mirror of
https://git.savannah.gnu.org/git/emacs.git
synced 2024-12-15 09:47:20 +00:00
986 lines
34 KiB
Plaintext
986 lines
34 KiB
Plaintext
\input texinfo @c -*-texinfo-*-
|
|
@c documentation for forms-mode
|
|
@c Written by Johan Vromans, and edited by Richard Stallman
|
|
|
|
@comment %**start of header (This is for running Texinfo on a region.)
|
|
@setfilename ../info/forms
|
|
@settitle Forms Mode User's Manual
|
|
@syncodeindex vr cp
|
|
@syncodeindex fn cp
|
|
@syncodeindex ky cp
|
|
@iftex
|
|
@finalout
|
|
@setchapternewpage odd
|
|
@end iftex
|
|
@c @smallbook
|
|
@comment %**end of header (This is for running Texinfo on a region.)
|
|
|
|
@copying
|
|
This file documents Forms mode, a form-editing major mode for GNU Emacs.
|
|
|
|
Copyright @copyright{} 1989, 1997, 2001, 2002, 2003, 2004,
|
|
2005, 2006, 2007 Free Software Foundation, Inc.
|
|
|
|
@quotation
|
|
Permission is granted to copy, distribute and/or modify this document
|
|
under the terms of the GNU Free Documentation License, Version 1.2 or
|
|
any later version published by the Free Software Foundation; with no
|
|
Invariant Sections, with the Front-Cover texts being ``A GNU
|
|
Manual,'' and with the Back-Cover Texts as in (a) below. A copy of the
|
|
license is included in the section entitled ``GNU Free Documentation
|
|
License'' in the Emacs manual.
|
|
|
|
(a) The FSF's Back-Cover Text is: ``You have freedom to copy and modify
|
|
this GNU Manual, like GNU software. Copies published by the Free
|
|
Software Foundation raise funds for GNU development.''
|
|
|
|
This document is part of a collection distributed under the GNU Free
|
|
Documentation License. If you want to distribute this document
|
|
separately from the collection, you can do so by adding a copy of the
|
|
license to the document, as described in section 6 of the license.
|
|
@end quotation
|
|
@end copying
|
|
|
|
@dircategory Emacs
|
|
@direntry
|
|
* Forms: (forms). Emacs package for editing data bases
|
|
by filling in forms.
|
|
@end direntry
|
|
|
|
@titlepage
|
|
@sp 6
|
|
@center @titlefont{Forms Mode User's Manual}
|
|
@sp 4
|
|
@center Forms-Mode version 2
|
|
@sp 1
|
|
@center for GNU Emacs 22.1
|
|
@sp 1
|
|
@center April 2007
|
|
@sp 5
|
|
@center Johan Vromans
|
|
@center @i{jvromans@@squirrel.nl}
|
|
@page
|
|
@vskip 0pt plus 1filll
|
|
@insertcopying
|
|
@end titlepage
|
|
|
|
@ifnottex
|
|
@node Top
|
|
@top Forms Mode
|
|
|
|
Forms mode is an Emacs major mode for working with simple textual data
|
|
bases in a forms-oriented manner. In Forms mode, the information in
|
|
these files is presented in an Emacs window in a user-defined format,
|
|
one record at a time. The user can view records or modify their
|
|
contents.
|
|
|
|
Forms mode is not a simple major mode, but requires two files to do its
|
|
job: a control file and a data file. The data file holds the
|
|
actual data to be presented. The control file describes
|
|
how to present it.
|
|
|
|
@menu
|
|
* Forms Example:: An example: editing the password data base.
|
|
* Entering and Exiting Forms Mode::
|
|
How to visit a file in Forms mode.
|
|
* Forms Commands:: Special commands to use while in Forms mode.
|
|
* Data File Format:: How to format the data file.
|
|
* Control File Format:: How to control forms mode.
|
|
* Format Description:: How to define the forms layout.
|
|
* Modifying Forms Contents:: How to modify.
|
|
* Miscellaneous:: Forms mode messages and other remarks.
|
|
* Error Messages:: List of error messages forms mode can produce.
|
|
* Long Example:: A more complex control file example.
|
|
* GNU Free Documentation License:: The license for this documentation.
|
|
* Credits:: Thanks everyone.
|
|
* Index:: Index to this manual.
|
|
@end menu
|
|
@end ifnottex
|
|
|
|
@node Forms Example
|
|
@chapter Forms Example
|
|
|
|
Let's illustrate Forms mode with an example. Suppose you are looking at
|
|
the @file{/etc/passwd} file, and the screen looks like this:
|
|
|
|
@example
|
|
====== /etc/passwd ======
|
|
|
|
User : root Uid: 0 Gid: 1
|
|
|
|
Name : Super User
|
|
|
|
Home : /
|
|
|
|
Shell: /bin/sh
|
|
@end example
|
|
|
|
As you can see, the familiar fields from the entry for the super user
|
|
are all there, but instead of being colon-separated on one single line,
|
|
they make up a forms.
|
|
|
|
The contents of the forms consist of the contents of the fields of the
|
|
record (e.g. @samp{root}, @samp{0}, @samp{1}, @samp{Super User})
|
|
interspersed with normal text (e.g @samp{User : }, @samp{Uid: }).
|
|
|
|
If you modify the contents of the fields, Forms mode will analyze your
|
|
changes and update the file appropriately. You cannot modify the
|
|
interspersed explanatory text (unless you go to some trouble about it),
|
|
because that is marked read-only (@pxref{Text Properties,,, elisp, The
|
|
Emacs Lisp Reference Manual}).
|
|
|
|
The Forms mode control file specifies the relationship between the
|
|
format of @file{/etc/passwd} and what appears on the screen in Forms
|
|
mode. @xref{Control File Format}.
|
|
|
|
@node Entering and Exiting Forms Mode
|
|
@chapter Entering and Exiting Forms Mode
|
|
|
|
@table @kbd
|
|
@findex forms-find-file
|
|
@item M-x forms-find-file @key{RET} @var{control-file} @key{RET}
|
|
Visit a database using Forms mode. Specify the name of the
|
|
@strong{control file}, not the data file!
|
|
|
|
@findex forms-find-file-other-window
|
|
@item M-x forms-find-file-other-window @key{RET} @var{control-file} @key{RET}
|
|
Similar, but displays the file in another window.
|
|
@end table
|
|
|
|
The command @code{forms-find-file} evaluates the file
|
|
@var{control-file}, and also visits it in Forms mode. What you see in
|
|
its buffer is not the contents of this file, but rather a single record
|
|
of the corresponding data file that is visited in its own buffer. So
|
|
there are two buffers involved in Forms mode: the @dfn{forms buffer}
|
|
that is initially used to visit the control file and that shows the
|
|
records being browsed, and the @dfn{data buffer} that holds the data
|
|
file being visited. The latter buffer is normally not visible.
|
|
|
|
Initially, the first record is displayed in the forms buffer.
|
|
The mode line displays the major mode name @samp{Forms}, followed by the
|
|
minor mode @samp{View} if the data base is read-only. The number of the
|
|
current record (@var{n}) and the total number of records in the
|
|
file(@var{t}) are shown in the mode line as @samp{@var{n}/@var{t}}. For
|
|
example:
|
|
|
|
@example
|
|
--%%-Emacs: passwd-demo (Forms View 1/54)----All-------
|
|
@end example
|
|
|
|
If the buffer is not read-only, you may change the buffer to modify the
|
|
fields in the record. When you move to a different record, the contents
|
|
of the buffer are parsed using the specifications in
|
|
@code{forms-format-list}, and the data file is updated. If the record
|
|
has fields that aren't included in the display, they are not changed.
|
|
|
|
@vindex forms-mode-hooks
|
|
Entering Forms mode runs the normal hook @code{forms-mode-hooks} to
|
|
perform user-defined customization.
|
|
|
|
To save any modified data, you can use @kbd{C-x C-s}
|
|
(@code{forms-save-buffer}). This does not save the forms buffer (which would
|
|
be rather useless), but instead saves the buffer visiting the data file.
|
|
|
|
To terminate Forms mode, you can use @kbd{C-x C-s} (@code{forms-save-buffer})
|
|
and then kill the forms buffer. However, the data buffer will still
|
|
remain. If this is not desired, you have to kill this buffer too.
|
|
|
|
@node Forms Commands
|
|
@chapter Forms Commands
|
|
|
|
The commands of Forms mode belong to the @kbd{C-c} prefix, with one
|
|
exception: @key{TAB}, which moves to the next field. Forms mode uses
|
|
different key maps for normal mode and read-only mode. In read-only
|
|
Forms mode, you can access most of the commands without the @kbd{C-c}
|
|
prefix, but you must type ordinary letters instead of control
|
|
characters; for example, type @kbd{n} instead of @kbd{C-c C-n}.
|
|
|
|
If your Emacs has been built with X-toolkit support, Forms mode will
|
|
provide its own menu with a number of Forms mode commands.
|
|
|
|
@table @kbd
|
|
@findex forms-next-record
|
|
@kindex C-c C-n
|
|
@item C-c C-n
|
|
Show the next record (@code{forms-next-record}). With a numeric
|
|
argument @var{n}, show the @var{n}th next record.
|
|
|
|
@findex forms-prev-record
|
|
@kindex C-c C-p
|
|
@item C-c C-p
|
|
Show the previous record (@code{forms-prev-record}). With a numeric
|
|
argument @var{n}, show the @var{n}th previous record.
|
|
|
|
@findex forms-jump-record
|
|
@kindex C-c C-l
|
|
@item C-c C-l
|
|
Jump to a record by number (@code{forms-jump-record}). Specify
|
|
the record number with a numeric argument.
|
|
|
|
@findex forms-first-record
|
|
@kindex C-c <
|
|
@item C-c <
|
|
Jump to the first record (@code{forms-first-record}).
|
|
|
|
@findex forms-last-record
|
|
@kindex C-c >
|
|
@item C-c >
|
|
Jump to the last record (@code{forms-last-record}). This command also
|
|
recalculates the number of records in the data file.
|
|
|
|
@findex forms-next-field
|
|
@kindex TAB
|
|
@item @key{TAB}
|
|
@kindex C-c TAB
|
|
@itemx C-c @key{TAB}
|
|
Jump to the next field in the current record (@code{forms-next-field}).
|
|
With a numeric argument @var{n}, jump forward @var{n} fields. If this command
|
|
would move past the last field, it wraps around to the first field.
|
|
|
|
@findex forms-toggle-read-only
|
|
@kindex C-c C-q
|
|
@item C-c C-q
|
|
Toggles read-only mode (@code{forms-toggle-read-only}). In read-only
|
|
Forms mode, you cannot edit the fields; most Forms mode commands can be
|
|
accessed without the prefix @kbd{C-c} if you use the normal letter
|
|
instead (for example, type @kbd{n} instead of @kbd{C-c C-n}). In edit
|
|
mode, you can edit the fields and thus change the contents of the data
|
|
base; you must begin Forms mode commands with @code{C-c}. Switching
|
|
to edit mode is allowed only if you have write access to the data file.
|
|
|
|
@findex forms-insert-record
|
|
@kindex C-c C-o
|
|
@item C-c C-o
|
|
Create a new record and insert it before the current record
|
|
(@code{forms-insert-record}). It starts out with empty (or default)
|
|
contents for its fields; you can then edit the fields. With a numeric
|
|
argument, the new record is created @emph{after} the current one.
|
|
See also @code{forms-modified-record-filter} in @ref{Modifying Forms
|
|
Contents}.
|
|
|
|
@findex forms-delete-record
|
|
@kindex C-c C-k
|
|
@item C-c C-k
|
|
Delete the current record (@code{forms-delete-record}). You are
|
|
prompted for confirmation before the record is deleted unless a numeric
|
|
argument has been provided.
|
|
|
|
@findex forms-search-forward
|
|
@kindex C-c C-s @var{regexp} @key{RET}
|
|
@item C-c C-s @var{regexp} @key{RET}
|
|
Search forward for @var{regexp} in all records following this one
|
|
(@code{forms-search-forward}). If found, this record is shown.
|
|
If you give an empty argument, the previous regexp is used again.
|
|
|
|
@findex forms-search-backward
|
|
@kindex C-c C-r @var{regexp} @key{RET}
|
|
@item C-c C-r @var{regexp} @key{RET}
|
|
Search backward for @var{regexp} in all records following this one
|
|
(@code{forms-search-backward}). If found, this record is shown.
|
|
If you give an empty argument, the previous regexp is used again.
|
|
|
|
@ignore
|
|
@findex forms-exit
|
|
@kindex C-c C-x
|
|
@item C-c C-x
|
|
Terminate Forms mode processing (@code{forms-exit}). The data file is
|
|
saved if it has been modified.
|
|
|
|
@findex forms-exit-no-save
|
|
@item M-x forms-exit-no-save
|
|
Terminates forms mode processing without saving modified data first.
|
|
@end ignore
|
|
|
|
@findex forms-prev-field
|
|
@item M-x forms-prev-field
|
|
Similar to @code{forms-next-field} but moves backwards.
|
|
|
|
@findex forms-save-buffer
|
|
@item M-x forms-save-buffer
|
|
@kindex C-x C-s
|
|
@itemx C-x C-s
|
|
Forms mode replacement for @code{save-buffer}. When executed in the
|
|
forms buffer it will save the contents of the (modified) data buffer
|
|
instead. In Forms mode this function will be bound to @kbd{C-x C-s}.
|
|
|
|
@findex forms-print
|
|
@item M-x forms-print
|
|
This command can be used to make a formatted print
|
|
of the contents of the data file.
|
|
|
|
@end table
|
|
|
|
In addition the command @kbd{M-x revert-buffer} is useful in Forms mode
|
|
just as in other modes.
|
|
|
|
@ignore
|
|
@vindex forms-forms-scroll
|
|
@findex scroll-up
|
|
@findex scroll-down
|
|
If the variable @code{forms-forms-scrolls} is set to a value other
|
|
than @code{nil} (which it is, by default), the Emacs functions
|
|
@code{scroll-up} and @code{scroll-down} will perform a
|
|
@code{forms-next-record} and @code{forms-prev-record} when in forms
|
|
mode. So you can use your favorite page commands to page through the
|
|
data file.
|
|
|
|
@vindex forms-forms-jump
|
|
@findex beginning-of-buffer
|
|
@findex end-of-buffer
|
|
Likewise, if the variable @code{forms-forms-jump} is not @code{nil}
|
|
(which it is, by default), Emacs functions @code{beginning-of-buffer}
|
|
and @code{end-of-buffer} will perform @code{forms-first-record} and
|
|
@code{forms-last-record} when in forms mode.
|
|
@end ignore
|
|
|
|
The following function key definitions are set up in Forms mode
|
|
(whether read-only or not):
|
|
|
|
@table @kbd
|
|
@kindex next
|
|
@item next
|
|
forms-next-record
|
|
|
|
@kindex prior
|
|
@item prior
|
|
forms-prev-record
|
|
|
|
@kindex begin
|
|
@item begin
|
|
forms-first-record
|
|
|
|
@kindex end
|
|
@item end
|
|
forms-last-record
|
|
|
|
@kindex S-Tab
|
|
@findex forms-prev-field
|
|
@item S-Tab
|
|
forms-prev-field
|
|
@end table
|
|
|
|
@node Data File Format
|
|
@chapter Data File Format
|
|
|
|
@cindex record
|
|
@cindex field
|
|
@vindex forms-field-sep
|
|
Files for use with Forms mode are very simple---each @dfn{record}
|
|
(usually one line) forms the contents of one form. Each record consists
|
|
of a number of @dfn{fields}, which are separated by the value of the
|
|
string @code{forms-field-sep}, which is @code{"\t"} (a Tab) by default.
|
|
|
|
@vindex forms-read-file-filter
|
|
@vindex forms-write-file-filter
|
|
If the format of the data file is not suitable enough you can define the
|
|
filter functions @code{forms-read-file-filter} and
|
|
@code{forms-write-file-filter}. @code{forms-read-file-filter} is called
|
|
when the data file is read from disk into the data buffer. It operates
|
|
on the data buffer, ignoring read-only protections. When the data file
|
|
is saved to disk @code{forms-write-file-filter} is called to cancel the
|
|
effects of @code{forms-read-file-filter}. After being saved,
|
|
@code{forms-read-file-filter} is called again to prepare the data buffer
|
|
for further processing.
|
|
|
|
@cindex pseudo-newline
|
|
@vindex forms-multi-line
|
|
Fields may contain text which shows up in the forms in multiple lines.
|
|
These lines are separated in the field using a ``pseudo-newline''
|
|
character which is defined by the value of the string
|
|
@code{forms-multi-line}. Its default value is @code{"\^k"} (a Control-K
|
|
character). If it is
|
|
set to @code{nil}, multiple line fields are prohibited.
|
|
|
|
If the data file does not exist, it is automatically created.
|
|
|
|
@node Control File Format
|
|
@chapter Control File Format
|
|
|
|
@cindex control file
|
|
The Forms mode @dfn{control file} serves two purposes. First, it names
|
|
the data file to use, and defines its format and properties. Second,
|
|
the Emacs buffer it occupies is used by Forms mode to display the forms.
|
|
|
|
The contents of the control file are evaluated as a Lisp program. It
|
|
should set the following Lisp variables to suitable values:
|
|
|
|
@table @code
|
|
@vindex forms-file
|
|
@item forms-file
|
|
This variable specifies the name of the data file. Example:
|
|
|
|
@example
|
|
(setq forms-file "my/data-file")
|
|
@end example
|
|
|
|
If the control file doesn't set @code{forms-file}, Forms mode
|
|
reports an error.
|
|
|
|
@vindex forms-format-list
|
|
@item forms-format-list
|
|
This variable describes the way the fields of the record are formatted on
|
|
the screen. For details, see @ref{Format Description}.
|
|
|
|
@vindex forms-number-of-fields
|
|
@item forms-number-of-fields
|
|
This variable holds the number of fields in each record of the data
|
|
file. Example:
|
|
|
|
@example
|
|
(setq forms-number-of-fields 10)
|
|
@end example
|
|
@end table
|
|
|
|
If the control file does not set @code{forms-format-list} a default
|
|
format is used. In this situation, Forms mode will deduce the number of
|
|
fields from the data file providing this file exists and
|
|
@code{forms-number-of-records} has not been set in the control file.
|
|
|
|
The control file can optionally set the following additional Forms mode
|
|
variables. Most of them have default values that are good for most
|
|
applications.
|
|
|
|
@table @code
|
|
@vindex forms-field-sep
|
|
@item forms-field-sep
|
|
This variable may be used to designate the string which separates the
|
|
fields in the records of the data file. If not set, it defaults to the
|
|
string @code{"\t"} (a Tab character). Example:
|
|
|
|
@example
|
|
(setq forms-field-sep "\t")
|
|
@end example
|
|
|
|
@vindex forms-read-only
|
|
@item forms-read-only
|
|
If the value is non-@code{nil}, the data file is treated read-only. (Forms
|
|
mode also treats the data file as read-only if you don't have access to
|
|
write it.) Example:
|
|
|
|
@example
|
|
(set forms-read-only t)
|
|
@end example
|
|
|
|
@vindex forms-multi-line
|
|
@item forms-multi-line
|
|
This variable specifies the @dfn{pseudo newline} separator that allows
|
|
multi-line fields. This separator goes between the ``lines'' within a
|
|
field---thus, the field doesn't really contain multiple lines, but it
|
|
appears that way when displayed in Forms mode. If the value is
|
|
@code{nil}, multi-line text fields are prohibited. The pseudo newline
|
|
must not be a character contained in @code{forms-field-sep}.
|
|
|
|
The default value is @code{"\^k"}, the character Control-K. Example:
|
|
|
|
@example
|
|
(setq forms-multi-line "\^k")
|
|
@end example
|
|
|
|
@ignore
|
|
@vindex forms-forms-scroll
|
|
@item forms-forms-scroll
|
|
@xref{Forms Mode Commands}, for details.
|
|
|
|
@vindex forms-forms-jump
|
|
@item forms-forms-jump
|
|
@xref{Forms Mode Commands}, for details.
|
|
@end ignore
|
|
|
|
@findex forms-read-file-filter
|
|
@item forms-read-file-filter
|
|
This variable holds the name of a function to be called after the data
|
|
file has been read in. This can be used to transform the contents of the
|
|
data file into a format more suitable for forms processing.
|
|
If it is @code{nil}, no function is called. For example, to maintain a
|
|
gzipped database:
|
|
|
|
@example
|
|
(defun gzip-read-file-filter ()
|
|
(shell-command-on-region (point-min) (point-max)
|
|
"gzip -d" t t))
|
|
(setq forms-read-file-filter 'gzip-read-file-filter)
|
|
@end example
|
|
|
|
@findex forms-write-file-filter
|
|
@item forms-write-file-filter
|
|
This variable holds the name of a function to be called before writing
|
|
out the contents of the data file.
|
|
This can be used to undo the effects of @code{forms-read-file-filter}.
|
|
If it is @code{nil}, no function is called. Example:
|
|
|
|
@example
|
|
(defun gzip-write-file-filter ()
|
|
(make-variable-buffer-local 'require-final-newline)
|
|
(setq require-final-newline nil)
|
|
(shell-command-on-region (point-min) (point-max)
|
|
"gzip" t t))
|
|
(setq forms-write-file-filter 'gzip-write-file-filter)
|
|
@end example
|
|
|
|
@findex forms-new-record-filter
|
|
@item forms-new-record-filter
|
|
This variable holds a function to be called whenever a new record is created
|
|
to supply default values for fields. If it is @code{nil}, no function is
|
|
called.
|
|
@xref{Modifying Forms Contents}, for details.
|
|
|
|
@findex forms-modified-record-filter
|
|
@item forms-modified-record-filter
|
|
This variable holds a function to be called whenever a record is
|
|
modified, just before updating the Forms data file. If it is
|
|
@code{nil}, no function is called.
|
|
@xref{Modifying Forms Contents}, for details.
|
|
|
|
@findex forms-insert-after
|
|
@item forms-insert-after
|
|
If this variable is not @code{nil}, new records are created @emph{after} the
|
|
current record. Also, upon visiting a file, the initial position will be
|
|
at the last record instead of the first one.
|
|
|
|
@findex forms-check-number-of-fields
|
|
@item forms-check-number-of-fields
|
|
Normally each record is checked to contain the correct number of fields.
|
|
Under certain circumstances, this can be undesirable.
|
|
If this variable is set to @code{nil}, these checks will be bypassed.
|
|
@end table
|
|
|
|
@node Format Description
|
|
@chapter The Format Description
|
|
|
|
@vindex forms-format-list
|
|
The variable @code{forms-format-list} specifies the format of the data
|
|
in the data file, and how to convert the data for display in Forms mode.
|
|
Its value must be a list of Forms mode @dfn{formatting elements}, each
|
|
of which can be a string, a number, a Lisp list, or a Lisp symbol that
|
|
evaluates to one of those. The formatting elements are processed in the
|
|
order they appear in the list.
|
|
|
|
@table @var
|
|
@item string
|
|
A string formatting element is inserted in the forms ``as is,'' as text
|
|
that the user cannot alter.
|
|
|
|
@item number
|
|
A number element selects a field of the record. The contents of this
|
|
field are inserted in the display at this point. Field numbers count
|
|
starting from 1 (one).
|
|
|
|
@item list
|
|
A formatting element that is a list specifies a function call. This
|
|
function is called every time a record is displayed, and its result,
|
|
which must be a string, is inserted in the display text. The function
|
|
should do nothing but returning a string.
|
|
|
|
@vindex forms-fields
|
|
The function you call can access the fields of the record as a list in
|
|
the variable
|
|
@code{forms-fields}.
|
|
|
|
@item symbol
|
|
A symbol used as a formatting element should evaluate to a string, number,
|
|
or list; the value is interpreted as a formatting element, as described
|
|
above.
|
|
@end table
|
|
|
|
If a record does not contain the number of fields as specified in
|
|
@code{forms-number-of-fields}, a warning message will be printed. Excess
|
|
fields are ignored, missing fields are set to empty.
|
|
|
|
The control file which displays @file{/etc/passwd} file as demonstrated
|
|
in the beginning of this manual might look as follows:
|
|
|
|
@example
|
|
;; @r{This demo visits @file{/etc/passwd}.}
|
|
|
|
(setq forms-file "/etc/passwd")
|
|
(setq forms-number-of-fields 7)
|
|
(setq forms-read-only t) ; @r{to make sure}
|
|
(setq forms-field-sep ":")
|
|
;; @r{Don't allow multi-line fields.}
|
|
(setq forms-multi-line nil)
|
|
|
|
(setq forms-format-list
|
|
(list
|
|
"====== /etc/passwd ======\n\n"
|
|
"User : " 1
|
|
" Uid: " 3
|
|
" Gid: " 4
|
|
"\n\n"
|
|
"Name : " 5
|
|
"\n\n"
|
|
"Home : " 6
|
|
"\n\n"
|
|
"Shell: " 7
|
|
"\n"))
|
|
@end example
|
|
|
|
When you construct the value of @code{forms-format-list}, you should
|
|
usually either quote the whole value, like this,
|
|
|
|
@example
|
|
(setq forms-format-list
|
|
'(
|
|
"====== " forms-file " ======\n\n"
|
|
"User : " 1
|
|
(make-string 20 ?-)
|
|
@dots{}
|
|
))
|
|
@end example
|
|
|
|
@noindent
|
|
or quote the elements which are lists, like this:
|
|
|
|
@example
|
|
(setq forms-format-list
|
|
(list
|
|
"====== " forms-file " ======\n\n"
|
|
"User : " 1
|
|
'(make-string 20 ?-)
|
|
@dots{}
|
|
))
|
|
@end example
|
|
|
|
Forms mode validates the contents of @code{forms-format-list} when you
|
|
visit a database. If there are errors, processing is aborted with an
|
|
error message which includes a descriptive text. @xref{Error Messages},
|
|
for a detailed list of error messages.
|
|
|
|
If no @code{forms-format-list} is specified, Forms mode will supply a
|
|
default format list. This list contains the name of the file being
|
|
visited, and a simple label for each field indicating the field number.
|
|
|
|
@node Modifying Forms Contents
|
|
@chapter Modifying The Forms Contents
|
|
|
|
If @code{forms-read-only} is @code{nil}, the user can modify the fields
|
|
and records of the database.
|
|
|
|
All normal editing commands are available for editing the contents of the
|
|
displayed record. You cannot delete or modify the fixed, explanatory
|
|
text that comes from string formatting elements, but you can modify the
|
|
actual field contents.
|
|
|
|
@ignore
|
|
@c This is for the Emacs 18 version only.
|
|
If the contents of the forms cannot be recognized properly, this is
|
|
signaled using a descriptive text. @xref{Error Messages}, for more info.
|
|
The cursor will indicate the last part of the forms which was
|
|
successfully parsed. It's important to avoid entering field contents
|
|
that would cause confusion with the field-separating fixed text.
|
|
@end ignore
|
|
|
|
If the variable @code{forms-modified-record-filter} is non-@code{nil},
|
|
it is called as a function before the new data is written to the data
|
|
file. The function receives one argument, a vector that contains the
|
|
contents of the fields of the record.
|
|
|
|
The function can refer to fields with @code{aref} and modify them with
|
|
@code{aset}. The first field has number 1 (one); thus, element 0 of the
|
|
vector is not used. The function should return the same vector it was
|
|
passed; the (possibly modified) contents of the vector determine what is
|
|
actually written in the file. Here is an example:
|
|
|
|
@example
|
|
(defun my-modified-record-filter (record)
|
|
;; @r{Modify second field.}
|
|
(aset record 2 (current-time-string))
|
|
;; @r{Return the field vector.}
|
|
record)
|
|
|
|
(setq forms-modified-record-filter 'my-modified-record-filter)
|
|
@end example
|
|
|
|
If the variable @code{forms-new-record-filter} is non-@code{nil}, its
|
|
value is a function to be called to fill in default values for the
|
|
fields of a new record. The function is passed a vector of empty
|
|
strings, one for each field; it should return the same vector, with
|
|
the desired field values stored in it. Fields are numbered starting
|
|
from 1 (one). Example:
|
|
|
|
@example
|
|
(defun my-new-record-filter (fields)
|
|
(aset fields 5 (login-name))
|
|
(aset fields 1 (current-time-string))
|
|
fields)
|
|
|
|
(setq forms-new-record-filter 'my-new-record-filter)
|
|
@end example
|
|
|
|
@node Miscellaneous
|
|
@chapter Miscellaneous
|
|
|
|
@vindex forms-version
|
|
The global variable @code{forms-version} holds the version information
|
|
of the Forms mode software.
|
|
|
|
@findex forms-enumerate
|
|
It is very convenient to use symbolic names for the fields in a record.
|
|
The function @code{forms-enumerate} provides an elegant means to define
|
|
a series of variables whose values are consecutive integers. The
|
|
function returns the highest number used, so it can be used to set
|
|
@code{forms-number-of-fields} also. For example:
|
|
|
|
@example
|
|
(setq forms-number-of-fields
|
|
(forms-enumerate
|
|
'(field1 field2 field3 @dots{})))
|
|
@end example
|
|
|
|
This sets @code{field1} to 1, @code{field2} to 2, and so on.
|
|
|
|
Care has been taken to keep the Forms mode variables buffer-local, so it
|
|
is possible to visit multiple files in Forms mode simultaneously, even
|
|
if they have different properties.
|
|
|
|
@findex forms-mode
|
|
If you have visited the control file in normal fashion with
|
|
@code{find-file} or a like command, you can switch to Forms mode with
|
|
the command @code{M-x forms-mode}. If you put @samp{-*- forms -*-} in
|
|
the first line of the control file, then visiting it enables Forms mode
|
|
automatically. But this makes it hard to edit the control file itself,
|
|
so you'd better think twice before using this.
|
|
|
|
The default format for the data file, using @code{"\t"} to separate
|
|
fields and @code{"\^k"} to separate lines within a field, matches the
|
|
file format of some popular database programs, e.g. FileMaker. So
|
|
@code{forms-mode} can decrease the need to use proprietary software.
|
|
|
|
@node Error Messages
|
|
@chapter Error Messages
|
|
|
|
This section describes all error messages which can be generated by
|
|
forms mode. Error messages that result from parsing the control file
|
|
all start with the text @samp{Forms control file error}. Messages
|
|
generated while analyzing the definition of @code{forms-format-list}
|
|
start with @samp{Forms format error}.
|
|
|
|
@table @code
|
|
@item Forms control file error: `forms-file' has not been set
|
|
The variable @code{forms-file} was not set by the control file.
|
|
|
|
@item Forms control file error: `forms-number-of-fields' has not been set
|
|
The variable @code{forms-number-of-fields} was not set by the control
|
|
file.
|
|
|
|
@item Forms control file error: `forms-number-of-fields' must be a number > 0
|
|
The variable @code{forms-number-of-fields} did not contain a positive
|
|
number.
|
|
|
|
@item Forms control file error: `forms-field-sep' is not a string
|
|
@itemx Forms control file error: `forms-multi-line' must be nil or a one-character string
|
|
The variable @code{forms-multi-line} was set to something other than
|
|
@code{nil} or a single-character string.
|
|
|
|
@item Forms control file error: `forms-multi-line' is equal to 'forms-field-sep'
|
|
The variable @code{forms-multi-line} may not be equal to
|
|
@code{forms-field-sep} for this would make it impossible to distinguish
|
|
fields and the lines in the fields.
|
|
|
|
@item Forms control file error: `forms-new-record-filter' is not a function
|
|
@itemx Forms control file error: `forms-modified-record-filter' is not a function
|
|
The variable has been set to something else than a function.
|
|
|
|
@item Forms control file error: `forms-format-list' is not a list
|
|
The variable @code{forms-format-list} was not set to a Lisp list
|
|
by the control file.
|
|
|
|
@item Forms format error: field number @var{xx} out of range 1..@var{nn}
|
|
A field number was supplied in @code{forms-format-list} with a value of
|
|
@var{xx}, which was not greater than zero and smaller than or equal to
|
|
the number of fields in the forms, @var{nn}.
|
|
|
|
@item Forms format error: @var{fun} is not a function
|
|
The first element of a list which is an element of
|
|
@code{forms-format-list} was not a valid Lisp function.
|
|
|
|
@item Forms format error: invalid element @var{xx}
|
|
A list element was supplied in @code{forms-format-list} which was not a
|
|
string, number or list.
|
|
|
|
@ignore
|
|
@c This applies to Emacs 18 only.
|
|
@c Error messages generated while a modified form is being analyzed.
|
|
|
|
@item Parse error: not looking at `...'
|
|
When re-parsing the contents of a forms, the text shown could not
|
|
be found.
|
|
|
|
@item Parse error: cannot find `...'
|
|
When re-parsing the contents of a forms, the text shown, which
|
|
separates two fields, could not be found.
|
|
|
|
@item Parse error: cannot parse adjacent fields @var{xx} and @var{yy}
|
|
Fields @var{xx} and @var{yy} were not separated by text, so could not be
|
|
parsed again.
|
|
@end ignore
|
|
|
|
@item Warning: this record has @var{xx} fields instead of @var{yy}
|
|
The number of fields in this record in the data file did not match
|
|
@code{forms-number-of-fields}. Missing fields will be made empty.
|
|
|
|
@item Multi-line fields in this record - update refused!
|
|
The current record contains newline characters, hence can not be written
|
|
back to the data file, for it would corrupt it. Probably you inserted a
|
|
newline in a field, while @code{forms-multi-line} was @code{nil}.
|
|
|
|
@item Field separator occurs in record - update refused!
|
|
The current record contains the field separator string inside one of the
|
|
fields. It can not be written back to the data file, for it would
|
|
corrupt it. Probably you inserted the field separator string in a field.
|
|
|
|
@item Record number @var{xx} out of range 1..@var{yy}
|
|
A jump was made to non-existing record @var{xx}. @var{yy} denotes the
|
|
number of records in the file.
|
|
|
|
@item Stuck at record @var{xx}
|
|
An internal error prevented a specific record from being retrieved.
|
|
|
|
@item No write access to @code{"}@var{file}@code{"}
|
|
An attempt was made to enable edit mode on a file that has been write
|
|
protected.
|
|
|
|
@item Search failed: @var{regexp}
|
|
The @var{regexp} could not be found in the data file. Forward searching
|
|
is done from the current location until the end of the file, then
|
|
retrying from the beginning of the file until the current location.
|
|
Backward searching is done from the current location until the beginning
|
|
of the file, then retrying from the end of the file until the current
|
|
location.
|
|
|
|
@item Wrapped
|
|
A search completed successfully after wrapping around.
|
|
|
|
@item Warning: number of records changed to @var{nn}
|
|
Forms mode's idea of the number of records has been adjusted to the
|
|
number of records actually present in the data file.
|
|
|
|
@item Problem saving buffers?
|
|
An error occurred while saving the data file buffer. Most likely, Emacs
|
|
did ask to confirm deleting the buffer because it had been modified, and
|
|
you said `no'.
|
|
@end table
|
|
|
|
@node Long Example
|
|
@chapter Long Example
|
|
|
|
The following example exploits most of the features of Forms mode.
|
|
This example is included in the distribution as file @file{forms-d2.el}.
|
|
|
|
@example
|
|
;; demo2 -- demo forms-mode -*- emacs-lisp -*-
|
|
|
|
;; @r{This sample forms exploit most of the features of forms mode.}
|
|
|
|
;; @r{Set the name of the data file.}
|
|
(setq forms-file "forms-d2.dat")
|
|
|
|
;; @r{Use @code{forms-enumerate} to set field names and number thereof.}
|
|
(setq forms-number-of-fields
|
|
(forms-enumerate
|
|
'(arch-newsgroup ; 1
|
|
arch-volume ; 2
|
|
arch-issue ; and ...
|
|
arch-article ; ... so
|
|
arch-shortname ; ... ... on
|
|
arch-parts
|
|
arch-from
|
|
arch-longname
|
|
arch-keywords
|
|
arch-date
|
|
arch-remarks)))
|
|
|
|
;; @r{The following functions are used by this form for layout purposes.}
|
|
;;
|
|
(defun arch-tocol (target &optional fill)
|
|
"Produces a string to skip to column TARGET.
|
|
Prepends newline if needed.
|
|
The optional FILL should be a character, used to fill to the column."
|
|
(if (null fill)
|
|
(setq fill ? ))
|
|
(if (< target (current-column))
|
|
(concat "\n" (make-string target fill))
|
|
(make-string (- target (current-column)) fill)))
|
|
;;
|
|
(defun arch-rj (target field &optional fill)
|
|
"Produces a string to skip to column TARGET\
|
|
minus the width of field FIELD.
|
|
Prepends newline if needed.
|
|
The optional FILL should be a character,
|
|
used to fill to the column."
|
|
(arch-tocol (- target (length (nth field forms-fields))) fill))
|
|
|
|
;; @r{Record filters.}
|
|
;;
|
|
(defun new-record-filter (the-record)
|
|
"Form a new record with some defaults."
|
|
(aset the-record arch-from (user-full-name))
|
|
(aset the-record arch-date (current-time-string))
|
|
the-record) ; return it
|
|
(setq forms-new-record-filter 'new-record-filter)
|
|
|
|
;; @r{The format list.}
|
|
(setq forms-format-list
|
|
(list
|
|
"====== Public Domain Software Archive ======\n\n"
|
|
arch-shortname
|
|
" - " arch-longname
|
|
"\n\n"
|
|
"Article: " arch-newsgroup
|
|
"/" arch-article
|
|
" "
|
|
'(arch-tocol 40)
|
|
"Issue: " arch-issue
|
|
" "
|
|
'(arch-rj 73 10)
|
|
"Date: " arch-date
|
|
"\n\n"
|
|
"Submitted by: " arch-from
|
|
"\n"
|
|
'(arch-tocol 79 ?-)
|
|
"\n"
|
|
"Keywords: " arch-keywords
|
|
"\n\n"
|
|
"Parts: " arch-parts
|
|
"\n\n====== Remarks ======\n\n"
|
|
arch-remarks
|
|
))
|
|
|
|
;; @r{That's all, folks!}
|
|
@end example
|
|
|
|
@node Credits
|
|
@chapter Credits
|
|
|
|
Bug fixes and other useful suggestions were supplied by
|
|
Harald Hanche-Olsen (@code{hanche@@imf.unit.no}),
|
|
@code{cwitty@@portia.stanford.edu},
|
|
Jonathan I. Kamens,
|
|
Per Cederqvist (@code{ceder@@signum.se}),
|
|
Michael Lipka (@code{lipka@@lip.hanse.de}),
|
|
Andy Piper (@code{ajp@@eng.cam.ac.uk}),
|
|
Frederic Pierresteguy (@code{F.Pierresteguy@@frcl.bull.fr}),
|
|
Ignatios Souvatzis
|
|
and Richard Stallman (@code{rms@@gnu.org}).
|
|
|
|
This documentation was slightly inspired by the documentation of ``rolo
|
|
mode'' by Paul Davis at Schlumberger Cambridge Research
|
|
(@code{davis%scrsu1%sdr.slb.com@@relay.cs.net}).
|
|
|
|
None of this would have been possible without GNU Emacs of the Free
|
|
Software Foundation. Thanks, Richard!
|
|
|
|
@node GNU Free Documentation License
|
|
@appendix GNU Free Documentation License
|
|
@include doclicense.texi
|
|
|
|
@node Index
|
|
@unnumbered Index
|
|
@printindex cp
|
|
|
|
@contents
|
|
@bye
|
|
|
|
@ignore
|
|
arch-tag: 2ac9810b-aa49-4ea6-8030-d7f1ecd467ed
|
|
@end ignore
|