"] ndirs = split(path, pathlist, ":") for (i = 1; i <= ndirs; i++) @{ if (pathlist[i] == "") pathlist[i] = "." @} @c endfile @end group @end example The stack is initialized with @code{ARGV[1]}, which will be @file{/tmp/ig.s.$$}. The main loop comes next. Input lines are read in succession. Lines that do not start with @samp{@@include} are printed verbatim. If the line does start with @samp{@@include}, the file name is in @code{$2}. @code{pathto} is called to generate the full path. If it could not, then we print an error message and continue. The next thing to check is if the file has been included already. The @code{processed} array is indexed by the full file name of each included file, and it tracks this information for us. If the file has been seen, a warning message is printed. Otherwise, the new file name is pushed onto the stack and processing continues. Finally, when @code{getline} encounters the end of the input file, the file is closed and the stack is popped. When @code{stackptr} is less than zero, the program is done. @example @c @group @c file eg/prog/igawk.sh stackptr = 0 input[stackptr] = ARGV[1] # ARGV[1] is first file for (; stackptr >= 0; stackptr--) @{ while ((getline < input[stackptr]) > 0) @{ if (tolower($1) != "@@include") @{ print continue @} fpath = pathto($2) if (fpath == "") @{ printf("igawk:%s:%d: cannot find %s\n", \ input[stackptr], FNR, $2) > "/dev/stderr" continue @} @group if (! (fpath in processed)) @{ processed[fpath] = input[stackptr] input[++stackptr] = fpath @} else print $2, "included in", input[stackptr], \ "already included in", \ processed[fpath] > "/dev/stderr" @} @end group @group close(input[stackptr]) @} @}' /tmp/ig.s.$$ > /tmp/ig.e.$$ @end group @c endfile @c @end group @end example The last step is to call @code{gawk} with the expanded program and the original options and command line arguments that the user supplied. @code{gawk}'s exit status is passed back on to @code{igawk}'s calling program. @c this causes more problems than it solves, so leave it out. @ignore The special file @file{/dev/null} is passed as a data file to @code{gawk} to handle an interesting case. Suppose that the user's program only has a @code{BEGIN} rule, and there are no data files to read. The program should exit without reading any data files. However, suppose that an included library file defines an @code{END} rule of its own. In this case, @code{gawk} will hang, reading standard input. In order to avoid this, @file{/dev/null} is explicitly to the command line. Reading from @file{/dev/null} always returns an immediate end of file indication. @c Hmm. Add /dev/null if $# is 0? Still messes up ARGV. Sigh. @end ignore @example @c @group @c file eg/prog/igawk.sh eval gawk -f /tmp/ig.e.$$ $opts -- "$@@" exit $? @c endfile @c @end group @end example This version of @code{igawk} represents my third attempt at this program. There are three key simplifications that made the program work better. @enumerate @item Using @samp{@@include} even for the files named with @samp{-f} makes building the initial collected @code{awk} program much simpler; all the @samp{@@include} processing can be done once. @item The @code{pathto} function doesn't try to save the line read with @code{getline} when testing for the file's accessibility. Trying to save this line for use with the main program complicates things considerably. @c what problem does this engender though - exercise @c answer, reading from "-" or /dev/stdin @item Using a @code{getline} loop in the @code{BEGIN} rule does it all in one place. It is not necessary to call out to a separate loop for processing nested @samp{@@include} statements. @end enumerate Also, this program illustrates that it is often worthwhile to combine @code{sh} and @code{awk} programming together. You can usually accomplish quite a lot, without having to resort to low-level programming in C or C++, and it is frequently easier to do certain kinds of string and argument manipulation using the shell than it is in @code{awk}. Finally, @code{igawk} shows that it is not always necessary to add new features to a program; they can often be layered on top. With @code{igawk}, there is no real reason to build @samp{@@include} processing into @code{gawk} itself. As an additional example of this, consider the idea of having two files in a directory in the search path. @table @file @item default.awk This file would contain a set of default library functions, such as @code{getopt} and @code{assert}. @item site.awk This file would contain library functions that are specific to a site or installation, i.e.@: locally developed functions. Having a separate file allows @file{default.awk} to change with new @code{gawk} releases, without requiring the system administrator to update it each time by adding the local functions. @end table One user @c Karl Berry, karl@ileaf.com, 10/95 suggested that @code{gawk} be modified to automatically read these files upon startup. Instead, it would be very simple to modify @code{igawk} to do this. Since @code{igawk} can process nested @samp{@@include} directives, @file{default.awk} could simply contain @samp{@@include} statements for the desired library functions. @c Exercise: make this change @node Language History, Gawk Summary, Sample Programs, Top @chapter The Evolution of the @code{awk} Language This @value{DOCUMENT} describes the GNU implementation of @code{awk}, which follows the POSIX specification. Many @code{awk} users are only familiar with the original @code{awk} implementation in Version 7 Unix. (This implementation was the basis for @code{awk} in Berkeley Unix, through 4.3--Reno. The 4.4 release of Berkeley Unix uses @code{gawk} 2.15.2 for its version of @code{awk}.) This chapter briefly describes the evolution of the @code{awk} language, with cross references to other parts of the @value{DOCUMENT} where you can find more information. @menu * V7/SVR3.1:: The major changes between V7 and System V Release 3.1. * SVR4:: Minor changes between System V Releases 3.1 and 4. * POSIX:: New features from the POSIX standard. * BTL:: New features from the Bell Laboratories version of @code{awk}. * POSIX/GNU:: The extensions in @code{gawk} not in POSIX @code{awk}. @end menu @node V7/SVR3.1, SVR4, Language History, Language History @section Major Changes between V7 and SVR3.1 The @code{awk} language evolved considerably between the release of Version 7 Unix (1978) and the new version first made generally available in System V Release 3.1 (1987). This section summarizes the changes, with cross-references to further details. @itemize @bullet @item The requirement for @samp{;} to separate rules on a line (@pxref{Statements/Lines, ,@code{awk} Statements Versus Lines}). @item User-defined functions, and the @code{return} statement (@pxref{User-defined, ,User-defined Functions}). @item The @code{delete} statement (@pxref{Delete, ,The @code{delete} Statement}). @item The @code{do}-@code{while} statement (@pxref{Do Statement, ,The @code{do}-@code{while} Statement}). @item The built-in functions @code{atan2}, @code{cos}, @code{sin}, @code{rand} and @code{srand} (@pxref{Numeric Functions, ,Numeric Built-in Functions}). @item The built-in functions @code{gsub}, @code{sub}, and @code{match} (@pxref{String Functions, ,Built-in Functions for String Manipulation}). @item The built-in functions @code{close}, and @code{system} (@pxref{I/O Functions, ,Built-in Functions for Input/Output}). @item The @code{ARGC}, @code{ARGV}, @code{FNR}, @code{RLENGTH}, @code{RSTART}, and @code{SUBSEP} built-in variables (@pxref{Built-in Variables}). @item The conditional expression using the ternary operator @samp{?:} (@pxref{Conditional Exp, ,Conditional Expressions}). @item The exponentiation operator @samp{^} (@pxref{Arithmetic Ops, ,Arithmetic Operators}) and its assignment operator form @samp{^=} (@pxref{Assignment Ops, ,Assignment Expressions}). @item C-compatible operator precedence, which breaks some old @code{awk} programs (@pxref{Precedence, ,Operator Precedence (How Operators Nest)}). @item Regexps as the value of @code{FS} (@pxref{Field Separators, ,Specifying How Fields are Separated}), and as the third argument to the @code{split} function (@pxref{String Functions, ,Built-in Functions for String Manipulation}). @item Dynamic regexps as operands of the @samp{~} and @samp{!~} operators (@pxref{Regexp Usage, ,How to Use Regular Expressions}). @item The escape sequences @samp{\b}, @samp{\f}, and @samp{\r} (@pxref{Escape Sequences}). (Some vendors have updated their old versions of @code{awk} to recognize @samp{\r}, @samp{\b}, and @samp{\f}, but this is not something you can rely on.) @item Redirection of input for the @code{getline} function (@pxref{Getline, ,Explicit Input with @code{getline}}). @item Multiple @code{BEGIN} and @code{END} rules (@pxref{BEGIN/END, ,The @code{BEGIN} and @code{END} Special Patterns}). @item Multi-dimensional arrays (@pxref{Multi-dimensional, ,Multi-dimensional Arrays}). @end itemize @node SVR4, POSIX, V7/SVR3.1, Language History @section Changes between SVR3.1 and SVR4 @cindex @code{awk} language, V.4 version The System V Release 4 version of Unix @code{awk} added these features (some of which originated in @code{gawk}): @itemize @bullet @item The @code{ENVIRON} variable (@pxref{Built-in Variables}). @item Multiple @samp{-f} options on the command line (@pxref{Options, ,Command Line Options}). @item The @samp{-v} option for assigning variables before program execution begins (@pxref{Options, ,Command Line Options}). @item The @samp{--} option for terminating command line options. @item The @samp{\a}, @samp{\v}, and @samp{\x} escape sequences (@pxref{Escape Sequences}). @item A defined return value for the @code{srand} built-in function (@pxref{Numeric Functions, ,Numeric Built-in Functions}). @item The @code{toupper} and @code{tolower} built-in string functions for case translation (@pxref{String Functions, ,Built-in Functions for String Manipulation}). @item A cleaner specification for the @samp{%c} format-control letter in the @code{printf} function (@pxref{Control Letters, ,Format-Control Letters}). @item The ability to dynamically pass the field width and precision (@code{"%*.*d"}) in the argument list of the @code{printf} function (@pxref{Control Letters, ,Format-Control Letters}). @item The use of regexp constants such as @code{/foo/} as expressions, where they are equivalent to using the matching operator, as in @samp{$0 ~ /foo/} (@pxref{Using Constant Regexps, ,Using Regular Expression Constants}). @end itemize @node POSIX, BTL, SVR4, Language History @section Changes between SVR4 and POSIX @code{awk} The POSIX Command Language and Utilities standard for @code{awk} introduced the following changes into the language: @itemize @bullet @item The use of @samp{-W} for implementation-specific options. @item The use of @code{CONVFMT} for controlling the conversion of numbers to strings (@pxref{Conversion, ,Conversion of Strings and Numbers}). @item The concept of a numeric string, and tighter comparison rules to go with it (@pxref{Typing and Comparison, ,Variable Typing and Comparison Expressions}). @item More complete documentation of many of the previously undocumented features of the language. @end itemize The following common extensions are not permitted by the POSIX standard: @c IMPORTANT! Keep this list in sync with the one in node Options @itemize @bullet @item @code{\x} escape sequences are not recognized (@pxref{Escape Sequences}). @item Newlines do not act as whitespace to separate fields when @code{FS} is equal to a single space. @item The synonym @code{func} for the keyword @code{function} is not recognized (@pxref{Definition Syntax, ,Function Definition Syntax}). @item The operators @samp{**} and @samp{**=} cannot be used in place of @samp{^} and @samp{^=} (@pxref{Arithmetic Ops, ,Arithmetic Operators}, and also @pxref{Assignment Ops, ,Assignment Expressions}). @item Specifying @samp{-Ft} on the command line does not set the value of @code{FS} to be a single tab character (@pxref{Field Separators, ,Specifying How Fields are Separated}). @item The @code{fflush} built-in function is not supported (@pxref{I/O Functions, , Built-in Functions for Input/Output}). @end itemize @node BTL, POSIX/GNU, POSIX, Language History @section Extensions in the Bell Laboratories @code{awk} @cindex Kernighan, Brian Brian Kernighan, one of the original designers of Unix @code{awk}, has made his version available via anonymous @code{ftp} (@pxref{Other Versions, ,Other Freely Available @code{awk} Implementations}). This section describes extensions in his version of @code{awk} that are not in POSIX @code{awk}. @itemize @bullet @item The @samp{-mf @var{NNN}} and @samp{-mr @var{NNN}} command line options to set the maximum number of fields, and the maximum record size, respectively (@pxref{Options, ,Command Line Options}). @item The @code{fflush} built-in function for flushing buffered output (@pxref{I/O Functions, ,Built-in Functions for Input/Output}). @ignore @item The @code{SYMTAB} array, that allows access to the internal symbol table of @code{awk}. This feature is not documented, largely because it is somewhat shakily implemented. For instance, you cannot access arrays or array elements through it. @end ignore @end itemize @node POSIX/GNU, , BTL, Language History @section Extensions in @code{gawk} Not in POSIX @code{awk} @cindex compatibility mode The GNU implementation, @code{gawk}, adds a number of features. This sections lists them in the order they were added to @code{gawk}. They can all be disabled with either the @samp{--traditional} or @samp{--posix} options (@pxref{Options, ,Command Line Options}). Version 2.10 of @code{gawk} introduced these features: @itemize @bullet @item The @code{AWKPATH} environment variable for specifying a path search for the @samp{-f} command line option (@pxref{Options, ,Command Line Options}). @item The @code{IGNORECASE} variable and its effects (@pxref{Case-sensitivity, ,Case-sensitivity in Matching}). @item The @file{/dev/stdin}, @file{/dev/stdout}, @file{/dev/stderr}, and @file{/dev/fd/@var{n}} file name interpretation (@pxref{Special Files, ,Special File Names in @code{gawk}}). @end itemize Version 2.13 of @code{gawk} introduced these features: @itemize @bullet @item The @code{FIELDWIDTHS} variable and its effects (@pxref{Constant Size, ,Reading Fixed-width Data}). @item The @code{systime} and @code{strftime} built-in functions for obtaining and printing time stamps (@pxref{Time Functions, ,Functions for Dealing with Time Stamps}). @item The @samp{-W lint} option to provide source code and run time error and portability checking (@pxref{Options, ,Command Line Options}). @item The @samp{-W compat} option to turn off these extensions (@pxref{Options, ,Command Line Options}). @item The @samp{-W posix} option for full POSIX compliance (@pxref{Options, ,Command Line Options}). @end itemize Version 2.14 of @code{gawk} introduced these features: @itemize @bullet @item The @code{next file} statement for skipping to the next data file (@pxref{Nextfile Statement, ,The @code{nextfile} Statement}). @end itemize Version 2.15 of @code{gawk} introduced these features: @itemize @bullet @item The @code{ARGIND} variable, that tracks the movement of @code{FILENAME} through @code{ARGV} (@pxref{Built-in Variables}). @item The @code{ERRNO} variable, that contains the system error message when @code{getline} returns @minus{}1, or when @code{close} fails (@pxref{Built-in Variables}). @item The ability to use GNU-style long named options that start with @samp{--} (@pxref{Options, ,Command Line Options}). @item The @samp{--source} option for mixing command line and library file source code (@pxref{Options, ,Command Line Options}). @item The @file{/dev/pid}, @file{/dev/ppid}, @file{/dev/pgrpid}, and @file{/dev/user} file name interpretation (@pxref{Special Files, ,Special File Names in @code{gawk}}). @end itemize Version 3.0 of @code{gawk} introduced these features: @itemize @bullet @item The @code{next file} statement became @code{nextfile} (@pxref{Nextfile Statement, ,The @code{nextfile} Statement}). @item The @samp{--lint-old} option to warn about constructs that are not available in the original Version 7 Unix version of @code{awk} (@pxref{V7/SVR3.1, , Major Changes between V7 and SVR3.1}). @item The @samp{--traditional} option was added as a better name for @samp{--compat} (@pxref{Options, ,Command Line Options}). @item The ability for @code{FS} to be a null string, and for the third argument to @code{split} to be the null string (@pxref{Single Character Fields, , Making Each Character a Separate Field}). @item The ability for @code{RS} to be a regexp (@pxref{Records, , How Input is Split into Records}). @item The @code{RT} variable (@pxref{Records, , How Input is Split into Records}). @item The @code{gensub} function for more powerful text manipulation (@pxref{String Functions, , Built-in Functions for String Manipulation}). @item The @code{strftime} function acquired a default time format, allowing it to be called with no arguments (@pxref{Time Functions, , Functions for Dealing with Time Stamps}). @item Full support for both POSIX and GNU regexps (@pxref{Regexp, , Regular Expressions}). @item The @samp{--re-interval} option to provide interval expressions in regexps (@pxref{Regexp Operators, , Regular Expression Operators}). @item @code{IGNORECASE} changed, now applying to string comparison as well as regexp operations (@pxref{Case-sensitivity, ,Case-sensitivity in Matching}). @item The @samp{-m} option and the @code{fflush} function from the Bell Labs research version of @code{awk} (@pxref{Options, ,Command Line Options}; also @pxref{I/O Functions, ,Built-in Functions for Input/Output}). @item The use of GNU Autoconf to control the configuration process (@pxref{Quick Installation, , Compiling @code{gawk} for Unix}). @item Amiga support (@pxref{Amiga Installation, ,Installing @code{gawk} on an Amiga}). @c XXX ADD MORE STUFF HERE @end itemize @node Gawk Summary, Installation, Language History, Top @appendix @code{gawk} Summary This appendix provides a brief summary of the @code{gawk} command line and the @code{awk} language. It is designed to serve as ``quick reference.'' It is therefore terse, but complete. @menu * Command Line Summary:: Recapitulation of the command line. * Language Summary:: A terse review of the language. * Variables/Fields:: Variables, fields, and arrays. * Rules Summary:: Patterns and Actions, and their component parts. * Actions Summary:: Quick overview of actions. * Functions Summary:: Defining and calling functions. * Historical Features:: Some undocumented but supported ``features''. @end menu @node Command Line Summary, Language Summary, Gawk Summary, Gawk Summary @appendixsec Command Line Options Summary The command line consists of options to @code{gawk} itself, the @code{awk} program text (if not supplied via the @samp{-f} option), and values to be made available in the @code{ARGC} and @code{ARGV} predefined @code{awk} variables: @example gawk @r{[@var{POSIX or GNU style options}]} -f @var{source-file} @r{[@code{--}]} @var{file} @dots{} gawk @r{[@var{POSIX or GNU style options}]} @r{[@code{--}]} '@var{program}' @var{file} @dots{} @end example The options that @code{gawk} accepts are: @table @code @item -F @var{fs} @itemx --field-separator @var{fs} Use @var{fs} for the input field separator (the value of the @code{FS} predefined variable). @item -f @var{program-file} @itemx --file @var{program-file} Read the @code{awk} program source from the file @var{program-file}, instead of from the first command line argument. @item -mf @var{NNN} @itemx -mr @var{NNN} The @samp{f} flag sets the maximum number of fields, and the @samp{r} flag sets the maximum record size. These options are ignored by @code{gawk}, since @code{gawk} has no predefined limits; they are only for compatibility with the Bell Labs research version of Unix @code{awk}. @item -v @var{var}=@var{val} @itemx --assign @var{var}=@var{val} Assign the variable @var{var} the value @var{val} before program execution begins. @item -W traditional @itemx -W compat @itemx --traditional @itemx --compat Use compatibility mode, in which @code{gawk} extensions are turned off. @item -W copyleft @itemx -W copyright @itemx --copyleft @itemx --copyright Print the short version of the General Public License on the standard output, and exit. This option may disappear in a future version of @code{gawk}. @item -W help @itemx -W usage @itemx --help @itemx --usage Print a relatively short summary of the available options on the standard output, and exit. @item -W lint @itemx --lint Give warnings about dubious or non-portable @code{awk} constructs. @item -W lint-old @itemx --lint-old Warn about constructs that are not available in the original Version 7 Unix version of @code{awk}. @item -W posix @itemx --posix Use POSIX compatibility mode, in which @code{gawk} extensions are turned off and additional restrictions apply. @item -W re-interval @itemx --re-interval Allow interval expressions (@pxref{Regexp Operators, , Regular Expression Operators}), in regexps. @item -W source=@var{program-text} @itemx --source @var{program-text} Use @var{program-text} as @code{awk} program source code. This option allows mixing command line source code with source code from files, and is particularly useful for mixing command line programs with library functions. @item -W version @itemx --version Print version information for this particular copy of @code{gawk} on the error output. @item -- Signal the end of options. This is useful to allow further arguments to the @code{awk} program itself to start with a @samp{-}. This is mainly for consistency with POSIX argument parsing conventions. @end table Any other options are flagged as invalid, but are otherwise ignored. @xref{Options, ,Command Line Options}, for more details. @node Language Summary, Variables/Fields, Command Line Summary, Gawk Summary @appendixsec Language Summary An @code{awk} program consists of a sequence of zero or more pattern-action statements and optional function definitions. One or the other of the pattern and action may be omitted. @example @var{pattern} @{ @var{action statements} @} @var{pattern} @{ @var{action statements} @} function @var{name}(@var{parameter list}) @{ @var{action statements} @} @end example @code{gawk} first reads the program source from the @var{program-file}(s), if specified, or from the first non-option argument on the command line. The @samp{-f} option may be used multiple times on the command line. @code{gawk} reads the program text from all the @var{program-file} files, effectively concatenating them in the order they are specified. This is useful for building libraries of @code{awk} functions, without having to include them in each new @code{awk} program that uses them. To use a library function in a file from a program typed in on the command line, specify @samp{--source '@var{program}'}, and type your program in between the single quotes. @xref{Options, ,Command Line Options}. The environment variable @code{AWKPATH} specifies a search path to use when finding source files named with the @samp{-f} option. The default path, which is @samp{.:/usr/local/share/awk}@footnote{The path may use a directory other than @file{/usr/local/share/awk}, depending upon how @code{gawk} was built and installed.} is used if @code{AWKPATH} is not set. If a file name given to the @samp{-f} option contains a @samp{/} character, no path search is performed. @xref{AWKPATH Variable, ,The @code{AWKPATH} Environment Variable}. @code{gawk} compiles the program into an internal form, and then proceeds to read each file named in the @code{ARGV} array. The initial values of @code{ARGV} come from the command line arguments. If there are no files named on the command line, @code{gawk} reads the standard input. If a ``file'' named on the command line has the form @samp{@var{var}=@var{val}}, it is treated as a variable assignment: the variable @var{var} is assigned the value @var{val}. If any of the files have a value that is the null string, that element in the list is skipped. For each record in the input, @code{gawk} tests to see if it matches any @var{pattern} in the @code{awk} program. For each pattern that the record matches, the associated @var{action} is executed. @node Variables/Fields, Rules Summary, Language Summary, Gawk Summary @appendixsec Variables and Fields @code{awk} variables are not declared; they come into existence when they are first used. Their values are either floating-point numbers or strings. @code{awk} also has one-dimensional arrays; multiple-dimensional arrays may be simulated. There are several predefined variables that @code{awk} sets as a program runs; these are summarized below. @menu * Fields Summary:: Input field splitting. * Built-in Summary:: @code{awk}'s built-in variables. * Arrays Summary:: Using arrays. * Data Type Summary:: Values in @code{awk} are numbers or strings. @end menu @node Fields Summary, Built-in Summary, Variables/Fields, Variables/Fields @appendixsubsec Fields As each input line is read, @code{gawk} splits the line into @var{fields}, using the value of the @code{FS} variable as the field separator. If @code{FS} is a single character, fields are separated by that character. Otherwise, @code{FS} is expected to be a full regular expression. In the special case that @code{FS} is a single space, fields are separated by runs of spaces, tabs and/or newlines.@footnote{In POSIX @code{awk}, newline does not separate fields.} If @code{FS} is the null string (@code{""}), then each individual character in the record becomes a separate field. Note that the value of @code{IGNORECASE} (@pxref{Case-sensitivity, ,Case-sensitivity in Matching}) also affects how fields are split when @code{FS} is a regular expression. Each field in the input line may be referenced by its position, @code{$1}, @code{$2}, and so on. @code{$0} is the whole line. The value of a field may be assigned to as well. Field numbers need not be constants: @example n = 5 print $n @end example @noindent prints the fifth field in the input line. The variable @code{NF} is set to the total number of fields in the input line. References to non-existent fields (i.e.@: fields after @code{$NF}) return the null string. However, assigning to a non-existent field (e.g., @code{$(NF+2) = 5}) increases the value of @code{NF}, creates any intervening fields with the null string as their value, and causes the value of @code{$0} to be recomputed, with the fields being separated by the value of @code{OFS}. Decrementing @code{NF} causes the values of fields past the new value to be lost, and the value of @code{$0} to be recomputed, with the fields being separated by the value of @code{OFS}. @xref{Reading Files, ,Reading Input Files}. @node Built-in Summary, Arrays Summary, Fields Summary, Variables/Fields @appendixsubsec Built-in Variables @code{gawk}'s built-in variables are: @table @code @item ARGC The number of elements in @code{ARGV}. See below for what is actually included in @code{ARGV}. @item ARGIND The index in @code{ARGV} of the current file being processed. When @code{gawk} is processing the input data files, it is always true that @samp{FILENAME == ARGV[ARGIND]}. @item ARGV The array of command line arguments. The array is indexed from zero to @code{ARGC} @minus{} 1. Dynamically changing @code{ARGC} and the contents of @code{ARGV} can control the files used for data. A null-valued element in @code{ARGV} is ignored. @code{ARGV} does not include the options to @code{awk} or the text of the @code{awk} program itself. @item CONVFMT The conversion format to use when converting numbers to strings. @item FIELDWIDTHS A space separated list of numbers describing the fixed-width input data. @item ENVIRON An array of environment variable values. The array is indexed by variable name, each element being the value of that variable. Thus, the environment variable @code{HOME} is @code{ENVIRON["HOME"]}. One possible value might be @file{/home/arnold}. Changing this array does not affect the environment seen by programs which @code{gawk} spawns via redirection or the @code{system} function. (This may change in a future version of @code{gawk}.) Some operating systems do not have environment variables. The @code{ENVIRON} array is empty when running on these systems. @item ERRNO The system error message when an error occurs using @code{getline} or @code{close}. @item FILENAME The name of the current input file. If no files are specified on the command line, the value of @code{FILENAME} is the null string. @item FNR The input record number in the current input file. @item FS The input field separator, a space by default. @item IGNORECASE The case-sensitivity flag for string comparisons and regular expression operations. If @code{IGNORECASE} has a non-zero value, then pattern matching in rules, record separating with @code{RS}, field splitting with @code{FS}, regular expression matching with @samp{~} and @samp{!~}, and the @code{gensub}, @code{gsub}, @code{index}, @code{match}, @code{split} and @code{sub} built-in functions all ignore case when doing regular expression operations, and all string comparisons are done ignoring case. The value of @code{IGNORECASE} does @emph{not} affect array subscripting. @item NF The number of fields in the current input record. @item NR The total number of input records seen so far. @item OFMT The output format for numbers for the @code{print} statement, @code{"%.6g"} by default. @item OFS The output field separator, a space by default. @item ORS The output record separator, by default a newline. @item RS The input record separator, by default a newline. If @code{RS} is set to the null string, then records are separated by blank lines. When @code{RS} is set to the null string, then the newline character always acts as a field separator, in addition to whatever value @code{FS} may have. If @code{RS} is set to a multi-character string, it denotes a regexp; input text matching the regexp separates records. @item RT The input text that matched the text denoted by @code{RS}, the record separator. @item RSTART The index of the first character last matched by @code{match}; zero if no match. @item RLENGTH The length of the string last matched by @code{match}; @minus{}1 if no match. @item SUBSEP The string used to separate multiple subscripts in array elements, by default @code{"\034"}. @end table @xref{Built-in Variables}, for more information. @node Arrays Summary, Data Type Summary, Built-in Summary, Variables/Fields @appendixsubsec Arrays Arrays are subscripted with an expression between square brackets (@samp{[} and @samp{]}). Array subscripts are @emph{always} strings; numbers are converted to strings as necessary, following the standard conversion rules (@pxref{Conversion, ,Conversion of Strings and Numbers}). If you use multiple expressions separated by commas inside the square brackets, then the array subscript is a string consisting of the concatenation of the individual subscript values, converted to strings, separated by the subscript separator (the value of @code{SUBSEP}). The special operator @code{in} may be used in a conditional context to see if an array has an index consisting of a particular value. @example if (val in array) print array[val] @end example If the array has multiple subscripts, use @samp{(i, j, @dots{}) in @var{array}} to test for existence of an element. The @code{in} construct may also be used in a @code{for} loop to iterate over all the elements of an array. @xref{Scanning an Array, ,Scanning All Elements of an Array}. You can remove an element from an array using the @code{delete} statement. You can clear an entire array using @samp{delete @var{array}}. @xref{Arrays, ,Arrays in @code{awk}}. @node Data Type Summary, , Arrays Summary, Variables/Fields @appendixsubsec Data Types The value of an @code{awk} expression is always either a number or a string. Some contexts (such as arithmetic operators) require numeric values. They convert strings to numbers by interpreting the text of the string as a number. If the string does not look like a number, it converts to zero. Other contexts (such as concatenation) require string values. They convert numbers to strings by effectively printing them with @code{sprintf}. @xref{Conversion, ,Conversion of Strings and Numbers}, for the details. To force conversion of a string value to a number, simply add zero to it. If the value you start with is already a number, this does not change it. To force conversion of a numeric value to a string, concatenate it with the null string. Comparisons are done numerically if both operands are numeric, or if one is numeric and the other is a numeric string. Otherwise one or both operands are converted to strings and a string comparison is performed. Fields, @code{getline} input, @code{FILENAME}, @code{ARGV} elements, @code{ENVIRON} elements and the elements of an array created by @code{split} are the only items that can be numeric strings. String constants, such as @code{"3.1415927"} are not numeric strings, they are string constants. The full rules for comparisons are described in @ref{Typing and Comparison, ,Variable Typing and Comparison Expressions}. Uninitialized variables have the string value @code{""} (the null, or empty, string). In contexts where a number is required, this is equivalent to zero. @xref{Variables}, for more information on variable naming and initialization; @pxref{Conversion, ,Conversion of Strings and Numbers}, for more information on how variable values are interpreted. @node Rules Summary, Actions Summary, Variables/Fields, Gawk Summary @appendixsec Patterns @menu * Pattern Summary:: Quick overview of patterns. * Regexp Summary:: Quick overview of regular expressions. @end menu An @code{awk} program is mostly composed of rules, each consisting of a pattern followed by an action. The action is enclosed in @samp{@{} and @samp{@}}. Either the pattern may be missing, or the action may be missing, but not both. If the pattern is missing, the action is executed for every input record. A missing action is equivalent to @samp{@w{@{ print @}}}, which prints the entire line. @c These paragraphs repeated for both patterns and actions. I don't @c like this, but I also don't see any way around it. Update both copies @c if they need fixing. Comments begin with the @samp{#} character, and continue until the end of the line. Blank lines may be used to separate statements. Statements normally end with a newline; however, this is not the case for lines ending in a @samp{,}, @samp{@{}, @samp{?}, @samp{:}, @samp{&&}, or @samp{||}. Lines ending in @code{do} or @code{else} also have their statements automatically continued on the following line. In other cases, a line can be continued by ending it with a @samp{\}, in which case the newline is ignored. Multiple statements may be put on one line by separating each one with a @samp{;}. This applies to both the statements within the action part of a rule (the usual case), and to the rule statements. @xref{Comments, ,Comments in @code{awk} Programs}, for information on @code{awk}'s commenting convention; @pxref{Statements/Lines, ,@code{awk} Statements Versus Lines}, for a description of the line continuation mechanism in @code{awk}. @node Pattern Summary, Regexp Summary, Rules Summary, Rules Summary @appendixsubsec Pattern Summary @code{awk} patterns may be one of the following: @example /@var{regular expression}/ @var{relational expression} @var{pattern} && @var{pattern} @var{pattern} || @var{pattern} @var{pattern} ? @var{pattern} : @var{pattern} (@var{pattern}) ! @var{pattern} @var{pattern1}, @var{pattern2} BEGIN END @end example @code{BEGIN} and @code{END} are two special kinds of patterns that are not tested against the input. The action parts of all @code{BEGIN} rules are concatenated as if all the statements had been written in a single @code{BEGIN} rule. They are executed before any of the input is read. Similarly, all the @code{END} rules are concatenated, and executed when all the input is exhausted (or when an @code{exit} statement is executed). @code{BEGIN} and @code{END} patterns cannot be combined with other patterns in pattern expressions. @code{BEGIN} and @code{END} rules cannot have missing action parts. For @code{/@var{regular-expression}/} patterns, the associated statement is executed for each input record that matches the regular expression. Regular expressions are summarized below. A @var{relational expression} may use any of the operators defined below in the section on actions. These generally test whether certain fields match certain regular expressions. The @samp{&&}, @samp{||}, and @samp{!} operators are logical ``and,'' logical ``or,'' and logical ``not,'' respectively, as in C. They do short-circuit evaluation, also as in C, and are used for combining more primitive pattern expressions. As in most languages, parentheses may be used to change the order of evaluation. The @samp{?:} operator is like the same operator in C. If the first pattern matches, then the second pattern is matched against the input record; otherwise, the third is matched. Only one of the second and third patterns is matched. The @samp{@var{pattern1}, @var{pattern2}} form of a pattern is called a range pattern. It matches all input lines starting with a line that matches @var{pattern1}, and continuing until a line that matches @var{pattern2}, inclusive. A range pattern cannot be used as an operand of any of the pattern operators. @xref{Pattern Overview, ,Pattern Elements}. @node Regexp Summary, , Pattern Summary, Rules Summary @appendixsubsec Regular Expressions Regular expressions are based on POSIX EREs (extended regular expressions). The escape sequences allowed in string constants are also valid in regular expressions (@pxref{Escape Sequences}). Regexps are composed of characters as follows: @table @code @item @var{c} matches the character @var{c} (assuming @var{c} is none of the characters listed below). @item \@var{c} matches the literal character @var{c}. @item . matches any character, @emph{including} newline. In strict POSIX mode, @samp{.} does not match the @sc{nul} character, which is a character with all bits equal to zero. @item ^ matches the beginning of a string. @item $ matches the end of a string. @item [@var{abc}@dots{}] matches any of the characters @var{abc}@dots{} (character list). @item [[:@var{class}:]] matches any character in the character class @var{class}. Allowable classes are @code{alnum}, @code{alpha}, @code{blank}, @code{cntrl}, @code{digit}, @code{graph}, @code{lower}, @code{print}, @code{punct}, @code{space}, @code{upper}, and @code{xdigit}. @item [[.@var{symbol}.]] matches the multi-character collating symbol @var{symbol}. @code{gawk} does not currently support collating symbols. @item [[=@var{classname}=]] matches any of the equivalent characters in the current locale named by the equivalence class @var{classname}. @code{gawk} does not currently support equivalence classes. @item [^@var{abc}@dots{}] matches any character except @var{abc}@dots{} (negated character list). @item @var{r1}|@var{r2} matches either @var{r1} or @var{r2} (alternation). @item @var{r1r2} matches @var{r1}, and then @var{r2} (concatenation). @item @var{r}+ matches one or more @var{r}'s. @item @var{r}* matches zero or more @var{r}'s. @item @var{r}? matches zero or one @var{r}'s. @item (@var{r}) matches @var{r} (grouping). @item @var{r}@{@var{n}@} @itemx @var{r}@{@var{n},@} @itemx @var{r}@{@var{n},@var{m}@} matches at least @var{n}, @var{n} to any number, or @var{n} to @var{m} occurrences of @var{r} (interval expressions). @item \y matches the empty string at either the beginning or the end of a word. @item \B matches the empty string within a word. @item \< matches the empty string at the beginning of a word. @item \> matches the empty string at the end of a word. @item \w matches any word-constituent character (alphanumeric characters and the underscore). @item \W matches any character that is not word-constituent. @item \` matches the empty string at the beginning of a buffer (same as a string in @code{gawk}). @item \' matches the empty string at the end of a buffer. @end table The various command line options control how @code{gawk} interprets characters in regexps. @c NOTE!!! Keep this in sync with the same table in the regexp chapter! @table @asis @item No options In the default case, @code{gawk} provide all the facilities of POSIX regexps and the GNU regexp operators described above. However, interval expressions are not supported. @item @code{--posix} Only POSIX regexps are supported, the GNU operators are not special (e.g., @samp{\w} matches a literal @samp{w}). Interval expressions are allowed. @item @code{--traditional} Traditional Unix @code{awk} regexps are matched. The GNU operators are not special, interval expressions are not available, and neither are the POSIX character classes (@code{[[:alnum:]]} and so on). Characters described by octal and hexadecimal escape sequences are treated literally, even if they represent regexp metacharacters. @item @code{--re-interval} Allow interval expressions in regexps, even if @samp{--traditional} has been provided. @end table @xref{Regexp, ,Regular Expressions}. @node Actions Summary, Functions Summary, Rules Summary, Gawk Summary @appendixsec Actions Action statements are enclosed in braces, @samp{@{} and @samp{@}}. A missing action statement is equivalent to @samp{@w{@{ print @}}}. Action statements consist of the usual assignment, conditional, and looping statements found in most languages. The operators, control statements, and Input/Output statements available are similar to those in C. @c These paragraphs repeated for both patterns and actions. I don't @c like this, but I also don't see any way around it. Update both copies @c if they need fixing. Comments begin with the @samp{#} character, and continue until the end of the line. Blank lines may be used to separate statements. Statements normally end with a newline; however, this is not the case for lines ending in a @samp{,}, @samp{@{}, @samp{?}, @samp{:}, @samp{&&}, or @samp{||}. Lines ending in @code{do} or @code{else} also have their statements automatically continued on the following line. In other cases, a line can be continued by ending it with a @samp{\}, in which case the newline is ignored. Multiple statements may be put on one line by separating each one with a @samp{;}. This applies to both the statements within the action part of a rule (the usual case), and to the rule statements. @xref{Comments, ,Comments in @code{awk} Programs}, for information on @code{awk}'s commenting convention; @pxref{Statements/Lines, ,@code{awk} Statements Versus Lines}, for a description of the line continuation mechanism in @code{awk}. @menu * Operator Summary:: @code{awk} operators. * Control Flow Summary:: The control statements. * I/O Summary:: The I/O statements. * Printf Summary:: A summary of @code{printf}. * Special File Summary:: Special file names interpreted internally. * Built-in Functions Summary:: Built-in numeric and string functions. * Time Functions Summary:: Built-in time functions. * String Constants Summary:: Escape sequences in strings. @end menu @node Operator Summary, Control Flow Summary, Actions Summary, Actions Summary @appendixsubsec Operators The operators in @code{awk}, in order of decreasing precedence, are: @table @code @item (@dots{}) Grouping. @item $ Field reference. @item ++ -- Increment and decrement, both prefix and postfix. @item ^ Exponentiation (@samp{**} may also be used, and @samp{**=} for the assignment operator, but they are not specified in the POSIX standard). @item + - ! Unary plus, unary minus, and logical negation. @item * / % Multiplication, division, and modulus. @item + - Addition and subtraction. @item @var{space} String concatenation. @item < <= > >= != == The usual relational operators. @item ~ !~ Regular expression match, negated match. @item in Array membership. @item && Logical ``and''. @item || Logical ``or''. @item ?: A conditional expression. This has the form @samp{@var{expr1} ? @var{expr2} : @var{expr3}}. If @var{expr1} is true, the value of the expression is @var{expr2}; otherwise it is @var{expr3}. Only one of @var{expr2} and @var{expr3} is evaluated. @item = += -= *= /= %= ^= Assignment. Both absolute assignment (@code{@var{var}=@var{value}}) and operator assignment (the other forms) are supported. @end table @xref{Expressions}. @node Control Flow Summary, I/O Summary, Operator Summary, Actions Summary @appendixsubsec Control Statements The control statements are as follows: @example if (@var{condition}) @var{statement} @r{[} else @var{statement} @r{]} while (@var{condition}) @var{statement} do @var{statement} while (@var{condition}) for (@var{expr1}; @var{expr2}; @var{expr3}) @var{statement} for (@var{var} in @var{array}) @var{statement} break continue delete @var{array}[@var{index}] delete @var{array} exit @r{[} @var{expression} @r{]} @{ @var{statements} @} @end example @xref{Statements, ,Control Statements in Actions}. @node I/O Summary, Printf Summary, Control Flow Summary, Actions Summary @appendixsubsec I/O Statements The Input/Output statements are as follows: @table @code @item getline Set @code{$0} from next input record; set @code{NF}, @code{NR}, @code{FNR}. @xref{Getline, ,Explicit Input with @code{getline}}. @item getline <@var{file} Set @code{$0} from next record of @var{file}; set @code{NF}. @item getline @var{var} Set @var{var} from next input record; set @code{NR}, @code{FNR}. @item getline @var{var} <@var{file} Set @var{var} from next record of @var{file}. @item @var{command} | getline Run @var{command}, piping its output into @code{getline}; sets @code{$0}, @code{NF}, @code{NR}. @item @var{command} | getline @code{var} Run @var{command}, piping its output into @code{getline}; sets @var{var}. @item next Stop processing the current input record. The next input record is read and processing starts over with the first pattern in the @code{awk} program. If the end of the input data is reached, the @code{END} rule(s), if any, are executed. @xref{Next Statement, ,The @code{next} Statement}. @item nextfile Stop processing the current input file. The next input record read comes from the next input file. @code{FILENAME} is updated, @code{FNR} is set to one, @code{ARGIND} is incremented, and processing starts over with the first pattern in the @code{awk} program. If the end of the input data is reached, the @code{END} rule(s), if any, are executed. Earlier versions of @code{gawk} used @samp{next file}; this usage is still supported, but is considered to be deprecated. @xref{Nextfile Statement, ,The @code{nextfile} Statement}. @item print Prints the current record. @xref{Printing, ,Printing Output}. @item print @var{expr-list} Prints expressions. @item print @var{expr-list} > @var{file} Prints expressions to @var{file}. If @var{file} does not exist, it is created. If it does exist, its contents are deleted the first time the @code{print} is executed. @item print @var{expr-list} >> @var{file} Prints expressions to @var{file}. The previous contents of @var{file} are retained, and the output of @code{print} is appended to the file. @item print @var{expr-list} | @var{command} Prints expressions, sending the output down a pipe to @var{command}. The pipeline to the command stays open until the @code{close} function is called. @item printf @var{fmt, expr-list} Format and print. @item printf @var{fmt, expr-list} > file Format and print to @var{file}. If @var{file} does not exist, it is created. If it does exist, its contents are deleted the first time the @code{printf} is executed. @item printf @var{fmt, expr-list} >> @var{file} Format and print to @var{file}. The previous contents of @var{file} are retained, and the output of @code{printf} is appended to the file. @item printf @var{fmt, expr-list} | @var{command} Format and print, sending the output down a pipe to @var{command}. The pipeline to the command stays open until the @code{close} function is called. @end table @code{getline} returns zero on end of file, and @minus{}1 on an error. In the event of an error, @code{getline} will set @code{ERRNO} to the value of a system-dependent string that describes the error. @node Printf Summary, Special File Summary, I/O Summary, Actions Summary @appendixsubsec @code{printf} Summary Conversion specification have the form @code{%}[@var{flag}][@var{width}][@code{.}@var{prec}]@var{format}. @c whew! Items in brackets are optional. The @code{awk} @code{printf} statement and @code{sprintf} function accept the following conversion specification formats: @table @code @item %c An ASCII character. If the argument used for @samp{%c} is numeric, it is treated as a character and printed. Otherwise, the argument is assumed to be a string, and the only first character of that string is printed. @item %d @itemx %i A decimal number (the integer part). @item %e @itemx %E A floating point number of the form @samp{@r{[}-@r{]}d.dddddde@r{[}+-@r{]}dd}. The @samp{%E} format uses @samp{E} instead of @samp{e}. @item %f A floating point number of the form @r{[}@code{-}@r{]}@code{ddd.dddddd}. @item %g @itemx %G Use either the @samp{%e} or @samp{%f} formats, whichever produces a shorter string, with non-significant zeros suppressed. @samp{%G} will use @samp{%E} instead of @samp{%e}. @item %o An unsigned octal number (again, an integer). @item %s A character string. @item %x @itemx %X An unsigned hexadecimal number (an integer). The @samp{%X} format uses @samp{A} through @samp{F} instead of @samp{a} through @samp{f} for decimal 10 through 15. @item %% A single @samp{%} character; no argument is converted. @end table There are optional, additional parameters that may lie between the @samp{%} and the control letter: @table @code @item - The expression should be left-justified within its field. @item @var{space} For numeric conversions, prefix positive values with a space, and negative values with a minus sign. @item + The plus sign, used before the width modifier (see below), says to always supply a sign for numeric conversions, even if the data to be formatted is positive. The @samp{+} overrides the space modifier. @item # Use an ``alternate form'' for certain control letters. For @samp{o}, supply a leading zero. For @samp{x}, and @samp{X}, supply a leading @samp{0x} or @samp{0X} for a non-zero result. For @samp{e}, @samp{E}, and @samp{f}, the result will always contain a decimal point. For @samp{g}, and @samp{G}, trailing zeros are not removed from the result. @item 0 A leading @samp{0} (zero) acts as a flag, that indicates output should be padded with zeros instead of spaces. This applies even to non-numeric output formats. This flag only has an effect when the field width is wider than the value to be printed. @item @var{width} The field should be padded to this width. The field is normally padded with spaces. If the @samp{0} flag has been used, it is padded with zeros. @item .@var{prec} A number that specifies the precision to use when printing. For the @samp{e}, @samp{E}, and @samp{f} formats, this specifies the number of digits you want printed to the right of the decimal point. For the @samp{g}, and @samp{G} formats, it specifies the maximum number of significant digits. For the @samp{d}, @samp{o}, @samp{i}, @samp{u}, @samp{x}, and @samp{X} formats, it specifies the minimum number of digits to print. For the @samp{s} format, it specifies the maximum number of characters from the string that should be printed. @end table Either or both of the @var{width} and @var{prec} values may be specified as @samp{*}. In that case, the particular value is taken from the argument list. @xref{Printf, ,Using @code{printf} Statements for Fancier Printing}. @node Special File Summary, Built-in Functions Summary, Printf Summary, Actions Summary @appendixsubsec Special File Names When doing I/O redirection from either @code{print} or @code{printf} into a file, or via @code{getline} from a file, @code{gawk} recognizes certain special file names internally. These file names allow access to open file descriptors inherited from @code{gawk}'s parent process (usually the shell). The file names are: @table @file @item /dev/stdin The standard input. @item /dev/stdout The standard output. @item /dev/stderr The standard error output. @item /dev/fd/@var{n} The file denoted by the open file descriptor @var{n}. @end table In addition, reading the following files provides process related information about the running @code{gawk} program. All returned records are terminated with a newline. @table @file @item /dev/pid Returns the process ID of the current process. @item /dev/ppid Returns the parent process ID of the current process. @item /dev/pgrpid Returns the process group ID of the current process. @item /dev/user At least four space-separated fields, containing the return values of the @code{getuid}, @code{geteuid}, @code{getgid}, and @code{getegid} system calls. If there are any additional fields, they are the group IDs returned by @code{getgroups} system call. (Multiple groups may not be supported on all systems.) @end table @noindent These file names may also be used on the command line to name data files. These file names are only recognized internally if you do not actually have files with these names on your system. @xref{Special Files, ,Special File Names in @code{gawk}}, for a longer description that provides the motivation for this feature. @node Built-in Functions Summary, Time Functions Summary, Special File Summary, Actions Summary @appendixsubsec Built-in Functions @code{awk} provides a number of built-in functions for performing numeric operations, string related operations, and I/O related operations. The built-in arithmetic functions are: @table @code @item atan2(@var{y}, @var{x}) the arctangent of @var{y/x} in radians. @item cos(@var{expr}) the cosine of @var{expr}, which is in radians. @item exp(@var{expr}) the exponential function (@code{e ^ @var{expr}}). @item int(@var{expr}) truncates to integer. @item log(@var{expr}) the natural logarithm of @code{expr}. @item rand() a random number between zero and one. @item sin(@var{expr}) the sine of @var{expr}, which is in radians. @item sqrt(@var{expr}) the square root function. @item srand(@r{[}@var{expr}@r{]}) use @var{expr} as a new seed for the random number generator. If no @var{expr} is provided, the time of day is used. The return value is the previous seed for the random number generator. @end table @code{awk} has the following built-in string functions: @table @code @item gensub(@var{regex}, @var{subst}, @var{how} @r{[}, @var{target}@r{]}) If @var{how} is a string beginning with @samp{g} or @samp{G}, then replace each match of @var{regex} in @var{target} with @var{subst}. Otherwise, replace the @var{how}'th occurrence. If @var{target} is not supplied, use @code{$0}. The return value is the changed string; the original @var{target} is not modified. Within @var{subst}, @samp{\@var{n}}, where @var{n} is a digit from one to nine, can be used to indicate the text that matched the @var{n}'th parenthesized subexpression. This function is @code{gawk}-specific. @item gsub(@var{regex}, @var{subst} @r{[}, @var{target}@r{]}) for each substring matching the regular expression @var{regex} in the string @var{target}, substitute the string @var{subst}, and return the number of substitutions. If @var{target} is not supplied, use @code{$0}. @item index(@var{str}, @var{search}) returns the index of the string @var{search} in the string @var{str}, or zero if @var{search} is not present. @item length(@r{[}@var{str}@r{]}) returns the length of the string @var{str}. The length of @code{$0} is returned if no argument is supplied. @item match(@var{str}, @var{regex}) returns the position in @var{str} where the regular expression @var{regex} occurs, or zero if @var{regex} is not present, and sets the values of @code{RSTART} and @code{RLENGTH}. @item split(@var{str}, @var{arr} @r{[}, @var{regex}@r{]}) splits the string @var{str} into the array @var{arr} on the regular expression @var{regex}, and returns the number of elements. If @var{regex} is omitted, @code{FS} is used instead. @var{regex} can be the null string, causing each character to be placed into its own array element. The array @var{arr} is cleared first. @item sprintf(@var{fmt}, @var{expr-list}) prints @var{expr-list} according to @var{fmt}, and returns the resulting string. @item sub(@var{regex}, @var{subst} @r{[}, @var{target}@r{]}) just like @code{gsub}, but only the first matching substring is replaced. @item substr(@var{str}, @var{index} @r{[}, @var{len}@r{]}) returns the @var{len}-character substring of @var{str} starting at @var{index}. If @var{len} is omitted, the rest of @var{str} is used. @item tolower(@var{str}) returns a copy of the string @var{str}, with all the upper-case characters in @var{str} translated to their corresponding lower-case counterparts. Non-alphabetic characters are left unchanged. @item toupper(@var{str}) returns a copy of the string @var{str}, with all the lower-case characters in @var{str} translated to their corresponding upper-case counterparts. Non-alphabetic characters are left unchanged. @end table The I/O related functions are: @table @code @item close(@var{expr}) Close the open file or pipe denoted by @var{expr}. @item fflush(@r{[}@var{expr}@r{]}) Flush any buffered output for the output file or pipe denoted by @var{expr}. If @var{expr} is omitted, standard output is flushed. If @var{expr} is the null string (@code{""}), all output buffers are flushed. @item system(@var{cmd-line}) Execute the command @var{cmd-line}, and return the exit status. If your operating system does not support @code{system}, calling it will generate a fatal error. @samp{system("")} can be used to force @code{awk} to flush any pending output. This is more portable, but less obvious, than calling @code{fflush}. @end table @node Time Functions Summary, String Constants Summary, Built-in Functions Summary, Actions Summary @appendixsubsec Time Functions The following two functions are available for getting the current time of day, and for formatting time stamps. They are specific to @code{gawk}. @table @code @item systime() returns the current time of day as the number of seconds since a particular epoch (Midnight, January 1, 1970 UTC, on POSIX systems). @item strftime(@r{[}@var{format}@r{[}, @var{timestamp}@r{]]}) formats @var{timestamp} according to the specification in @var{format}. The current time of day is used if no @var{timestamp} is supplied. A default format equivalent to the output of the @code{date} utility is used if no @var{format} is supplied. @xref{Time Functions, ,Functions for Dealing with Time Stamps}, for the details on the conversion specifiers that @code{strftime} accepts. @end table @iftex @xref{Built-in, ,Built-in Functions}, for a description of all of @code{awk}'s built-in functions. @end iftex @node String Constants Summary, , Time Functions Summary, Actions Summary @appendixsubsec String Constants String constants in @code{awk} are sequences of characters enclosed in double quotes (@code{"}). Within strings, certain @dfn{escape sequences} are recognized, as in C. These are: @table @code @item \\ A literal backslash. @item \a The ``alert'' character; usually the ASCII BEL character. @item \b Backspace. @item \f Formfeed. @item \n Newline. @item \r Carriage return. @item \t Horizontal tab. @item \v Vertical tab. @item \x@var{hex digits} The character represented by the string of hexadecimal digits following the @samp{\x}. As in ANSI C, all following hexadecimal digits are considered part of the escape sequence. E.g., @code{"\x1B"} is a string containing the ASCII ESC (escape) character. (The @samp{\x} escape sequence is not in POSIX @code{awk}.) @item \@var{ddd} The character represented by the one, two, or three digit sequence of octal digits. Thus, @code{"\033"} is also a string containing the ASCII ESC (escape) character. @item \@var{c} The literal character @var{c}, if @var{c} is not one of the above. @end table The escape sequences may also be used inside constant regular expressions (e.g., the regexp @code{@w{/[@ \t\f\n\r\v]/}} matches whitespace characters). @xref{Escape Sequences}. @node Functions Summary, Historical Features, Actions Summary, Gawk Summary @appendixsec User-defined Functions Functions in @code{awk} are defined as follows: @example function @var{name}(@var{parameter list}) @{ @var{statements} @} @end example Actual parameters supplied in the function call are used to instantiate the formal parameters declared in the function. Arrays are passed by reference, other variables are passed by value. If there are fewer arguments passed than there are names in @var{parameter-list}, the extra names are given the null string as their value. Extra names have the effect of local variables. The open-parenthesis in a function call of a user-defined function must immediately follow the function name, without any intervening white space. This is to avoid a syntactic ambiguity with the concatenation operator. The word @code{func} may be used in place of @code{function} (but not in POSIX @code{awk}). Use the @code{return} statement to return a value from a function. @xref{User-defined, ,User-defined Functions}. @node Historical Features, , Functions Summary, Gawk Summary @appendixsec Historical Features @cindex historical features There are two features of historical @code{awk} implementations that @code{gawk} supports. First, it is possible to call the @code{length} built-in function not only with no arguments, but even without parentheses! @example a = length @end example @noindent is the same as either of @example a = length() a = length($0) @end example @noindent For example: @example $ echo abcdef | awk '@{ print length @}' @print{} 6 @end example @noindent This feature is marked as ``deprecated'' in the POSIX standard, and @code{gawk} will issue a warning about its use if @samp{--lint} is specified on the command line. (The ability to use @code{length} this way was actually an accident of the original Unix @code{awk} implementation. If any built-in function used @code{$0} as its default argument, it was possible to call that function without the parentheses. In particular, it was common practice to use the @code{length} function in this fashion, and this usage was documented in the @code{awk} manual page.) The other historical feature is the use of either the @code{break} statement, or the @code{continue} statement outside the body of a @code{while}, @code{for}, or @code{do} loop. Traditional @code{awk} implementations have treated such usage as equivalent to the @code{next} statement. More recent versions of Unix @code{awk} do not allow it. @code{gawk} supports this usage if @samp{--traditional} has been specified. @xref{Options, ,Command Line Options}, for more information about the @samp{--posix} and @samp{--lint} options. @node Installation, Notes, Gawk Summary, Top @appendix Installing @code{gawk} This appendix provides instructions for installing @code{gawk} on the various platforms that are supported by the developers. The primary developers support Unix (and one day, GNU), while the other ports were contributed. The file @file{ACKNOWLEDGMENT} in the @code{gawk} distribution lists the electronic mail addresses of the people who did the respective ports, and they are also provided in @ref{Bugs, , Reporting Problems and Bugs}. @menu * Gawk Distribution:: What is in the @code{gawk} distribution. * Unix Installation:: Installing @code{gawk} under various versions of Unix. * VMS Installation:: Installing @code{gawk} on VMS. * PC Installation:: Installing and Compiling @code{gawk} on MS-DOS and OS/2 * Atari Installation:: Installing @code{gawk} on the Atari ST. * Amiga Installation:: Installing @code{gawk} on an Amiga. * Bugs:: Reporting Problems and Bugs. * Other Versions:: Other freely available @code{awk} implementations. @end menu @node Gawk Distribution, Unix Installation, Installation, Installation @appendixsec The @code{gawk} Distribution This section first describes how to get the @code{gawk} distribution, how to extract it, and then what is in the various files and subdirectories. @menu * Getting:: How to get the distribution. * Extracting:: How to extract the distribution. * Distribution contents:: What is in the distribution. @end menu @node Getting, Extracting, Gawk Distribution, Gawk Distribution @appendixsubsec Getting the @code{gawk} Distribution @cindex getting @code{gawk} @cindex anonymous @code{ftp} @cindex @code{ftp}, anonymous @cindex Free Software Foundation There are three ways you can get GNU software. @enumerate @item You can copy it from someone else who already has it. @cindex Free Software Foundation @item You can order @code{gawk} directly from the Free Software Foundation. Software distributions are available for Unix, MS-DOS, and VMS, on tape and CD-ROM. The address is: @quotation Free Software Foundation @* 59 Temple Place---Suite 330 @* Boston, MA 02111-1307 USA @* Phone: +1-617-542-5942 @* Fax (including Japan): +1-617-542-2652 @* E-mail: @code{gnu@@gnu.org} @* @end quotation @noindent Ordering from the FSF directly contributes to the support of the foundation and to the production of more free software. @item You can get @code{gawk} by using anonymous @code{ftp} to the Internet host @code{gnudist.gnu.org}, in the directory @file{/gnu/gawk}. Here is a list of alternate @code{ftp} sites from which you can obtain GNU software. When a site is listed as ``@var{site}@code{:}@var{directory}'' the @var{directory} indicates the directory where GNU software is kept. You should use a site that is geographically close to you. @table @asis @item Asia: @table @code @item cair-archive.kaist.ac.kr:/pub/gnu @itemx ftp.cs.titech.ac.jp @itemx ftp.nectec.or.th:/pub/mirrors/gnu @itemx utsun.s.u-tokyo.ac.jp:/ftpsync/prep @end table @item Australia: @table @code @item archie.au:/gnu (@code{archie.oz} or @code{archie.oz.au} for ACSnet) @end table @item Africa: @table @code @item ftp.sun.ac.za:/pub/gnu @end table @item Middle East: @table @code @item ftp.technion.ac.il:/pub/unsupported/gnu @end table @item Europe: @table @code @item archive.eu.net @itemx ftp.denet.dk @itemx ftp.eunet.ch @itemx ftp.funet.fi:/pub/gnu @itemx ftp.ieunet.ie:pub/gnu @itemx ftp.informatik.rwth-aachen.de:/pub/gnu @itemx ftp.informatik.tu-muenchen.de @itemx ftp.luth.se:/pub/unix/gnu @itemx ftp.mcc.ac.uk @itemx ftp.stacken.kth.se @itemx ftp.sunet.se:/pub/gnu @itemx ftp.univ-lyon1.fr:pub/gnu @itemx ftp.win.tue.nl:/pub/gnu @itemx irisa.irisa.fr:/pub/gnu @itemx isy.liu.se @itemx nic.switch.ch:/mirror/gnu @itemx src.doc.ic.ac.uk:/gnu @itemx unix.hensa.ac.uk:/pub/uunet/systems/gnu @end table @item South America: @table @code @item ftp.inf.utfsm.cl:/pub/gnu @itemx ftp.unicamp.br:/pub/gnu @end table @item Western Canada: @table @code @item ftp.cs.ubc.ca:/mirror2/gnu @end table @item USA: @table @code @item col.hp.com:/mirrors/gnu @itemx f.ms.uky.edu:/pub3/gnu @itemx ftp.cc.gatech.edu:/pub/gnu @itemx ftp.cs.columbia.edu:/archives/gnu/prep @itemx ftp.digex.net:/pub/gnu @itemx ftp.hawaii.edu:/mirrors/gnu @itemx ftp.kpc.com:/pub/mirror/gnu @end table @c NEEDED @page @item USA (continued): @table @code @itemx ftp.uu.net:/systems/gnu @itemx gatekeeper.dec.com:/pub/GNU @itemx jaguar.utah.edu:/gnustuff @itemx labrea.stanford.edu @itemx mrcnext.cso.uiuc.edu:/pub/gnu @itemx vixen.cso.uiuc.edu:/gnu @itemx wuarchive.wustl.edu:/systems/gnu @end table @end table @end enumerate @node Extracting, Distribution contents, Getting, Gawk Distribution @appendixsubsec Extracting the Distribution @code{gawk} is distributed as a @code{tar} file compressed with the GNU Zip program, @code{gzip}. Once you have the distribution (for example, @file{gawk-@value{VERSION}.@value{PATCHLEVEL}.tar.gz}), first use @code{gzip} to expand the file, and then use @code{tar} to extract it. You can use the following pipeline to produce the @code{gawk} distribution: @example # Under System V, add 'o' to the tar flags gzip -d -c gawk-@value{VERSION}.@value{PATCHLEVEL}.tar.gz | tar -xvpf - @end example @noindent This will create a directory named @file{gawk-@value{VERSION}.@value{PATCHLEVEL}} in the current directory. The distribution file name is of the form @file{gawk-@var{V}.@var{R}.@var{n}.tar.gz}. The @var{V} represents the major version of @code{gawk}, the @var{R} represents the current release of version @var{V}, and the @var{n} represents a @dfn{patch level}, meaning that minor bugs have been fixed in the release. The current patch level is @value{PATCHLEVEL}, but when retrieving distributions, you should get the version with the highest version, release, and patch level. (Note that release levels greater than or equal to 90 denote ``beta,'' or non-production software; you may not wish to retrieve such a version unless you don't mind experimenting.) If you are not on a Unix system, you will need to make other arrangements for getting and extracting the @code{gawk} distribution. You should consult a local expert. @node Distribution contents, , Extracting, Gawk Distribution @appendixsubsec Contents of the @code{gawk} Distribution The @code{gawk} distribution has a number of C source files, documentation files, subdirectories and files related to the configuration process (@pxref{Unix Installation, ,Compiling and Installing @code{gawk} on Unix}), and several subdirectories related to different, non-Unix, operating systems. @table @asis @item various @samp{.c}, @samp{.y}, and @samp{.h} files These files are the actual @code{gawk} source code. @end table @table @file @item README @itemx README_d/README.* Descriptive files: @file{README} for @code{gawk} under Unix, and the rest for the various hardware and software combinations. @item INSTALL A file providing an overview of the configuration and installation process. @item PORTS A list of systems to which @code{gawk} has been ported, and which have successfully run the test suite. @item ACKNOWLEDGMENT A list of the people who contributed major parts of the code or documentation. @item ChangeLog A detailed list of source code changes as bugs are fixed or improvements made. @item NEWS A list of changes to @code{gawk} since the last release or patch. @item COPYING The GNU General Public License. @item FUTURES A brief list of features and/or changes being contemplated for future releases, with some indication of the time frame for the feature, based on its difficulty. @item LIMITATIONS A list of those factors that limit @code{gawk}'s performance. Most of these depend on the hardware or operating system software, and are not limits in @code{gawk} itself. @item POSIX.STD A description of one area where the POSIX standard for @code{awk} is incorrect, and how @code{gawk} handles the problem. @item PROBLEMS A file describing known problems with the current release. @cindex artificial intelligence, using @code{gawk} @cindex AI programming, using @code{gawk} @item doc/awkforai.txt A short article describing why @code{gawk} is a good language for AI (Artificial Intelligence) programming. @item doc/README.card @itemx doc/ad.block @itemx doc/awkcard.in @itemx doc/cardfonts @itemx doc/colors @itemx doc/macros @itemx doc/no.colors @itemx doc/setter.outline The @code{troff} source for a five-color @code{awk} reference card. A modern version of @code{troff}, such as GNU Troff (@code{groff}) is needed to produce the color version. See the file @file{README.card} for instructions if you have an older @code{troff}. @item doc/gawk.1 The @code{troff} source for a manual page describing @code{gawk}. This is distributed for the convenience of Unix users. @item doc/gawk.texi The Texinfo source file for this @value{DOCUMENT}. It should be processed with @TeX{} to produce a printed document, and with @code{makeinfo} to produce an Info file. @item doc/gawk.info The generated Info file for this @value{DOCUMENT}. @item doc/igawk.1 The @code{troff} source for a manual page describing the @code{igawk} program presented in @ref{Igawk Program, ,An Easy Way to Use Library Functions}. @item doc/Makefile.in The input file used during the configuration process to generate the actual @file{Makefile} for creating the documentation. @item Makefile.in @itemx acconfig.h @itemx aclocal.m4 @itemx configh.in @itemx configure.in @itemx configure @itemx custom.h @itemx missing/* These files and subdirectory are used when configuring @code{gawk} for various Unix systems. They are explained in detail in @ref{Unix Installation, ,Compiling and Installing @code{gawk} on Unix}. @item awklib/extract.awk @itemx awklib/Makefile.in The @file{awklib} directory contains a copy of @file{extract.awk} (@pxref{Extract Program, ,Extracting Programs from Texinfo Source Files}), which can be used to extract the sample programs from the Texinfo source file for this @value{DOCUMENT}, and a @file{Makefile.in} file, which @code{configure} uses to generate a @file{Makefile}. As part of the process of building @code{gawk}, the library functions from @ref{Library Functions, , A Library of @code{awk} Functions}, and the @code{igawk} program from @ref{Igawk Program, , An Easy Way to Use Library Functions}, are extracted into ready to use files. They are installed as part of the installation process. @item atari/* Files needed for building @code{gawk} on an Atari ST. @xref{Atari Installation, ,Installing @code{gawk} on the Atari ST}, for details. @item pc/* Files needed for building @code{gawk} under MS-DOS and OS/2. @xref{PC Installation, ,MS-DOS and OS/2 Installation and Compilation}, for details. @item vms/* Files needed for building @code{gawk} under VMS. @xref{VMS Installation, ,How to Compile and Install @code{gawk} on VMS}, for details. @item test/* A test suite for @code{gawk}. You can use @samp{make check} from the top level @code{gawk} directory to run your version of @code{gawk} against the test suite. If @code{gawk} successfully passes @samp{make check} then you can be confident of a successful port. @end table @node Unix Installation, VMS Installation, Gawk Distribution, Installation @appendixsec Compiling and Installing @code{gawk} on Unix Usually, you can compile and install @code{gawk} by typing only two commands. However, if you do use an unusual system, you may need to configure @code{gawk} for your system yourself. @menu * Quick Installation:: Compiling @code{gawk} under Unix. * Configuration Philosophy:: How it's all supposed to work. @end menu @node Quick Installation, Configuration Philosophy, Unix Installation, Unix Installation @appendixsubsec Compiling @code{gawk} for Unix @cindex installation, unix After you have extracted the @code{gawk} distribution, @code{cd} to @file{gawk-@value{VERSION}.@value{PATCHLEVEL}}. Like most GNU software, @code{gawk} is configured automatically for your Unix system by running the @code{configure} program. This program is a Bourne shell script that was generated automatically using GNU @code{autoconf}. @iftex (The @code{autoconf} software is described fully in @cite{Autoconf---Generating Automatic Configuration Scripts}, which is available from the Free Software Foundation.) @end iftex @ifinfo (The @code{autoconf} software is described fully starting with @ref{Top, , Introduction, autoconf, Autoconf---Generating Automatic Configuration Scripts}.) @end ifinfo To configure @code{gawk}, simply run @code{configure}: @example sh ./configure @end example This produces a @file{Makefile} and @file{config.h} tailored to your system. The @file{config.h} file describes various facts about your system. You may wish to edit the @file{Makefile} to change the @code{CFLAGS} variable, which controls the command line options that are passed to the C compiler (such as optimization levels, or compiling for debugging). Alternatively, you can add your own values for most @code{make} variables, such as @code{CC} and @code{CFLAGS}, on the command line when running @code{configure}: @example CC=cc CFLAGS=-g sh ./configure @end example @noindent See the file @file{INSTALL} in the @code{gawk} distribution for all the details. After you have run @code{configure}, and possibly edited the @file{Makefile}, type: @example make @end example @noindent and shortly thereafter, you should have an executable version of @code{gawk}. That's all there is to it! (If these steps do not work, please send in a bug report; @pxref{Bugs, ,Reporting Problems and Bugs}.) @node Configuration Philosophy, , Quick Installation, Unix Installation @appendixsubsec The Configuration Process @cindex configuring @code{gawk} (This section is of interest only if you know something about using the C language and the Unix operating system.) The source code for @code{gawk} generally attempts to adhere to formal standards wherever possible. This means that @code{gawk} uses library routines that are specified by the ANSI C standard and by the POSIX operating system interface standard. When using an ANSI C compiler, function prototypes are used to help improve the compile-time checking. Many Unix systems do not support all of either the ANSI or the POSIX standards. The @file{missing} subdirectory in the @code{gawk} distribution contains replacement versions of those subroutines that are most likely to be missing. The @file{config.h} file that is created by the @code{configure} program contains definitions that describe features of the particular operating system where you are attempting to compile @code{gawk}. The three things described by this file are what header files are available, so that they can be correctly included, what (supposedly) standard functions are actually available in your C libraries, and other miscellaneous facts about your variant of Unix. For example, there may not be an @code{st_blksize} element in the @code{stat} structure. In this case @samp{HAVE_ST_BLKSIZE} would be undefined. @cindex @code{custom.h} configuration file It is possible for your C compiler to lie to @code{configure}. It may do so by not exiting with an error when a library function is not available. To get around this, you can edit the file @file{custom.h}. Use an @samp{#ifdef} that is appropriate for your system, and either @code{#define} any constants that @code{configure} should have defined but didn't, or @code{#undef} any constants that @code{configure} defined and should not have. @file{custom.h} is automatically included by @file{config.h}. It is also possible that the @code{configure} program generated by @code{autoconf} will not work on your system in some other fashion. If you do have a problem, the file @file{configure.in} is the input for @code{autoconf}. You may be able to change this file, and generate a new version of @code{configure} that will work on your system. @xref{Bugs, ,Reporting Problems and Bugs}, for information on how to report problems in configuring @code{gawk}. The same mechanism may be used to send in updates to @file{configure.in} and/or @file{custom.h}. @node VMS Installation, PC Installation, Unix Installation, Installation @appendixsec How to Compile and Install @code{gawk} on VMS @c based on material from Pat Rankin @cindex installation, vms This section describes how to compile and install @code{gawk} under VMS. @menu * VMS Compilation:: How to compile @code{gawk} under VMS. * VMS Installation Details:: How to install @code{gawk} under VMS. * VMS Running:: How to run @code{gawk} under VMS. * VMS POSIX:: Alternate instructions for VMS POSIX. @end menu @node VMS Compilation, VMS Installation Details, VMS Installation, VMS Installation @appendixsubsec Compiling @code{gawk} on VMS To compile @code{gawk} under VMS, there is a @code{DCL} command procedure that will issue all the necessary @code{CC} and @code{LINK} commands, and there is also a @file{Makefile} for use with the @code{MMS} utility. From the source directory, use either @example $ @@[.VMS]VMSBUILD.COM @end example @noindent or @example $ MMS/DESCRIPTION=[.VMS]DESCRIP.MMS GAWK @end example Depending upon which C compiler you are using, follow one of the sets of instructions in this table: @table @asis @item VAX C V3.x Use either @file{vmsbuild.com} or @file{descrip.mms} as is. These use @code{CC/OPTIMIZE=NOLINE}, which is essential for Version 3.0. @item VAX C V2.x You must have Version 2.3 or 2.4; older ones won't work. Edit either @file{vmsbuild.com} or @file{descrip.mms} according to the comments in them. For @file{vmsbuild.com}, this just entails removing two @samp{!} delimiters. Also edit @file{config.h} (which is a copy of file @file{[.config]vms-conf.h}) and comment out or delete the two lines @samp{#define __STDC__ 0} and @samp{#define VAXC_BUILTINS} near the end. @item GNU C Edit @file{vmsbuild.com} or @file{descrip.mms}; the changes are different from those for VAX C V2.x, but equally straightforward. No changes to @file{config.h} should be needed. @item DEC C Edit @file{vmsbuild.com} or @file{descrip.mms} according to their comments. No changes to @file{config.h} should be needed. @end table @code{gawk} has been tested under VAX/VMS 5.5-1 using VAX C V3.2, GNU C 1.40 and 2.3. It should work without modifications for VMS V4.6 and up. @node VMS Installation Details, VMS Running, VMS Compilation, VMS Installation @appendixsubsec Installing @code{gawk} on VMS To install @code{gawk}, all you need is a ``foreign'' command, which is a @code{DCL} symbol whose value begins with a dollar sign. For example: @example $ GAWK :== $disk1:[gnubin]GAWK @end example @noindent (Substitute the actual location of @code{gawk.exe} for @samp{$disk1:[gnubin]}.) The symbol should be placed in the @file{login.com} of any user who wishes to run @code{gawk}, so that it will be defined every time the user logs on. Alternatively, the symbol may be placed in the system-wide @file{sylogin.com} procedure, which will allow all users to run @code{gawk}. Optionally, the help entry can be loaded into a VMS help library: @example $ LIBRARY/HELP SYS$HELP:HELPLIB [.VMS]GAWK.HLP @end example @noindent (You may want to substitute a site-specific help library rather than the standard VMS library @samp{HELPLIB}.) After loading the help text, @example $ HELP GAWK @end example @noindent will provide information about both the @code{gawk} implementation and the @code{awk} programming language. The logical name @samp{AWK_LIBRARY} can designate a default location for @code{awk} program files. For the @samp{-f} option, if the specified filename has no device or directory path information in it, @code{gawk} will look in the current directory first, then in the directory specified by the translation of @samp{AWK_LIBRARY} if the file was not found. If after searching in both directories, the file still is not found, then @code{gawk} appends the suffix @samp{.awk} to the filename and the file search will be re-tried. If @samp{AWK_LIBRARY} is not defined, that portion of the file search will fail benignly. @node VMS Running, VMS POSIX, VMS Installation Details, VMS Installation @appendixsubsec Running @code{gawk} on VMS Command line parsing and quoting conventions are significantly different on VMS, so examples in this @value{DOCUMENT} or from other sources often need minor changes. They @emph{are} minor though, and all @code{awk} programs should run correctly. Here are a couple of trivial tests: @example $ gawk -- "BEGIN @{print ""Hello, World!""@}" $ gawk -"W" version ! could also be -"W version" or "-W version" @end example @noindent Note that upper-case and mixed-case text must be quoted. The VMS port of @code{gawk} includes a @code{DCL}-style interface in addition to the original shell-style interface (see the help entry for details). One side-effect of dual command line parsing is that if there is only a single parameter (as in the quoted string program above), the command becomes ambiguous. To work around this, the normally optional @samp{--} flag is required to force Unix style rather than @code{DCL} parsing. If any other dash-type options (or multiple parameters such as data files to be processed) are present, there is no ambiguity and @samp{--} can be omitted. The default search path when looking for @code{awk} program files specified by the @samp{-f} option is @code{"SYS$DISK:[],AWK_LIBRARY:"}. The logical name @samp{AWKPATH} can be used to override this default. The format of @samp{AWKPATH} is a comma-separated list of directory specifications. When defining it, the value should be quoted so that it retains a single translation, and not a multi-translation @code{RMS} searchlist. @node VMS POSIX, , VMS Running, VMS Installation @appendixsubsec Building and Using @code{gawk} on VMS POSIX Ignore the instructions above, although @file{vms/gawk.hlp} should still be made available in a help library. The source tree should be unpacked into a container file subsystem rather than into the ordinary VMS file system. Make sure that the two scripts, @file{configure} and @file{vms/posix-cc.sh}, are executable; use @samp{chmod +x} on them if necessary. Then execute the following two commands: @example @group psx> CC=vms/posix-cc.sh configure psx> make CC=c89 gawk @end group @end example @noindent The first command will construct files @file{config.h} and @file{Makefile} out of templates, using a script to make the C compiler fit @code{configure}'s expectations. The second command will compile and link @code{gawk} using the C compiler directly; ignore any warnings from @code{make} about being unable to redefine @code{CC}. @code{configure} will take a very long time to execute, but at least it provides incremental feedback as it runs. This has been tested with VAX/VMS V6.2, VMS POSIX V2.0, and DEC C V5.2. Once built, @code{gawk} will work like any other shell utility. Unlike the normal VMS port of @code{gawk}, no special command line manipulation is needed in the VMS POSIX environment. @c Rewritten by Scott Deifik @c and Darrel Hankerson @node PC Installation, Atari Installation, VMS Installation, Installation @appendixsec MS-DOS and OS/2 Installation and Compilation @cindex installation, MS-DOS and OS/2 If you have received a binary distribution prepared by the DOS maintainers, then @code{gawk} and the necessary support files will appear under the @file{gnu} directory, with executables in @file{gnu/bin}, libraries in @file{gnu/lib/awk}, and manual pages under @file{gnu/man}. This is designed for easy installation to a @file{/gnu} directory on your drive, but the files can be installed anywhere provided @code{AWKPATH} is set properly. Regardless of the installation directory, the first line of @file{igawk.cmd} and @file{igawk.bat} (in @file{gnu/bin}) may need to be edited. The binary distribution will contain a separate file describing the contents. In particular, it may include more than one version of the @code{gawk} executable. OS/2 binary distributions may have a different arrangement, but installation is similar. The OS/2 and MS-DOS versions of @code{gawk} search for program files as described in @ref{AWKPATH Variable, ,The @code{AWKPATH} Environment Variable}. However, semicolons (rather than colons) separate elements in the @code{AWKPATH} variable. If @code{AWKPATH} is not set or is empty, then the default search path is @code{@w{".;c:/lib/awk;c:/gnu/lib/awk"}}. An @code{sh}-like shell (as opposed to @code{command.com} under MS-DOS or @code{cmd.exe} under OS/2) may be useful for @code{awk} programming. Ian Stewartson has written an excellent shell for MS-DOS and OS/2, and a @code{ksh} clone and GNU Bash are available for OS/2. The file @file{README_d/README.pc} in the @code{gawk} distribution contains information on these shells. Users of Stewartson's shell on DOS should examine its documentation on handling of command-lines. In particular, the setting for @code{gawk} in the shell configuration may need to be changed, and the @code{ignoretype} option may also be of interest. @code{gawk} can be compiled for MS-DOS and OS/2 using the GNU development tools from DJ Delorie (DJGPP, MS-DOS-only) or Eberhard Mattes (EMX, MS-DOS and OS/2). Microsoft C can be used to build 16-bit versions for MS-DOS and OS/2. The file @file{README_d/README.pc} in the @code{gawk} distribution contains additional notes, and @file{pc/Makefile} contains important notes on compilation options. To build @code{gawk}, copy the files in the @file{pc} directory (@emph{except} for @file{ChangeLog}) to the directory with the rest of the @code{gawk} sources. The @file{Makefile} contains a configuration section with comments, and may need to be edited in order to work with your @code{make} utility. The @file{Makefile} contains a number of targets for building various MS-DOS and OS/2 versions. A list of targets will be printed if the @code{make} command is given without a target. As an example, to build @code{gawk} using the DJGPP tools, enter @samp{make djgpp}. Using @code{make} to run the standard tests and to install @code{gawk} requires additional Unix-like tools, including @code{sh}, @code{sed}, and @code{cp}. In order to run the tests, the @file{test/*.ok} files may need to be converted so that they have the usual DOS-style end-of-line markers. Most of the tests will work properly with Stewartson's shell along with the companion utilities or appropriate GNU utilities. However, some editing of @file{test/Makefile} is required. It is recommended that the file @file{pc/Makefile.tst} be copied to @file{test/Makefile} as a replacement. Details can be found in @file{README_d/README.pc}. @node Atari Installation, Amiga Installation, PC Installation, Installation @appendixsec Installing @code{gawk} on the Atari ST @c based on material from Michal Jaegermann @cindex atari @cindex installation, atari There are no substantial differences when installing @code{gawk} on various Atari models. Compiled @code{gawk} executables do not require a large amount of memory with most @code{awk} programs and should run on all Motorola processor based models (called further ST, even if that is not exactly right). In order to use @code{gawk}, you need to have a shell, either text or graphics, that does not map all the characters of a command line to upper-case. Maintaining case distinction in option flags is very important (@pxref{Options, ,Command Line Options}). These days this is the default, and it may only be a problem for some very old machines. If your system does not preserve the case of option flags, you will need to upgrade your tools. Support for I/O redirection is necessary to make it easy to import @code{awk} programs from other environments. Pipes are nice to have, but not vital. @menu * Atari Compiling:: Compiling @code{gawk} on Atari * Atari Using:: Running @code{gawk} on Atari @end menu @node Atari Compiling, Atari Using, Atari Installation, Atari Installation @appendixsubsec Compiling @code{gawk} on the Atari ST A proper compilation of @code{gawk} sources when @code{sizeof(int)} differs from @code{sizeof(void *)} requires an ANSI C compiler. An initial port was done with @code{gcc}. You may actually prefer executables where @code{int}s are four bytes wide, but the other variant works as well. You may need quite a bit of memory when trying to recompile the @code{gawk} sources, as some source files (@file{regex.c} in particular) are quite big. If you run out of memory compiling such a file, try reducing the optimization level for this particular file; this may help. @cindex Linux With a reasonable shell (Bash will do), and in particular if you run Linux, MiNT or a similar operating system, you have a pretty good chance that the @code{configure} utility will succeed. Otherwise sample versions of @file{config.h} and @file{Makefile.st} are given in the @file{atari} subdirectory and can be edited and copied to the corresponding files in the main source directory. Even if @code{configure} produced something, it might be advisable to compare its results with the sample versions and possibly make adjustments. Some @code{gawk} source code fragments depend on a preprocessor define @samp{atarist}. This basically assumes the TOS environment with @code{gcc}. Modify these sections as appropriate if they are not right for your environment. Also see the remarks about @code{AWKPATH} and @code{envsep} in @ref{Atari Using, ,Running @code{gawk} on the Atari ST}. As shipped, the sample @file{config.h} claims that the @code{system} function is missing from the libraries, which is not true, and an alternative implementation of this function is provided in @file{atari/system.c}. Depending upon your particular combination of shell and operating system, you may wish to change the file to indicate that @code{system} is available. @node Atari Using, , Atari Compiling, Atari Installation @appendixsubsec Running @code{gawk} on the Atari ST An executable version of @code{gawk} should be placed, as usual, anywhere in your @code{PATH} where your shell can find it. While executing, @code{gawk} creates a number of temporary files. When using @code{gcc} libraries for TOS, @code{gawk} looks for either of the environment variables @code{TEMP} or @code{TMPDIR}, in that order. If either one is found, its value is assumed to be a directory for temporary files. This directory must exist, and if you can spare the memory, it is a good idea to put it on a RAM drive. If neither @code{TEMP} nor @code{TMPDIR} are found, then @code{gawk} uses the current directory for its temporary files. The ST version of @code{gawk} searches for its program files as described in @ref{AWKPATH Variable, ,The @code{AWKPATH} Environment Variable}. The default value for the @code{AWKPATH} variable is taken from @code{DEFPATH} defined in @file{Makefile}. The sample @code{gcc}/TOS @file{Makefile} for the ST in the distribution sets @code{DEFPATH} to @code{@w{".,c:\lib\awk,c:\gnu\lib\awk"}}. The search path can be modified by explicitly setting @code{AWKPATH} to whatever you wish. Note that colons cannot be used on the ST to separate elements in the @code{AWKPATH} variable, since they have another, reserved, meaning. Instead, you must use a comma to separate elements in the path. When recompiling, the separating character can be modified by initializing the @code{envsep} variable in @file{atari/gawkmisc.atr} to another value. Although @code{awk} allows great flexibility in doing I/O redirections from within a program, this facility should be used with care on the ST running under TOS. In some circumstances the OS routines for file handle pool processing lose track of certain events, causing the computer to crash, and requiring a reboot. Often a warm reboot is sufficient. Fortunately, this happens infrequently, and in rather esoteric situations. In particular, avoid having one part of an @code{awk} program using @code{print} statements explicitly redirected to @code{"/dev/stdout"}, while other @code{print} statements use the default standard output, and a calling shell has redirected standard output to a file. When @code{gawk} is compiled with the ST version of @code{gcc} and its usual libraries, it will accept both @samp{/} and @samp{\} as path separators. While this is convenient, it should be remembered that this removes one, technically valid, character (@samp{/}) from your file names, and that it may create problems for external programs, called via the @code{system} function, which may not support this convention. Whenever it is possible that a file created by @code{gawk} will be used by some other program, use only backslashes. Also remember that in @code{awk}, backslashes in strings have to be doubled in order to get literal backslashes (@pxref{Escape Sequences}). @node Amiga Installation, Bugs, Atari Installation, Installation @appendixsec Installing @code{gawk} on an Amiga @cindex amiga @cindex installation, amiga You can install @code{gawk} on an Amiga system using a Unix emulation environment available via anonymous @code{ftp} from @code{ftp.ninemoons.com} in the directory @file{pub/ade/current}. This includes a shell based on @code{pdksh}. The primary component of this environment is a Unix emulation library, @file{ixemul.lib}. @c could really use more background here, who wrote this, etc. A more complete distribution for the Amiga is available on the Geek Gadgets CD-ROM from: @quotation CRONUS @* 1840 E. Warner Road #105-265 @* Tempe, AZ 85284 USA @* US Toll Free: (800) 804-0833 @* Phone: +1-602-491-0442 @* FAX: +1-602-491-0048 @* Email: @code{info@@ninemoons.com} @* WWW: @code{http://www.ninemoons.com} @* Anonymous @code{ftp} site: @code{ftp.ninemoons.com} @* @end quotation Once you have the distribution, you can configure @code{gawk} simply by running @code{configure}: @example configure -v m68k-amigaos @end example Then run @code{make}, and you should be all set! (If these steps do not work, please send in a bug report; @pxref{Bugs, ,Reporting Problems and Bugs}.) @node Bugs, Other Versions, Amiga Installation, Installation @appendixsec Reporting Problems and Bugs @display @i{There is nothing more dangerous than a bored archeologist.} The Hitchhiker's Guide to the Galaxy @c the radio show, not the book. :-) @end display @sp 1 If you have problems with @code{gawk} or think that you have found a bug, please report it to the developers; we cannot promise to do anything but we might well want to fix it. Before reporting a bug, make sure you have actually found a real bug. Carefully reread the documentation and see if it really says you can do what you're trying to do. If it's not clear whether you should be able to do something or not, report that too; it's a bug in the documentation! Before reporting a bug or trying to fix it yourself, try to isolate it to the smallest possible @code{awk} program and input data file that reproduces the problem. Then send us the program and data file, some idea of what kind of Unix system you're using, and the exact results @code{gawk} gave you. Also say what you expected to occur; this will help us decide whether the problem was really in the documentation. Once you have a precise problem, there are two e-mail addresses you can send mail to. @table @asis @item Internet: @samp{bug-gnu-utils@@gnu.org} @item UUCP: @samp{uunet!gnu.org!bug-gnu-utils} @end table Please include the version number of @code{gawk} you are using. You can get this information with the command @samp{gawk --version}. You should send a carbon copy of your mail to Arnold Robbins, who can be reached at @samp{arnold@@gnu.org}. @cindex @code{comp.lang.awk} @strong{Important!} Do @emph{not} try to report bugs in @code{gawk} by posting to the Usenet/Internet newsgroup @code{comp.lang.awk}. While the @code{gawk} developers do occasionally read this newsgroup, there is no guarantee that we will see your posting. The steps described above are the official, recognized ways for reporting bugs. Non-bug suggestions are always welcome as well. If you have questions about things that are unclear in the documentation or are just obscure features, ask Arnold Robbins; he will try to help you out, although he may not have the time to fix the problem. You can send him electronic mail at the Internet address above. If you find bugs in one of the non-Unix ports of @code{gawk}, please send an electronic mail message to the person who maintains that port. They are listed below, and also in the @file{README} file in the @code{gawk} distribution. Information in the @file{README} file should be considered authoritative if it conflicts with this @value{DOCUMENT}. @c NEEDED for looks @page The people maintaining the non-Unix ports of @code{gawk} are: @cindex Deifik, Scott @cindex Fish, Fred @cindex Hankerson, Darrel @cindex Jaegermann, Michal @cindex Rankin, Pat @cindex Rommel, Kai Uwe @table @asis @item MS-DOS Scott Deifik, @samp{scottd@@amgen.com}, and Darrel Hankerson, @samp{hankedr@@mail.auburn.edu}. @item OS/2 Kai Uwe Rommel, @samp{rommel@@ars.de}. @item VMS Pat Rankin, @samp{rankin@@eql.caltech.edu}. @item Atari ST Michal Jaegermann, @samp{michal@@gortel.phys.ualberta.ca}. @item Amiga Fred Fish, @samp{fnf@@ninemoons.com}. @end table If your bug is also reproducible under Unix, please send copies of your report to the general GNU bug list, as well as to Arnold Robbins, at the addresses listed above. @node Other Versions, , Bugs, Installation @appendixsec Other Freely Available @code{awk} Implementations @cindex Brennan, Michael @ignore From: emory!amc.com!brennan (Michael Brennan) Subject: C++ comments in awk programs To: arnold@gnu.ai.mit.edu (Arnold Robbins) Date: Wed, 4 Sep 1996 08:11:48 -0700 (PDT) @end ignore @display @i{It's kind of fun to put comments like this in your awk code.} @code{// Do C++ comments work? answer: yes! of course} Michael Brennan @end display @sp 1 There are two other freely available @code{awk} implementations. This section briefly describes where to get them. @table @asis @cindex Kernighan, Brian @cindex anonymous @code{ftp} @cindex @code{ftp}, anonymous @item Unix @code{awk} Brian Kernighan has been able to make his implementation of @code{awk} freely available. You can get it via anonymous @code{ftp} to the host @code{@w{netlib.bell-labs.com}}. Change directory to @file{/netlib/research}. Use ``binary'' or ``image'' mode, and retrieve @file{awk.bundle.gz}. This is a shell archive that has been compressed with the GNU @code{gzip} utility. It can be uncompressed with the @code{gunzip} utility. You can also retrieve this version via the World Wide Web from @uref{http://cm.bell-labs.com/who/bwk, Brian Kernighan's home page}. This version requires an ANSI C compiler; GCC (the GNU C compiler) works quite nicely. @cindex Brennan, Michael @cindex @code{mawk} @item @code{mawk} Michael Brennan has written an independent implementation of @code{awk}, called @code{mawk}. It is available under the GPL (@pxref{Copying, ,GNU GENERAL PUBLIC LICENSE}), just as @code{gawk} is. You can get it via anonymous @code{ftp} to the host @code{@w{ftp.whidbey.net}}. Change directory to @file{/pub/brennan}. Use ``binary'' or ``image'' mode, and retrieve @file{mawk1.3.3.tar.gz} (or the latest version that is there). @code{gunzip} may be used to decompress this file. Installation is similar to @code{gawk}'s (@pxref{Unix Installation, , Compiling and Installing @code{gawk} on Unix}). @end table @node Notes, Glossary, Installation, Top @appendix Implementation Notes This appendix contains information mainly of interest to implementors and maintainers of @code{gawk}. Everything in it applies specifically to @code{gawk}, and not to other implementations. @menu * Compatibility Mode:: How to disable certain @code{gawk} extensions. * Additions:: Making Additions To @code{gawk}. * Future Extensions:: New features that may be implemented one day. * Improvements:: Suggestions for improvements by volunteers. @end menu @node Compatibility Mode, Additions, Notes, Notes @appendixsec Downward Compatibility and Debugging @xref{POSIX/GNU, ,Extensions in @code{gawk} Not in POSIX @code{awk}}, for a summary of the GNU extensions to the @code{awk} language and program. All of these features can be turned off by invoking @code{gawk} with the @samp{--traditional} option, or with the @samp{--posix} option. If @code{gawk} is compiled for debugging with @samp{-DDEBUG}, then there is one more option available on the command line: @table @code @item -W parsedebug @itemx --parsedebug Print out the parse stack information as the program is being parsed. @end table This option is intended only for serious @code{gawk} developers, and not for the casual user. It probably has not even been compiled into your version of @code{gawk}, since it slows down execution. @node Additions, Future Extensions, Compatibility Mode, Notes @appendixsec Making Additions to @code{gawk} If you should find that you wish to enhance @code{gawk} in a significant fashion, you are perfectly free to do so. That is the point of having free software; the source code is available, and you are free to change it as you wish (@pxref{Copying, ,GNU GENERAL PUBLIC LICENSE}). This section discusses the ways you might wish to change @code{gawk}, and any considerations you should bear in mind. @menu * Adding Code:: Adding code to the main body of @code{gawk}. * New Ports:: Porting @code{gawk} to a new operating system. @end menu @node Adding Code, New Ports, Additions, Additions @appendixsubsec Adding New Features @cindex adding new features @cindex features, adding You are free to add any new features you like to @code{gawk}. However, if you want your changes to be incorporated into the @code{gawk} distribution, there are several steps that you need to take in order to make it possible for me to include your changes. @enumerate 1 @item Get the latest version. It is much easier for me to integrate changes if they are relative to the most recent distributed version of @code{gawk}. If your version of @code{gawk} is very old, I may not be able to integrate them at all. @xref{Getting, ,Getting the @code{gawk} Distribution}, for information on getting the latest version of @code{gawk}. @item @iftex Follow the @cite{GNU Coding Standards}. @end iftex @ifinfo See @inforef{Top, , Version, standards, GNU Coding Standards}. @end ifinfo This document describes how GNU software should be written. If you haven't read it, please do so, preferably @emph{before} starting to modify @code{gawk}. (The @cite{GNU Coding Standards} are available as part of the Autoconf distribution, from the FSF.) @cindex @code{gawk} coding style @cindex coding style used in @code{gawk} @item Use the @code{gawk} coding style. The C code for @code{gawk} follows the instructions in the @cite{GNU Coding Standards}, with minor exceptions. The code is formatted using the traditional ``K&R'' style, particularly as regards the placement of braces and the use of tabs. In brief, the coding rules for @code{gawk} are: @itemize @bullet @item Use old style (non-prototype) function headers when defining functions. @item Put the name of the function at the beginning of its own line. @item Put the return type of the function, even if it is @code{int}, on the line above the line with the name and arguments of the function. @item The declarations for the function arguments should not be indented. @item Put spaces around parentheses used in control structures (@code{if}, @code{while}, @code{for}, @code{do}, @code{switch} and @code{return}). @item Do not put spaces in front of parentheses used in function calls. @item Put spaces around all C operators, and after commas in function calls. @item Do not use the comma operator to produce multiple side-effects, except in @code{for} loop initialization and increment parts, and in macro bodies. @item Use real tabs for indenting, not spaces. @item Use the ``K&R'' brace layout style. @item Use comparisons against @code{NULL} and @code{'\0'} in the conditions of @code{if}, @code{while} and @code{for} statements, and in the @code{case}s of @code{switch} statements, instead of just the plain pointer or character value. @item Use the @code{TRUE}, @code{FALSE}, and @code{NULL} symbolic constants, and the character constant @code{'\0'} where appropriate, instead of @code{1} and @code{0}. @item Provide one-line descriptive comments for each function. @item Do not use @samp{#elif}. Many older Unix C compilers cannot handle it. @item Do not use the @code{alloca} function for allocating memory off the stack. Its use causes more portability trouble than the minor benefit of not having to free the storage. Instead, use @code{malloc} and @code{free}. @end itemize If I have to reformat your code to follow the coding style used in @code{gawk}, I may not bother. @item Be prepared to sign the appropriate paperwork. In order for the FSF to distribute your changes, you must either place those changes in the public domain, and submit a signed statement to that effect, or assign the copyright in your changes to the FSF. Both of these actions are easy to do, and @emph{many} people have done so already. If you have questions, please contact me (@pxref{Bugs, , Reporting Problems and Bugs}), or @code{gnu@@gnu.org}. @item Update the documentation. Along with your new code, please supply new sections and or chapters for this @value{DOCUMENT}. If at all possible, please use real Texinfo, instead of just supplying unformatted ASCII text (although even that is better than no documentation at all). Conventions to be followed in @cite{@value{TITLE}} are provided after the @samp{@@bye} at the end of the Texinfo source file. If possible, please update the man page as well. You will also have to sign paperwork for your documentation changes. @item Submit changes as context diffs or unified diffs. Use @samp{diff -c -r -N} or @samp{diff -u -r -N} to compare the original @code{gawk} source tree with your version. (I find context diffs to be more readable, but unified diffs are more compact.) I recommend using the GNU version of @code{diff}. Send the output produced by either run of @code{diff} to me when you submit your changes. @xref{Bugs, , Reporting Problems and Bugs}, for the electronic mail information. Using this format makes it easy for me to apply your changes to the master version of the @code{gawk} source code (using @code{patch}). If I have to apply the changes manually, using a text editor, I may not do so, particularly if there are lots of changes. @end enumerate Although this sounds like a lot of work, please remember that while you may write the new code, I have to maintain it and support it, and if it isn't possible for me to do that with a minimum of extra work, then I probably will not. @node New Ports, , Adding Code, Additions @appendixsubsec Porting @code{gawk} to a New Operating System @cindex porting @code{gawk} If you wish to port @code{gawk} to a new operating system, there are several steps to follow. @enumerate 1 @item Follow the guidelines in @ref{Adding Code, ,Adding New Features}, concerning coding style, submission of diffs, and so on. @item When doing a port, bear in mind that your code must co-exist peacefully with the rest of @code{gawk}, and the other ports. Avoid gratuitous changes to the system-independent parts of the code. If at all possible, avoid sprinkling @samp{#ifdef}s just for your port throughout the code. If the changes needed for a particular system affect too much of the code, I probably will not accept them. In such a case, you will, of course, be able to distribute your changes on your own, as long as you comply with the GPL (@pxref{Copying, ,GNU GENERAL PUBLIC LICENSE}). @item A number of the files that come with @code{gawk} are maintained by other people at the Free Software Foundation. Thus, you should not change them unless it is for a very good reason. I.e.@: changes are not out of the question, but changes to these files will be scrutinized extra carefully. The files are @file{alloca.c}, @file{getopt.h}, @file{getopt.c}, @file{getopt1.c}, @file{regex.h}, @file{regex.c}, @file{dfa.h}, @file{dfa.c}, @file{install-sh}, and @file{mkinstalldirs}. @item Be willing to continue to maintain the port. Non-Unix operating systems are supported by volunteers who maintain the code needed to compile and run @code{gawk} on their systems. If no-one volunteers to maintain a port, that port becomes unsupported, and it may be necessary to remove it from the distribution. @item Supply an appropriate @file{gawkmisc.???} file. Each port has its own @file{gawkmisc.???} that implements certain operating system specific functions. This is cleaner than a plethora of @samp{#ifdef}s scattered throughout the code. The @file{gawkmisc.c} in the main source directory includes the appropriate @file{gawkmisc.???} file from each subdirectory. Be sure to update it as well. Each port's @file{gawkmisc.???} file has a suffix reminiscent of the machine or operating system for the port. For example, @file{pc/gawkmisc.pc} and @file{vms/gawkmisc.vms}. The use of separate suffixes, instead of plain @file{gawkmisc.c}, makes it possible to move files from a port's subdirectory into the main subdirectory, without accidentally destroying the real @file{gawkmisc.c} file. (Currently, this is only an issue for the MS-DOS and OS/2 ports.) @item Supply a @file{Makefile} and any other C source and header files that are necessary for your operating system. All your code should be in a separate subdirectory, with a name that is the same as, or reminiscent of, either your operating system or the computer system. If possible, try to structure things so that it is not necessary to move files out of the subdirectory into the main source directory. If that is not possible, then be sure to avoid using names for your files that duplicate the names of files in the main source directory. @item Update the documentation. Please write a section (or sections) for this @value{DOCUMENT} describing the installation and compilation steps needed to install and/or compile @code{gawk} for your system. @item Be prepared to sign the appropriate paperwork. In order for the FSF to distribute your code, you must either place your code in the public domain, and submit a signed statement to that effect, or assign the copyright in your code to the FSF. @ifinfo Both of these actions are easy to do, and @emph{many} people have done so already. If you have questions, please contact me, or @code{gnu@@gnu.org}. @end ifinfo @end enumerate Following these steps will make it much easier to integrate your changes into @code{gawk}, and have them co-exist happily with the code for other operating systems that is already there. In the code that you supply, and that you maintain, feel free to use a coding style and brace layout that suits your taste. @node Future Extensions, Improvements, Additions, Notes @appendixsec Probable Future Extensions @ignore From emory!scalpel.netlabs.com!lwall Tue Oct 31 12:43:17 1995 Return-Path: Message-Id: <9510311732.AA28472@scalpel.netlabs.com> To: arnold@skeeve.atl.ga.us (Arnold D. Robbins) Subject: Re: May I quote you? In-Reply-To: Your message of "Tue, 31 Oct 95 09:11:00 EST." Date: Tue, 31 Oct 95 09:32:46 -0800 From: Larry Wall : Greetings. I am working on the release of gawk 3.0. Part of it will be a : thoroughly updated manual. One of the sections deals with planned future : extensions and enhancements. I have the following at the beginning : of it: : : @cindex PERL : @cindex Wall, Larry : @display : @i{AWK is a language similar to PERL, only considerably more elegant.} @* : Arnold Robbins : @sp 1 : @i{Hey!} @* : Larry Wall : @end display : : Before I actually release this for publication, I wanted to get your : permission to quote you. (Hopefully, in the spirit of much of GNU, the : implied humor is visible... :-) I think that would be fine. Larry @end ignore @cindex PERL @cindex Wall, Larry @display @i{AWK is a language similar to PERL, only considerably more elegant.} Arnold Robbins @i{Hey!} Larry Wall @end display @sp 1 This section briefly lists extensions and possible improvements that indicate the directions we are currently considering for @code{gawk}. The file @file{FUTURES} in the @code{gawk} distributions lists these extensions as well. This is a list of probable future changes that will be usable by the @code{awk} language programmer. @c these are ordered by likelihood @table @asis @item Localization The GNU project is starting to support multiple languages. It will at least be possible to make @code{gawk} print its warnings and error messages in languages other than English. It may be possible for @code{awk} programs to also use the multiple language facilities, separate from @code{gawk} itself. @item Databases It may be possible to map a GDBM/NDBM/SDBM file into an @code{awk} array. @item A @code{PROCINFO} Array The special files that provide process-related information (@pxref{Special Files, ,Special File Names in @code{gawk}}) may be superseded by a @code{PROCINFO} array that would provide the same information, in an easier to access fashion. @item More @code{lint} warnings There are more things that could be checked for portability. @item Control of subprocess environment Changes made in @code{gawk} to the array @code{ENVIRON} may be propagated to subprocesses run by @code{gawk}. @ignore @item @code{RECLEN} variable for fixed length records Along with @code{FIELDWIDTHS}, this would speed up the processing of fixed-length records. @item A @code{restart} keyword After modifying @code{$0}, @code{restart} would restart the pattern matching loop, without reading a new record from the input. @item A @samp{|&} redirection The @samp{|&} redirection, in place of @samp{|}, would open a two-way pipeline for communication with a sub-process (via @code{getline} and @code{print} and @code{printf}). @item Function valued variables It would be possible to assign the name of a user-defined or built-in function to a regular @code{awk} variable, and then call the function indirectly, by using the regular variable. This would make it possible to write general purpose sorting and comparing routines, for example, by simply passing the name of one function into another. @item A built-in @code{stat} function The @code{stat} function would provide an easy-to-use hook to the @code{stat} system call so that @code{awk} programs could determine information about files. @item A built-in @code{ftw} function Combined with function valued variables and the @code{stat} function, @code{ftw} (file tree walk) would make it easy for an @code{awk} program to walk an entire file tree. @end ignore @end table This is a list of probable improvements that will make @code{gawk} perform better. @table @asis @item An Improved Version of @code{dfa} The @code{dfa} pattern matcher from GNU @code{grep} has some problems. Either a new version or a fixed one will deal with some important regexp matching issues. @item Use of GNU @code{malloc} The GNU version of @code{malloc} could potentially speed up @code{gawk}, since it relies heavily on the use of dynamic memory allocation. @end table @node Improvements, , Future Extensions, Notes @appendixsec Suggestions for Improvements Here are some projects that would-be @code{gawk} hackers might like to take on. They vary in size from a few days to a few weeks of programming, depending on which one you choose and how fast a programmer you are. Please send any improvements you write to the maintainers at the GNU project. @xref{Adding Code, , Adding New Features}, for guidelines to follow when adding new features to @code{gawk}. @xref{Bugs, ,Reporting Problems and Bugs}, for information on contacting the maintainers. @enumerate @item Compilation of @code{awk} programs: @code{gawk} uses a Bison (YACC-like) parser to convert the script given it into a syntax tree; the syntax tree is then executed by a simple recursive evaluator. This method incurs a lot of overhead, since the recursive evaluator performs many procedure calls to do even the simplest things. It should be possible for @code{gawk} to convert the script's parse tree into a C program which the user would then compile, using the normal C compiler and a special @code{gawk} library to provide all the needed functions (regexps, fields, associative arrays, type coercion, and so on). An easier possibility might be for an intermediate phase of @code{awk} to convert the parse tree into a linear byte code form like the one used in GNU Emacs Lisp. The recursive evaluator would then be replaced by a straight line byte code interpreter that would be intermediate in speed between running a compiled program and doing what @code{gawk} does now. @item The programs in the test suite could use documenting in this @value{DOCUMENT}. @item See the @file{FUTURES} file for more ideas. Contact us if you would seriously like to tackle any of the items listed there. @end enumerate @node Glossary, Copying, Notes, Top @appendix Glossary @table @asis @item Action A series of @code{awk} statements attached to a rule. If the rule's pattern matches an input record, @code{awk} executes the rule's action. Actions are always enclosed in curly braces. @xref{Action Overview, ,Overview of Actions}. @item Amazing @code{awk} Assembler Henry Spencer at the University of Toronto wrote a retargetable assembler completely as @code{awk} scripts. It is thousands of lines long, including machine descriptions for several eight-bit microcomputers. It is a good example of a program that would have been better written in another language. @item Amazingly Workable Formatter (@code{awf}) Henry Spencer at the University of Toronto wrote a formatter that accepts a large subset of the @samp{nroff -ms} and @samp{nroff -man} formatting commands, using @code{awk} and @code{sh}. @item ANSI The American National Standards Institute. This organization produces many standards, among them the standards for the C and C++ programming languages. @item Assignment An @code{awk} expression that changes the value of some @code{awk} variable or data object. An object that you can assign to is called an @dfn{lvalue}. The assigned values are called @dfn{rvalues}. @xref{Assignment Ops, ,Assignment Expressions}. @item @code{awk} Language The language in which @code{awk} programs are written. @item @code{awk} Program An @code{awk} program consists of a series of @dfn{patterns} and @dfn{actions}, collectively known as @dfn{rules}. For each input record given to the program, the program's rules are all processed in turn. @code{awk} programs may also contain function definitions. @item @code{awk} Script Another name for an @code{awk} program. @item Bash The GNU version of the standard shell (the Bourne-Again shell). See ``Bourne Shell.'' @item BBS See ``Bulletin Board System.'' @item Boolean Expression Named after the English mathematician Boole. See ``Logical Expression.'' @item Bourne Shell The standard shell (@file{/bin/sh}) on Unix and Unix-like systems, originally written by Steven R.@: Bourne. Many shells (Bash, @code{ksh}, @code{pdksh}, @code{zsh}) are generally upwardly compatible with the Bourne shell. @item Built-in Function The @code{awk} language provides built-in functions that perform various numerical, time stamp related, and string computations. Examples are @code{sqrt} (for the square root of a number) and @code{substr} (for a substring of a string). @xref{Built-in, ,Built-in Functions}. @item Built-in Variable @code{ARGC}, @code{ARGIND}, @code{ARGV}, @code{CONVFMT}, @code{ENVIRON}, @code{ERRNO}, @code{FIELDWIDTHS}, @code{FILENAME}, @code{FNR}, @code{FS}, @code{IGNORECASE}, @code{NF}, @code{NR}, @code{OFMT}, @code{OFS}, @code{ORS}, @code{RLENGTH}, @code{RSTART}, @code{RS}, @code{RT}, and @code{SUBSEP}, are the variables that have special meaning to @code{awk}. Changing some of them affects @code{awk}'s running environment. Several of these variables are specific to @code{gawk}. @xref{Built-in Variables}. @item Braces See ``Curly Braces.'' @item Bulletin Board System A computer system allowing users to log in and read and/or leave messages for other users of the system, much like leaving paper notes on a bulletin board. @item C The system programming language that most GNU software is written in. The @code{awk} programming language has C-like syntax, and this @value{DOCUMENT} points out similarities between @code{awk} and C when appropriate. @cindex ISO 8859-1 @cindex ISO Latin-1 @item Character Set The set of numeric codes used by a computer system to represent the characters (letters, numbers, punctuation, etc.) of a particular country or place. The most common character set in use today is ASCII (American Standard Code for Information Interchange). Many European countries use an extension of ASCII known as ISO-8859-1 (ISO Latin-1). @item CHEM A preprocessor for @code{pic} that reads descriptions of molecules and produces @code{pic} input for drawing them. It was written in @code{awk} by Brian Kernighan and Jon Bentley, and is available from @email{@w{netlib@@research.bell-labs.com}}. @item Compound Statement A series of @code{awk} statements, enclosed in curly braces. Compound statements may be nested. @xref{Statements, ,Control Statements in Actions}. @item Concatenation Concatenating two strings means sticking them together, one after another, giving a new string. For example, the string @samp{foo} concatenated with the string @samp{bar} gives the string @samp{foobar}. @xref{Concatenation, ,String Concatenation}. @item Conditional Expression An expression using the @samp{?:} ternary operator, such as @samp{@var{expr1} ? @var{expr2} : @var{expr3}}. The expression @var{expr1} is evaluated; if the result is true, the value of the whole expression is the value of @var{expr2}, otherwise the value is @var{expr3}. In either case, only one of @var{expr2} and @var{expr3} is evaluated. @xref{Conditional Exp, ,Conditional Expressions}. @item Comparison Expression A relation that is either true or false, such as @samp{(a < b)}. Comparison expressions are used in @code{if}, @code{while}, @code{do}, and @code{for} statements, and in patterns to select which input records to process. @xref{Typing and Comparison, ,Variable Typing and Comparison Expressions}. @item Curly Braces The characters @samp{@{} and @samp{@}}. Curly braces are used in @code{awk} for delimiting actions, compound statements, and function bodies. @item Dark Corner An area in the language where specifications often were (or still are) not clear, leading to unexpected or undesirable behavior. Such areas are marked in this @value{DOCUMENT} with ``(d.c.)'' in the text, and are indexed under the heading ``dark corner.'' @item Data Objects These are numbers and strings of characters. Numbers are converted into strings and vice versa, as needed. @xref{Conversion, ,Conversion of Strings and Numbers}. @item Double Precision An internal representation of numbers that can have fractional parts. Double precision numbers keep track of more digits than do single precision numbers, but operations on them are more expensive. This is the way @code{awk} stores numeric values. It is the C type @code{double}. @item Dynamic Regular Expression A dynamic regular expression is a regular expression written as an ordinary expression. It could be a string constant, such as @code{"foo"}, but it may also be an expression whose value can vary. @xref{Computed Regexps, , Using Dynamic Regexps}. @item Environment A collection of strings, of the form @var{name@code{=}val}, that each program has available to it. Users generally place values into the environment in order to provide information to various programs. Typical examples are the environment variables @code{HOME} and @code{PATH}. @item Empty String See ``Null String.'' @item Escape Sequences A special sequence of characters used for describing non-printing characters, such as @samp{\n} for newline, or @samp{\033} for the ASCII ESC (escape) character. @xref{Escape Sequences}. @item Field When @code{awk} reads an input record, it splits the record into pieces separated by whitespace (or by a separator regexp which you can change by setting the built-in variable @code{FS}). Such pieces are called fields. If the pieces are of fixed length, you can use the built-in variable @code{FIELDWIDTHS} to describe their lengths. @xref{Field Separators, ,Specifying How Fields are Separated}, and also see @xref{Constant Size, , Reading Fixed-width Data}. @item Floating Point Number Often referred to in mathematical terms as a ``rational'' number, this is just a number that can have a fractional part. See ``Double Precision'' and ``Single Precision.'' @item Format Format strings are used to control the appearance of output in the @code{printf} statement. Also, data conversions from numbers to strings are controlled by the format string contained in the built-in variable @code{CONVFMT}. @xref{Control Letters, ,Format-Control Letters}. @item Function A specialized group of statements used to encapsulate general or program-specific tasks. @code{awk} has a number of built-in functions, and also allows you to define your own. @xref{Built-in, ,Built-in Functions}, and @ref{User-defined, ,User-defined Functions}. @item FSF See ``Free Software Foundation.'' @item Free Software Foundation A non-profit organization dedicated to the production and distribution of freely distributable software. It was founded by Richard M.@: Stallman, the author of the original Emacs editor. GNU Emacs is the most widely used version of Emacs today. @item @code{gawk} The GNU implementation of @code{awk}. @item General Public License This document describes the terms under which @code{gawk} and its source code may be distributed. (@pxref{Copying, ,GNU GENERAL PUBLIC LICENSE}) @item GNU ``GNU's not Unix''. An on-going project of the Free Software Foundation to create a complete, freely distributable, POSIX-compliant computing environment. @item GPL See ``General Public License.'' @item Hexadecimal Base 16 notation, where the digits are @code{0}-@code{9} and @code{A}-@code{F}, with @samp{A} representing 10, @samp{B} representing 11, and so on up to @samp{F} for 15. Hexadecimal numbers are written in C using a leading @samp{0x}, to indicate their base. Thus, @code{0x12} is 18 (one times 16 plus 2). @item I/O Abbreviation for ``Input/Output,'' the act of moving data into and/or out of a running program. @item Input Record A single chunk of data read in by @code{awk}. Usually, an @code{awk} input record consists of one line of text. @xref{Records, ,How Input is Split into Records}. @item Integer A whole number, i.e.@: a number that does not have a fractional part. @item Keyword In the @code{awk} language, a keyword is a word that has special meaning. Keywords are reserved and may not be used as variable names. @code{gawk}'s keywords are: @code{BEGIN}, @code{END}, @code{if}, @code{else}, @code{while}, @code{do@dots{}while}, @code{for}, @code{for@dots{}in}, @code{break}, @code{continue}, @code{delete}, @code{next}, @code{nextfile}, @code{function}, @code{func}, and @code{exit}. @item Logical Expression An expression using the operators for logic, AND, OR, and NOT, written @samp{&&}, @samp{||}, and @samp{!} in @code{awk}. Often called Boolean expressions, after the mathematician who pioneered this kind of mathematical logic. @item Lvalue An expression that can appear on the left side of an assignment operator. In most languages, lvalues can be variables or array elements. In @code{awk}, a field designator can also be used as an lvalue. @item Null String A string with no characters in it. It is represented explicitly in @code{awk} programs by placing two double-quote characters next to each other (@code{""}). It can appear in input data by having two successive occurrences of the field separator appear next to each other. @item Number A numeric valued data object. The @code{gawk} implementation uses double precision floating point to represent numbers. Very old @code{awk} implementations use single precision floating point. @item Octal Base-eight notation, where the digits are @code{0}-@code{7}. Octal numbers are written in C using a leading @samp{0}, to indicate their base. Thus, @code{013} is 11 (one times 8 plus 3). @item Pattern Patterns tell @code{awk} which input records are interesting to which rules. A pattern is an arbitrary conditional expression against which input is tested. If the condition is satisfied, the pattern is said to @dfn{match} the input record. A typical pattern might compare the input record against a regular expression. @xref{Pattern Overview, ,Pattern Elements}. @item POSIX The name for a series of standards being developed by the IEEE that specify a Portable Operating System interface. The ``IX'' denotes the Unix heritage of these standards. The main standard of interest for @code{awk} users is @cite{IEEE Standard for Information Technology, Standard 1003.2-1992, Portable Operating System Interface (POSIX) Part 2: Shell and Utilities}. Informally, this standard is often referred to as simply ``P1003.2.'' @item Private Variables and/or functions that are meant for use exclusively by library functions, and not for the main @code{awk} program. Special care must be taken when naming such variables and functions. @xref{Library Names, , Naming Library Function Global Variables}. @item Range (of input lines) A sequence of consecutive lines from the input file. A pattern can specify ranges of input lines for @code{awk} to process, or it can specify single lines. @xref{Pattern Overview, ,Pattern Elements}. @item Recursion When a function calls itself, either directly or indirectly. If this isn't clear, refer to the entry for ``recursion.'' @item Redirection Redirection means performing input from other than the standard input stream, or output to other than the standard output stream. You can redirect the output of the @code{print} and @code{printf} statements to a file or a system command, using the @samp{>}, @samp{>>}, and @samp{|} operators. You can redirect input to the @code{getline} statement using the @samp{<} and @samp{|} operators. @xref{Redirection, ,Redirecting Output of @code{print} and @code{printf}}, and @ref{Getline, ,Explicit Input with @code{getline}}. @item Regexp Short for @dfn{regular expression}. A regexp is a pattern that denotes a set of strings, possibly an infinite set. For example, the regexp @samp{R.*xp} matches any string starting with the letter @samp{R} and ending with the letters @samp{xp}. In @code{awk}, regexps are used in patterns and in conditional expressions. Regexps may contain escape sequences. @xref{Regexp, ,Regular Expressions}. @item Regular Expression See ``regexp.'' @item Regular Expression Constant A regular expression constant is a regular expression written within slashes, such as @code{/foo/}. This regular expression is chosen when you write the @code{awk} program, and cannot be changed doing its execution. @xref{Regexp Usage, ,How to Use Regular Expressions}. @item Rule A segment of an @code{awk} program that specifies how to process single input records. A rule consists of a @dfn{pattern} and an @dfn{action}. @code{awk} reads an input record; then, for each rule, if the input record satisfies the rule's pattern, @code{awk} executes the rule's action. Otherwise, the rule does nothing for that input record. @item Rvalue A value that can appear on the right side of an assignment operator. In @code{awk}, essentially every expression has a value. These values are rvalues. @item @code{sed} See ``Stream Editor.'' @item Short-Circuit The nature of the @code{awk} logical operators @samp{&&} and @samp{||}. If the value of the entire expression can be deduced from evaluating just the left-hand side of these operators, the right-hand side will not be evaluated (@pxref{Boolean Ops, ,Boolean Expressions}). @item Side Effect A side effect occurs when an expression has an effect aside from merely producing a value. Assignment expressions, increment and decrement expressions and function calls have side effects. @xref{Assignment Ops, ,Assignment Expressions}. @item Single Precision An internal representation of numbers that can have fractional parts. Single precision numbers keep track of fewer digits than do double precision numbers, but operations on them are less expensive in terms of CPU time. This is the type used by some very old versions of @code{awk} to store numeric values. It is the C type @code{float}. @item Space The character generated by hitting the space bar on the keyboard. @item Special File A file name interpreted internally by @code{gawk}, instead of being handed directly to the underlying operating system. For example, @file{/dev/stderr}. @xref{Special Files, ,Special File Names in @code{gawk}}. @item Stream Editor A program that reads records from an input stream and processes them one or more at a time. This is in contrast with batch programs, which may expect to read their input files in entirety before starting to do anything, and with interactive programs, which require input from the user. @item String A datum consisting of a sequence of characters, such as @samp{I am a string}. Constant strings are written with double-quotes in the @code{awk} language, and may contain escape sequences. @xref{Escape Sequences}. @item Tab The character generated by hitting the @kbd{TAB} key on the keyboard. It usually expands to up to eight spaces upon output. @item Unix A computer operating system originally developed in the early 1970's at AT&T Bell Laboratories. It initially became popular in universities around the world, and later moved into commercial evnironments as a software development system and network server system. There are many commercial versions of Unix, as well as several work-alike systems whose source code is freely available (such as Linux, NetBSD, and FreeBSD). @item Whitespace A sequence of space, tab, or newline characters occurring inside an input record or a string. @end table @node Copying, Index, Glossary, Top @unnumbered GNU GENERAL PUBLIC LICENSE @center Version 2, June 1991 @display Copyright @copyright{} 1989, 1991 Free Software Foundation, Inc. 59 Temple Place --- Suite 330, Boston, MA 02111-1307, USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. @end display @c fakenode --- for prepinfo @unnumberedsec Preamble The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software---to make sure the software is free for all its users. This General Public License applies to most of the Free Software Foundation's software and to any other program whose authors commit to using it. (Some other Free Software Foundation software is covered by the GNU Library General Public License instead.) You can apply it to your programs, too. When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things. To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it. For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights. We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software. Also, for each author's protection and ours, we want to make certain that everyone understands that there is no warranty for this free software. If the software is modified by someone else and passed on, we want its recipients to know that what they have is not the original, so that any problems introduced by others will not reflect on the original authors' reputations. Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all. The precise terms and conditions for copying, distribution and modification follow. @iftex @c fakenode --- for prepinfo @unnumberedsec TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION @end iftex @ifinfo @center TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION @end ifinfo @enumerate 0 @item This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The ``Program'', below, refers to any such program or work, and a ``work based on the Program'' means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term ``modification''.) Each licensee is addressed as ``you''. Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does. @item You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program. You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee. @item You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions: @enumerate a @item You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change. @item You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. @item If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.) @end enumerate These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program. In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. @item You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following: @enumerate a @item Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, @item Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, @item Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for non-commercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.) @end enumerate The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable. If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code. @item You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance. @item You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it. @item Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License. @item If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program. If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply and the section as a whole is intended to apply in other circumstances. It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system, which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice. This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License. @item If the distribution and/or use of the Program is restricted in certain countries either by patents or by copyrighted interfaces, the original copyright holder who places the Program under this License may add an explicit geographical distribution limitation excluding those countries, so that distribution is permitted only in or among countries not thus excluded. In such case, this License incorporates the limitation as if written in the body of this License. @item The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and ``any later version'', you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. @item If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission. For software which is copyrighted by the Free Software Foundation, write to the Free Software Foundation; we sometimes make exceptions for this. Our decision will be guided by the two goals of preserving the free status of all derivatives of our free software and of promoting the sharing and reuse of software generally. @iftex @c fakenode --- for prepinfo @heading NO WARRANTY @end iftex @ifinfo @center NO WARRANTY @end ifinfo @item BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW@. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM ``AS IS'' WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE@. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU@. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. @item IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. @end enumerate @iftex @c fakenode --- for prepinfo @heading END OF TERMS AND CONDITIONS @end iftex @ifinfo @center END OF TERMS AND CONDITIONS @end ifinfo @page @c fakenode --- for prepinfo @unnumberedsec How to Apply These Terms to Your New Programs If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms. To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively convey the exclusion of warranty; and each file should have at least the ``copyright'' line and a pointer to where the full notice is found. @smallexample @var{one line to give the program's name and an idea of what it does.} Copyright (C) 19@var{yy} @var{name of author} This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE@. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place --- Suite 330, Boston, MA 02111-1307, USA. @end smallexample Also add information on how to contact you by electronic and paper mail. If the program is interactive, make it output a short notice like this when it starts in an interactive mode: @smallexample Gnomovision version 69, Copyright (C) 19@var{yy} @var{name of author} Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details. @end smallexample The hypothetical commands @samp{show w} and @samp{show c} should show the appropriate parts of the General Public License. Of course, the commands you use may be called something other than @samp{show w} and @samp{show c}; they could even be mouse-clicks or menu items---whatever suits your program. You should also get your employer (if you work as a programmer) or your school, if any, to sign a ``copyright disclaimer'' for the program, if necessary. Here is a sample; alter the names: @smallexample @group Yoyodyne, Inc., hereby disclaims all copyright interest in the program `Gnomovision' (which makes passes at compilers) written by James Hacker. @var{signature of Ty Coon}, 1 April 1989 Ty Coon, President of Vice @end group @end smallexample This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Library General Public License instead of this License. @node Index, , Copying, Top @unnumbered Index @printindex cp @summarycontents @contents @bye Unresolved Issues: ------------------ 1. From ADR. Robert J. Chassell points out that awk programs should have some indication of how to use them. It would be useful to perhaps have a "programming style" section of the manual that would include this and other tips. 2. The default AWKPATH search path should be configurable via `configure' The default and how this changes needs to be documented. Consistency issues: /.../ regexps are in @code, not @samp ".." strings are in @code, not @samp no @print before @dots values of expressions in the text (@code{x} has the value 15), should be in roman, not @code Use tab and not TAB Use ESC and not ESCAPE Use space and not blank to describe the space bar's character The term "blank" is thus basically reserved for "blank lines" etc. The `(d.c.)' should appear inside the closing `.' of a sentence It should come before (pxref{...}) " " should have an @w{} around it Use "non-" everywhere Use @code{ftp} when talking about anonymous ftp Use upper-case and lower-case, not "upper case" and "lower case" Use alphanumeric, not alpha-numeric Use --foo, not -Wfoo when describing long options Use findex for all programs and functions in the example chapters Use "Bell Laboratories", but not "Bell Labs". Use "behavior" instead of "behaviour". Use "zeros" instead of "zeroes". Use "Input/Output", not "input/output". Also "I/O", not "i/o". Use @code{do}, and not @code{do}-@code{while}, except where actually discussing the do-while. The words "a", "and", "as", "between", "for", "from", "in", "of", "on", "that", "the", "to", "with", and "without", should not be capitalized in @chapter, @section etc. "Into" and "How" should. Search for @dfn; make sure important items are also indexed. "e.g." should always be followed by a comma. "i.e." should never be followed by a comma, and should be followed by `@:'. The numbers zero through ten should be spelled out, except when talking about file descriptor numbers. > 10 and < 0, it's ok to use numbers. In tables, put command line options in @code, while in the text, put them in @samp. When using @strong, use "Note:" or "Caution:" with colons and not exclamation points. Do not surround the paragraphs with @quotation ... @end quotation. Date: Wed, 13 Apr 94 15:20:52 -0400 From: rsm@gnu.ai.mit.edu (Richard Stallman) To: gnu-prog@gnu.ai.mit.edu Subject: A reminder: no pathnames in GNU It's a GNU convention to use the term "file name" for the name of a file, never "pathname". We use the term "path" for search paths, which are lists of file names. Using it for a single file name as well is potentially confusing to users. So please check any documentation you maintain, if you think you might have used "pathname". Note that "file name" should be two words when it appears as ordinary text. It's ok as one word when it's a metasyntactic variable, though. Suggestions: ------------ Enhance FIELDWIDTHS with some way to indicate "the rest of the record". E.g., a length of 0 or -1 or something. May be "n"? Make FIELDWIDTHS be an array? What if FIELDWIDTHS has invalid values in it?