emacs/man/emacs-mime.texi

\input texinfo                  @c -*-mode: texinfo; coding: latin-1 -*-

@setfilename ../info/emacs-mime
@settitle Emacs MIME Manual
@synindex fn cp
@synindex vr cp
@synindex pg cp
@dircategory Emacs
@direntry
* MIME: (emacs-mime).   Emacs MIME de/composition library.
@end direntry
@iftex
@finalout
@end iftex
@setchapternewpage odd

@ifnottex

This file documents the Emacs MIME interface functionality.

Copyright (C) 1998,99,2000, 2002 Free Software Foundation, Inc.

Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.1 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 ifnottex

@tex

@titlepage
@title Emacs MIME Manual

@author by Lars Magne Ingebrigtsen
@page

@vskip 0pt plus 1filll
Copyright @copyright{} 1998,99,2000, 2002 Free Software Foundation, Inc.

Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.1 or
any later version published by the Free Software Foundation; with the
Invariant Sections being none, 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 titlepage
@page

@end tex

@node Top
@top Emacs MIME

This manual documents the libraries used to compose and display
@sc{mime} messages.

This is not a manual meant for users; it's a manual directed at people
who want to write functions and commands that manipulate @sc{mime}
elements.

@sc{mime} is short for @dfn{Multipurpose Internet Mail Extensions}.
This standard is documented in a number of RFCs; mainly RFC2045 (Format
of Internet Message Bodies), RFC2046 (Media Types), RFC2047 (Message
Header Extensions for Non-ASCII Text), RFC2048 (Registration
Procedures), RFC2049 (Conformance Criteria and Examples).  It is highly
recommended that anyone who intends writing @sc{mime}-compliant software
read at least RFC2045 and RFC2047.

@menu
* Interface Functions::   An abstraction over the basic functions.
* Basic Functions::       Utility and basic parsing functions.
* Decoding and Viewing::  A framework for decoding and viewing.
* Composing::             MML; a language for describing MIME parts.
* Standards::             A summary of RFCs and working documents used.
* Index::                 Function and variable index.
@end menu


@node Interface Functions
@chapter Interface Functions
@cindex interface functions
@cindex mail-parse

The @code{mail-parse} library is an abstraction over the actual
low-level libraries that are described in the next chapter.

Standards change, and so programs have to change to fit in the new
mold.  For instance, RFC2045 describes a syntax for the
@code{Content-Type} header that only allows @sc{ascii} characters in the
parameter list.  RFC2231 expands on RFC2045 syntax to provide a scheme
for continuation headers and non-@sc{ascii} characters.

The traditional way to deal with this is just to update the library
functions to parse the new syntax.  However, this is sometimes the wrong
thing to do.  In some instances it may be vital to be able to understand
both the old syntax as well as the new syntax, and if there is only one
library, one must choose between the old version of the library and the
new version of the library.

The Emacs MIME library takes a different tack.  It defines a series of
low-level libraries (@file{rfc2047.el}, @file{rfc2231.el} and so on)
that parses strictly according to the corresponding standard.  However,
normal programs would not use the functions provided by these libraries
directly, but instead use the functions provided by the
@code{mail-parse} library.  The functions in this library are just
aliases to the corresponding functions in the latest low-level
libraries.  Using this scheme, programs get a consistent interface they
can use, and library developers are free to create write code that
handles new standards.

The following functions are defined by this library:

@defun mail-header-parse-content-type string
Parse @var{string}, a @code{Content-Type} header, and return a
content-type list in the following format:

@lisp
("type/subtype"
 (attribute1 . value1)
 (attribute2 . value2)
 @dots{})
@end lisp

Here's an example:

@example
(mail-header-parse-content-type
 "image/gif; name=\"b980912.gif\"")
@result{} ("image/gif" (name . "b980912.gif"))
@end example
@end defun

@defun mail-header-parse-content-disposition string
Parse @var{string}, a @code{Content-Disposition} header, and return a
content-type list in the format above.
@end defun

@defun mail-content-type-get ct attribute
@findex mail-content-type-get
Returns the value of the given @var{attribute} from the content-type
list @var{ct}.

@example
(mail-content-type-get
 '("image/gif" (name . "b980912.gif")) 'name)
@result{} "b980912.gif"
@end example
@end defun

@defun mail-header-encode-parameter param value
Takes a parameter string @samp{@var{param}=@var{value}} and returns an
encoded version of it.  This is used for parameters in headers like
@samp{Content-Type} and @samp{Content-Disposition}.
@end defun

@defun mail-header-remove-comments string
Return a comment-free version of @var{string}.

@example
(mail-header-remove-comments
 "Gnus/5.070027 (Pterodactyl Gnus v0.27) (Finnish Landrace)")
@result{} "Gnus/5.070027  "
@end example
@end defun

@defun mail-header-remove-whitespace string
Remove linear white space from @var{string}.  Space inside quoted
strings and comments is preserved.

@example
(mail-header-remove-whitespace
 "image/gif; name=\"Name with spaces\"")
@result{} "image/gif;name=\"Name with spaces\""
@end example
@end defun

@defun mail-header-get-comment string
Return the last comment in @var{string}.

@example
(mail-header-get-comment
 "Gnus/5.070027 (Pterodactyl Gnus v0.27) (Finnish Landrace)")
@result{} "Finnish Landrace"
@end example
@end defun


@defun mail-header-parse-address string
Parse an address string @var{string} and return a list containing the
mailbox and the plaintext name.

@example
(mail-header-parse-address
 "Hrvoje Niksic <hniksic@@srce.hr>")
@result{} ("hniksic@@srce.hr" . "Hrvoje Niksic")
@end example
@end defun

@defun mail-header-parse-addresses string
Parse @var{string} as a list of addresses and return a list of elements
like the one described above.

@example
(mail-header-parse-addresses
 "Hrvoje Niksic <hniksic@@srce.hr>, Steinar Bang <sb@@metis.no>")
@result{} (("hniksic@@srce.hr" . "Hrvoje Niksic")
     ("sb@@metis.no" . "Steinar Bang"))
@end example
@end defun

@defun mail-header-parse-date string
Parse a date @var{string} and return an Emacs time structure.
@end defun

@defun mail-narrow-to-head
Narrow the buffer to the header section of the buffer.  Point is placed
at the beginning of the narrowed buffer.
@end defun

@defun mail-header-narrow-to-field
Narrow the buffer to the header under point.
@end defun

@defun mail-encode-encoded-word-region start end
Encode the non-@sc{ascii} words in the region @var{start}to @var{end}.  For
instance, @samp{Na<4E>ve} is encoded as @samp{=?iso-8859-1?q?Na=EFve?=}.
@end defun

@defun mail-encode-encoded-word-buffer
Encode the non-@sc{ascii} words in the current buffer.  This function is
meant to be called with the buffer narrowed to the headers of a message.
@end defun

@defun mail-encode-encoded-word-string string
Encode the words that need encoding in @var{string}, and return the
result.

@example
(mail-encode-encoded-word-string
 "This is na<6E>ve, baby")
@result{} "This is =?iso-8859-1?q?na=EFve,?= baby"
@end example
@end defun

@defun mail-decode-encoded-word-region start end
Decode the encoded words in the region @var{start}to @var{end}.
@end defun

@defun mail-decode-encoded-word-string string
Decode the encoded words in @var{string} and return the result.

@example
(mail-decode-encoded-word-string
 "This is =?iso-8859-1?q?na=EFve,?= baby")
@result{} "This is na<6E>ve, baby"
@end example
@end defun

Currently, @code{mail-parse} is an abstraction over @code{ietf-drums},
@code{rfc2047}, @code{rfc2045} and @code{rfc2231}.  These are documented
in the subsequent sections.



@node Basic Functions
@chapter Basic Functions

This chapter describes the basic, ground-level functions for parsing and
handling.  Covered here is parsing @code{From} lines, removing comments
from header lines, decoding encoded words, parsing date headers and so
on.  High-level functionality is dealt with in the next chapter
(@pxref{Decoding and Viewing}).

@menu
* rfc2045::      Encoding @code{Content-Type} headers.
* rfc2231::      Parsing @code{Content-Type} headers.
* ietf-drums::   Handling mail headers defined by RFC822bis.
* rfc2047::      En/decoding encoded words in headers.
* time-date::    Functions for parsing dates and manipulating time.
* qp::           Quoted-Printable en/decoding.
* base64::       Base64 en/decoding.
* binhex::       Binhex decoding.
* uudecode::     Uuencode decoding.
* rfc1843::      Decoding HZ-encoded text.
* mailcap::      How parts are displayed is specified by mailcap files
@end menu


@node rfc2045
@section rfc2045

RFC2045 is the ``main'' @sc{mime} document, and as such, one would
imagine that there would be a lot to implement.  But there isn't, since
most of the implementation details are delegated to the subsequent
RFCs.

So @file{rfc2045.el} has only a single function:

@defun rfc2045-encode-string parameter value
@findex rfc2045-encode-string
Takes a @var{parameter} and a @var{value} and returns a
@samp{@var{param}=@var{value}} string.  @var{value} will be quoted if
there are non-safe characters in it.
@end defun


@node rfc2231
@section rfc2231

RFC2231 defines a syntax for the @samp{Content-Type} and
@samp{Content-Disposition} headers.  Its snappy name is @dfn{MIME
Parameter Value and Encoded Word Extensions: Character Sets, Languages,
and Continuations}.

In short, these headers look something like this:

@example
Content-Type: application/x-stuff;
 title*0*=us-ascii'en'This%20is%20even%20more%20;
 title*1*=%2A%2A%2Afun%2A%2A%2A%20;
 title*2="isn't it!"
@end example

They usually aren't this bad, though.

The following functions are defined by this library:

@defun rfc2231-parse-string string
Parse a @samp{Content-Type} header @var{string} and return a list
describing its elements.

@example
(rfc2231-parse-string
 "application/x-stuff;
 title*0*=us-ascii'en'This%20is%20even%20more%20;
 title*1*=%2A%2A%2Afun%2A%2A%2A%20;
 title*2=\"isn't it!\"")
@result{} ("application/x-stuff"
    (title . "This is even more ***fun*** isn't it!"))
@end example
@end defun

@defun rfc2231-get-value ct attribute
Takes a list @var{ct} of the format above and returns the value of the
specified @var{attribute}.
@end defun

@defun rfc2231-encode-string parameter value
Encode the string @samp{@var{parameter}=@var{value}} for inclusion in
headers likes @samp{Content-Type} and @samp{Content-Disposition}.
@end defun

@node ietf-drums
@section ietf-drums

@dfn{drums} is an IETF working group that is working on the replacement
for RFC822.

The functions provided by this library include:

@defun ietf-drums-remove-comments string
Remove the comments from @var{string} and return the result.
@end defun

@defun ietf-drums-remove-whitespace string
Remove linear white space from @var{string} and return the result.
Spaces inside quoted strings and comments are left untouched.
@end defun

@defun ietf-drums-get-comment string
Return the last most comment from @var{string}.
@end defun

@defun ietf-drums-parse-address string
Parse an address @var{string} and return a list of the mailbox and the
plain text name.
@end defun

@defun ietf-drums-parse-addresses string
Parse @var{string}, containing any number of comma-separated addresses,
and return a list of mailbox/plain text pairs.
@end defun

@defun ietf-drums-parse-date string
Parse the date @var{string} and return an Emacs time structure.
@end defun

@defun ietf-drums-narrow-to-header
Narrow the buffer to the header section of the current buffer.
@end defun


@node rfc2047
@section rfc2047

RFC2047 (Message Header Extensions for Non-ASCII Text) specifies how
non-@sc{ascii} text in headers are to be encoded.  This is actually rather
complicated, so a number of variables are necessary to tweak what this
library does.

The following variables are tweakable:

@defvar rfc2047-default-charset
Characters in this charset should not be decoded by this library.
This defaults to @samp{iso-8859-1}.
@end defvar

@defvar rfc2047-header-encoding-list
This is an alist of header / encoding-type pairs.  Its main purpose is
to prevent encoding of certain headers.
@end defvar

The keys can either be header regexps, or @code{t}.

The values can be either @code{nil}, in which case the header(s) in
question won't be encoded, or @code{mime}, which means that they will be
encoded.

@defvar rfc2047-charset-encoding-alist
RFC2047 specifies two forms of encoding---@code{Q} (a
Quoted-Printable-like encoding) and @code{B} (base64).  This alist
specifies which charset should use which encoding.
@end defvar

@defvar rfc2047-encoding-function-alist
This is an alist of encoding / function pairs.  The encodings are
@code{Q}, @code{B} and @code{nil}.
@end defvar

@defvar rfc2047-q-encoding-alist
The @code{Q} encoding isn't quite the same for all headers.  Some
headers allow a narrower range of characters, and that is what this
variable is for.  It's an alist of header regexps and allowable character
ranges.
@end defvar

@defvar rfc2047-encoded-word-regexp
When decoding words, this library looks for matches to this regexp.
@end defvar

Those were the variables, and these are the functions:

@defun rfc2047-narrow-to-field
Narrow the buffer to the header on the current line.
@end defun

@defun rfc2047-encode-message-header
Should be called narrowed to the header of a message.  Encodes according
to @code{rfc2047-header-encoding-alist}.
@end defun

@defun rfc2047-encode-region start end
Encodes all encodable words in the region @var{start} to @var{end}.
@end defun

@defun rfc2047-encode-string string
Encode @var{string} and return the result.
@end defun

@defun rfc2047-decode-region start end
Decode the encoded words in the region @var{start} to @var{end}.
@end defun

@defun rfc2047-decode-string string
Decode @var{string} and return the result.
@end defun



@node time-date
@section time-date

While not really a part of the @sc{mime} library, it is convenient to
document this library here.  It deals with parsing @samp{Date} headers
and manipulating time.  (Not by using tesseracts, though, I'm sorry to
say.)

These functions convert between five formats: a date string, an Emacs
time structure, a decoded time list, a number of seconds, and a day number.

The functions have quite self-explanatory names, so the following just
gives an overview of which functions are available.

@findex parse-time-string
@findex date-to-time
@findex time-to-seconds
@findex seconds-to-time
@findex time-to-day
@findex days-to-time
@findex time-since
@findex time-less-p
@findex subtract-time
@findex days-between
@findex date-leap-year-p
@findex time-to-day-in-year
@example
(parse-time-string "Sat Sep 12 12:21:54 1998 +0200")
@result{} (54 21 12 12 9 1998 6 nil 7200)

(date-to-time "Sat Sep 12 12:21:54 1998 +0200")
@result{} (13818 19266)

(time-to-seconds '(13818 19266))
@result{} 905595714.0

(seconds-to-time 905595714.0)
@result{} (13818 19266 0)

(time-to-day '(13818 19266))
@result{} 729644

(days-to-time 729644)
@result{} (961933 65536)

(time-since '(13818 19266))
@result{} (0 430)

(time-less-p '(13818 19266) '(13818 19145))
@result{} nil

(subtract-time '(13818 19266) '(13818 19145))
@result{} (0 121)

(days-between "Sat Sep 12 12:21:54 1998 +0200"
              "Sat Sep 07 12:21:54 1998 +0200")
@result{} 5

(date-leap-year-p 2000)
@result{} t

(time-to-day-in-year '(13818 19266))
@result{} 255
@end example

@findex safe-date-to-time
And finally, we have @code{safe-date-to-time}, which does the same as
@code{date-to-time}, but returns a zero time if the date is
syntactically malformed.



@node qp
@section qp

This library deals with decoding and encoding Quoted-Printable text.

Very briefly explained, QP encoding means translating all 8-bit
characters (and lots of control characters) into things that look like
@samp{=EF}; that is, an equal sign followed by the byte encoded as a hex
string.  It is defined in RFC 2045.

The following functions are defined by the library:

@deffn Command quoted-printable-decode-region @var{from} @var{to} &optional @var{coding-system}
QP-decode all the encoded text in the region.  If @var{coding-system}
is non-nil, decode bytes into characters with that coding-system.  It
is probably better not to use @var{coding-system}; instead decode into
a unibyte buffer, decode that appropriately and then interpret it as
multibyte.
@end deffn

@defun quoted-printable-decode-string @var{string} &optional @var{coding-system}
Return a QP-encoded copy of @var{string}.  If @var{coding-system} is
non-nil, decode bytes into characters with that coding-system.
@end defun

@deffn Command quoted-printable-encode-region @var{from} @var{to} &optional @var{fold} @var{class}
QP-encode all the region.  If @var{fold} is non-@var{nil}, fold lines
at 76 characters, as required by the RFC.  If @var{class} is
non-@code{nil}, translate the characters not matched by that regexp
class, which should be in the form expected by
@var{skip-chars-forward} and should probably not contain literal
eight-bit characters.  Specifying @var{class} makes sense to do extra
encoding in header fields.

If variable @var{mm-use-ultra-safe-encoding} is defined and
non-@code{nil}, fold lines unconditionally and encode @samp{From } and
@samp{-} at the start of lines..
@end deffn

@defun quoted-printable-encode-string string
Return a QP-encoded copy of @var{string}.
@end defun

@node base64
@section base64
@cindex base64

Base64 is an encoding that encodes three bytes into four characters,
thereby increasing the size by about 33%.  The alphabet used for
encoding is very resistant to mangling during transit.  @xref{Base
64,,Base 64 Encoding, elisp, The Emacs Lisp Reference Manual}.

@node binhex
@section binhex
@cindex binhex
@cindex Apple
@cindex Macintosh

Binhex is an encoding that originated in Macintosh environments.
The following function is supplied to deal with these:

@defun binhex-decode-region start end &optional header-only
Decode the encoded text in the region @var{start} to @var{end}.  If
@var{header-only} is non-@code{nil}, only decode the @samp{binhex}
header and return the file name.
@end defun


@node uudecode
@section uudecode
@cindex uuencode
@cindex uudecode

Uuencoding is probably still the most popular encoding of binaries
used on Usenet, although Base64 rules the mail world.

The following function is supplied by this package:

@defun uudecode-decode-region start end &optional file-name
Decode the text in the region @var{start} to @var{end}.  If
@var{file-name} is non-@code{nil}, save the result to @var{file-name}.
@end defun


@node rfc1843
@section rfc1843
@cindex rfc1843
@cindex HZ
@cindex Chinese

RFC1843 deals with mixing Chinese and @sc{ascii} characters in messages.  In
essence, RFC1843 switches between @sc{ascii} and Chinese by doing this:

@example
This sentence is in ASCII.
The next sentence is in GB.~@{<:Ky2;S@{#,NpJ)l6HK!#~@}Bye.
@end example

Simple enough, and widely used in China.

The following functions are available to handle this encoding:

@defun rfc1843-decode-region start end
Decode HZ-encoded text in the region @var{start} to @var{end}.
@end defun

@defun rfc1843-decode-string string
Decode the HZ-encoded @var{string} and return the result.
@end defun


@node mailcap
@section mailcap

As specified by RFC 1524, @sc{mime}-aware message handlers parse
@dfn{mailcap} files from a default list, which can be overridden by the
@code{MAILCAP} environment variable.  These describe how elements are
supposed to be displayed.  Here's an example file:

@example
image/*; gimp -8 %s
audio/wav; wavplayer %s
@end example

This says that all image files should be displayed with @command{gimp},
and that WAVE audio files should be played by @code{wavplayer}.

The @code{mailcap} library parses such files, and provides functions for
matching types.

@defvar mailcap-mime-data
This variable is an alist of alists containing backup viewing rules for
@sc{mime} types.  These are overridden by rules for a type found in
mailcap files.  The outer alist is keyed on the major content-type and
the inner alists are keyed on the minor content-type (which can be a
regular expression).

@c Fixme: document this properly!
For example:
@example
(("application"
  ("octet-stream"
   (viewer . mailcap-save-binary-file)
   (non-viewer . t)
   (type . "application/octet-stream"))
  ("plain"
   (viewer . view-mode)
   (test fboundp 'view-mode)
   (type . "text/plain")))
@end example
@end defvar

@defopt mailcap-default-mime-data
This variable is the default value of @code{mailcap-mime-data}.  It
exists to allow setting the value using Custom.  It is merged with
values from mailcap files by @code{mailcap-parse-mailcaps}.
@end defopt

Although it is not specified by the RFC, @sc{mime} tools normally use a
common means of associating file extensions with defualt @sc{mime} types
in the absence of other information about the type of a file.  The
information is found in per-user files @file{~/.mime.types} and system
@file{mime.types} files found in quasi-standard places.  Here is an
example:

@example
application/x-dvi	dvi
audio/mpeg		mpga mpega mp2 mp3
image/jpeg		jpeg jpg jpe
@end example


@defvar mailcap-mime-extensions
This variable is an alist @sc{mime} types keyed by file extensions.
This is overridden by entries found in @file{mime.types} files.
@end defvar

@defopt mailcap-default-mime-extensions
This variable is the default value of @code{mailcap-mime-extensions}.
It exists to allow setting the value using Custom.  It is merged with
values from mailcap files by @code{mailcap-parse-mimetypes}.
@end defopt

Interface functions:

@defun mailcap-parse-mailcaps &optional path force
Parse all the mailcap files specified in a path string @var{path} and
merge them with the values from @code{mailcap-mime-data}.  Components of
@var{path} are separated by the @code{path-separator} character
appropriate for the system.  If @var{force} is non-@code{nil}, the files
are re-parsed even if they have been parsed already.  If @var{path} is
omitted, use the value of environment variable @code{MAILCAPS} if it is
set; otherwise (on GNU and Unix) use the path defined in RFC 1524, plus
@file{/usr/local/etc/mailcap}.
@end defun

@defun mailcap-parse-mimetypes &optional path force
Parse all the mimetypes specified in a path string @var{path}
and merge them with the values from @code{mailcap-mime-extensions}.
Components of @var{path} are separated by the @code{path-separator}
character appropriate for the system.  If @var{path} is omitted, use the
value of environment variable @code{MIMETYPES} if set; otherwise use a
default path consistent with that used by @code{mailcap-parse-mailcaps}.
If @var{force} is non-@code{nil}, the files are re-parsed even if they
have been parsed already.
@end defun

@defun mailcap-mime-info string &optional request
Gets the viewer command for content-type @var{string}.  @code{nil} is
returned if none is found.  Expects @var{string} to be a complete
content-type header line.

If @var{request} is non-@code{nil} it specifies what information to
return.  If it is nil or the empty string, the viewer (second field of
the mailcap entry) will be returned.  If it is a string, then the
mailcap field corresponding to that string will be returned
(@samp{print}, @samp{description}, whatever).  If it is a number, all
the information for this viewer is returned.  If it is @code{all}, then
all possible viewers for this type is returned.
@end defun

@defun mailcap-mime-types
This function returns a list of all the defined media types.
@end defun

@defun mailcap-extension-to-mime extension
This function returns the content type defined for a file with the given
@var{extension}.
@end defun


@node Decoding and Viewing
@chapter Decoding and Viewing

This chapter deals with decoding and viewing @sc{mime} messages on a
higher level.

The main idea is to first analyze a @sc{mime} article, and then allow
other programs to do things based on the list of @dfn{handles} that are
returned as a result of this analysis.

@menu
* Dissection::     Analyzing a @sc{mime} message.
* Handles::        Handle manipulations.
* Display::        Displaying handles.
* Customization::  Variables that affect display.
* New Viewers::    How to write your own viewers.
@end menu


@node Dissection
@section Dissection

The @code{mm-dissect-buffer} is the function responsible for dissecting
a @sc{mime} article.  If given a multipart message, it will recursively
descend the message, following the structure, and return a tree of
@sc{mime} handles that describes the structure of the message.


@node Handles
@section Handles

A @sc{mime} handle is a list that fully describes a @sc{mime} component.

The following macros can be used to access elements from the
@var{handle} argument:

@defmac mm-handle-buffer handle
Return the buffer that holds the contents of the undecoded @sc{mime}
part.
@end defmac

@defmac mm-handle-type handle
Return the parsed @samp{Content-Type} of the part.
@end defmac

@defmac mm-handle-encoding handle
Return the @samp{Content-Transfer-Encoding} of the part.
@end defmac

@defmac mm-handle-undisplayer handle
Return the function that can be used to remove the displayed part (if it
has been displayed).
@end defmac

@defmac mm-handle-set-undisplayer handle function
Set the undisplayer function for the part to function.
@end defmac

@defmac mm-handle-disposition
Return the parsed @samp{Content-Disposition} of the part.
@end defmac

@defmac mm-handle-disposition
Return the description of the part.
@end defmac

@defmac mm-get-content-id id
Returns the handle(s) referred to by @var{id}, the @samp{Content-ID} of
the part.
@end defmac


@node Display
@section Display

Functions for displaying, removing and saving.  In the descriptions
below, `the part' means the @sc{mime} part represented by the
@var{handle} argument.

@defun mm-display-part handle &optional no-default
Display the part.  Return @code{nil} if the part is removed,
@code{inline} if it is displayed inline or @code{external} if it is
displayed externally.  If @var{no-default} is non-@code{nil}, the part
is not displayed unless the @sc{mime} type of @var{handle} is defined to
be displayed inline or there is an display method defined for it; i.e.@:
no default external method will be used.
@end defun

@defun mm-remove-part handle
Remove the part if it has been displayed.
@end defun

@defun mm-inlinable-p handle
Return non-@code{nil} if the part can be displayed inline.
@end defun

@defun mm-automatic-display-p handle
Return non-@code{nil} if the user has requested automatic display of the
@sc{mime} type of the part.
@end defun

@defun mm-destroy-part handle
Free all the resources used by the part.
@end defun

@defun mm-save-part handle
Save the part to a file.  The user is prompted for a file name to use.
@end defun

@defun mm-pipe-part handle
Pipe the part through a shell command.  The user is prompted for the
command to use.
@end defun

@defun mm-interactively-view-part handle
Prompt for a mailcap method to use to view the part and display it
externally using that method.
@end defun


@node Customization
@section Customization

The display of @sc{mime} types may be customized with the following
options.

@defopt mm-inline-media-tests
This is an alist where the key is a @sc{mime} type, the second element
is a function to display the part @dfn{inline} (i.e., inside Emacs), and 
the third element is a form to be @code{eval}ed to say whether the part
can be displayed inline.

This variable specifies whether a part @emph{can} be displayed inline,
and, if so, how to do it.  It does not say whether parts are
@emph{actually} displayed inline.
@end defopt

@defopt mm-inlined-types
This, on the other hand, says what types are to be displayed inline, if
they satisfy the conditions set by the variable above.  It's a list of
@sc{mime} media types.
@end defopt

@defopt mm-automatic-display
This is a list of types that are to be displayed ``automatically'', but
only if the above variable allows it.  That is, only inlinable parts can
be displayed automatically.
@end defopt

@defopt mm-attachment-override-types
Some @sc{mime} agents create parts that have a content-disposition of
@samp{attachment}.  This variable allows overriding that disposition and 
displaying the part inline.  (Note that the disposition is only
overridden if we are able to, and want to, display the part inline.)
@end defopt

@defopt mm-discouraged-alternatives
List of @sc{mime} types that are discouraged when viewing
@samp{multipart/alternative}.  Viewing agents are supposed to view the
last possible part of a message, as that is supposed to be the richest.
However, users may prefer other types instead, and this list says what
types are most unwanted.  If, for instance, @samp{text/html} parts are
very unwanted, and @samp{text/richtech} parts are somewhat unwanted,
then the value of this variable should be set to:

@lisp
("text/html" "text/richtext")
@end lisp
@end defopt

@defopt mm-inline-large-images-p
When displaying inline images that are larger than the window, XEmacs
does not enable scrolling, which means that you cannot see the whole
image.  To prevent this, the library tries to determine the image size
before displaying it inline, and if it doesn't fit the window, the
library will display it externally (e.g. with @samp{ImageMagick} or
@samp{xv}).  Setting this variable to @code{t} disables this check and
makes the library display all inline images as inline, regardless of
their size.
@end defopt

@defopt mm-inline-override-p
@code{mm-inlined-types} may include regular expressions, for example to
specify that all @samp{text/.*} parts be displayed inline.  If a user
prefers to have a type that matches such a regular expression be treated
as an attachment, that can be accomplished by setting this variable to a
list containing that type.  For example assuming @code{mm-inlined-types}
includes @samp{text/.*}, then including @samp{text/html} in this
variable will cause @samp{text/html} parts to be treated as attachments.
@end defopt


@node New Viewers
@section New Viewers

Here's an example viewer for displaying @samp{text/enriched} inline:

@lisp
(defun mm-display-enriched-inline (handle)
  (let (text)
    (with-temp-buffer
      (mm-insert-part handle)
      (save-window-excursion
        (enriched-decode (point-min) (point-max))
        (setq text (buffer-string))))
    (mm-insert-inline handle text)))
@end lisp

We see that the function takes a @sc{mime} handle as its parameter.  It
then goes to a temporary buffer, inserts the text of the part, does some 
work on the text, stores the result, goes back to the buffer it was
called from and inserts the result.

The two important helper functions here are @code{mm-insert-part} and
@code{mm-insert-inline}.  The first function inserts the text of the
handle in the current buffer.  It handles charset and/or content
transfer decoding.  The second function just inserts whatever text you
tell it to insert, but it also sets things up so that the text can be
``undisplayed' in a convenient manner.


@node Composing
@chapter Composing
@cindex Composing
@cindex MIME Composing
@cindex MML
@cindex MIME Meta Language

Creating a @sc{mime} message is boring and non-trivial.  Therefore, a
library called @code{mml} has been defined that parses a language called
MML (@sc{mime} Meta Language) and generates @sc{mime} messages.

@findex mml-generate-mime
The main interface function is @code{mml-generate-mime}.  It will
examine the contents of the current (narrowed-to) buffer and return a
string containing the @sc{mime} message.

@menu
* Simple MML Example::             An example MML document.
* MML Definition::                 All valid MML elements.
* Advanced MML Example::           Another example MML document.
* Charset Translation::            How charsets are mapped from Mule to MIME.
* Conversion::                     Going from @sc{mime} to MML and vice versa.
@end menu


@node Simple MML Example
@section Simple MML Example

Here's a simple @samp{multipart/alternative}:

@example
<#multipart type=alternative>
This is a plain text part.
<#part type=text/enriched>
<center>This is a centered enriched part</center>
<#/multipart>
@end example

After running this through @code{mml-generate-mime}, we get this:

@example
Content-Type: multipart/alternative; boundary="=-=-="


--=-=-=


This is a plain text part.

--=-=-=
Content-Type: text/enriched


<center>This is a centered enriched part</center>

--=-=-=--
@end example


@node MML Definition
@section MML Definition

The MML language is very simple.  It looks a bit like an SGML
application, but it's not.

The main concept of MML is the @dfn{part}.  Each part can be of a
different type or use a different charset.  The way to delineate a part
is with a @samp{<#part ...>} tag.  Multipart parts can be introduced
with the @samp{<#multipart ...>} tag.  Parts are ended by the
@samp{<#/part>} or @samp{<#/multipart>} tags.  Parts started with the
@samp{<#part ...>} tags are also closed by the next open tag.

There's also the @samp{<#external ...>} tag.  These introduce
@samp{external/message-body} parts.

Each tag can contain zero or more parameters on the form
@samp{parameter=value}.  The values may be enclosed in quotation marks,
but that's not necessary unless the value contains white space.  So
@samp{filename=/home/user/#hello$^yes} is perfectly valid.

The following parameters have meaning in MML; parameters that have no
meaning are ignored.  The MML parameter names are the same as the
@sc{mime} parameter names; the things in the parentheses say which
header it will be used in.

@table @samp
@item type
The @sc{mime} type of the part (@samp{Content-Type}).

@item filename
Use the contents of the file in the body of the part
(@samp{Content-Disposition}).

@item charset
The contents of the body of the part are to be encoded in the character
set specified (@samp{Content-Type}).

@item name
Might be used to suggest a file name if the part is to be saved
to a file (@samp{Content-Type}).

@item disposition
Valid values are @samp{inline} and @samp{attachment}
(@samp{Content-Disposition}).

@item encoding
Valid values are @samp{7bit}, @samp{8bit}, @samp{quoted-printable} and
@samp{base64} (@samp{Content-Transfer-Encoding}).

@item description
A description of the part (@samp{Content-Description}).

@item creation-date
RFC822 date when the part was created (@samp{Content-Disposition}).

@item modification-date
RFC822 date when the part was modified (@samp{Content-Disposition}).

@item read-date
RFC822 date when the part was read (@samp{Content-Disposition}).

@item size
The size (in octets) of the part (@samp{Content-Disposition}).

@end table

Parameters for @samp{application/octet-stream}:

@table @samp
@item type
Type of the part; informal---meant for human readers
(@samp{Content-Type}).
@end table

Parameters for @samp{message/external-body}:

@table @samp
@item access-type
A word indicating the supported access mechanism by which the file may
be obtained.  Values include @samp{ftp}, @samp{anon-ftp}, @samp{tftp},
@samp{localfile}, and @samp{mailserver}.  (@samp{Content-Type}.)

@item expiration
The RFC822 date after which the file may no longer be fetched.
(@samp{Content-Type}.)

@item size
The size (in octets) of the file.  (@samp{Content-Type}.)

@item permission
Valid values are @samp{read} and @samp{read-write}
(@samp{Content-Type}).

@end table


@node Advanced MML Example
@section Advanced MML Example

Here's a complex multipart message.  It's a @samp{multipart/mixed} that
contains many parts, one of which is a @samp{multipart/alternative}.

@example
<#multipart type=mixed>
<#part type=image/jpeg filename=~/rms.jpg disposition=inline>
<#multipart type=alternative>
This is a plain text part.
<#part type=text/enriched name=enriched.txt>
<center>This is a centered enriched part</center>
<#/multipart>
This is a new plain text part.
<#part disposition=attachment>
This plain text part is an attachment.
<#/multipart>
@end example

And this is the resulting @sc{mime} message:

@example
Content-Type: multipart/mixed; boundary="=-=-="


--=-=-=



--=-=-=
Content-Type: image/jpeg;
 filename="~/rms.jpg"
Content-Disposition: inline;
 filename="~/rms.jpg"
Content-Transfer-Encoding: base64
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--=-=-=
Content-Type: multipart/alternative; boundary="==-=-="


--==-=-=


This is a plain text part.

--==-=-=
Content-Type: text/enriched;
 name="enriched.txt"


<center>This is a centered enriched part</center>

--==-=-=--

--=-=-=

This is a new plain text part.

--=-=-=
Content-Disposition: attachment


This plain text part is an attachment.

--=-=-=--
@end example

@node Charset Translation
@section Charset Translation
@cindex charsets

During translation from MML to @sc{mime}, for each @sc{mime} part which
has been composed inside Emacs, an appropriate @sc{mime} charset has to
be chosen.

@vindex mail-parse-charset
@cindex unibyte Emacs
If you are running a non-Mule XEmacs, or Emacs in unibyte
mode@footnote{Deprecated!}, this process is simple: if the part
contains any non-@sc{ascii} (8-bit) characters, the @sc{mime} charset
given by @code{mail-parse-charset} (a symbol) is used.  (Never set this
variable directly, though.  If you want to change the default charset,
please consult the documentation of the package which you use to process
@sc{mime} messages.  @xref{Various Message Variables, , Various Message
Variables, message, Message Manual}, for example.)  If there are only
@sc{ascii} characters, the @sc{mime} charset @samp{US-ASCII} is used, of
course.

@cindex multibyte Emacs
@cindex @code{mime-charset} property
In a normal (multibyte) Emacs session, a list of coding systems is
derived that can encode the message part's content and correspond to
MIME charsets (according to their @code{mime-charset} property).  This
list is according to the normal priority rules and the highest priority
one is chosen to encode the part.  If no such coding system can encode
the part's contents, they are split into several parts such that each
can be encoded with an appropriate coding system/@sc{mime}
charset.@footnote{The part can only be split at line boundaries,
though---if more than one @sc{mime} charset is required to encode a
single line, it is not possible to encode the part.}  Note that this
procedure works with any correctly-defined coding systems, not just
built-in ones.  Given a suitably-defined UTF-8 coding system---one
capable of encoding the Emacs charsets you use---it is not normally
necessary to split a part by charset.

@vindex mm-mime-mule-charset-alist
@cindex XEmacs/Mule
It isn't possible to do this properly in XEmacs/Mule.  Instead, a list
of the Mule charsets used in the part is obtained, and the
corresponding @sc{mime} charsets are determined by lookup in
@code{mm-mime-mule-charset-alist}.  If the list elements all
correspond to a single @sc{mime} charset, that is used to encode the
part.  Otherwise, the part is split as above.

@node Conversion
@section Conversion

@findex mime-to-mml
A (multipart) @sc{mime} message can be converted to MML with the
@code{mime-to-mml} function.  It works on the message in the current
buffer, and substitutes MML markup for @sc{mime} boundaries.
Non-textual parts do not have their contents in the buffer, but instead
have the contents in separate buffers that are referred to from the MML
tags.

@findex mml-to-mime
An MML message can be converted back to @sc{mime} by the
@code{mml-to-mime} function.

These functions are in certain senses ``lossy''---you will not get back
an identical message if you run @sc{mime-to-mml} and then
@sc{mml-to-mime}.  Not only will trivial things like the order of the
headers differ, but the contents of the headers may also be different.
For instance, the original message may use base64 encoding on text,
while @sc{mml-to-mime} may decide to use quoted-printable encoding, and
so on.

In essence, however, these two functions should be the inverse of each
other.  The resulting contents of the message should remain equivalent,
if not identical.


@node Standards
@chapter Standards

The Emacs @sc{mime} library implements handling of various elements
according to a (somewhat) large number of RFCs, drafts and standards
documents.  This chapter lists the relevant ones.  They can all be
fetched from @samp{http://quimby.gnus.org/notes/}.

@table @dfn
@item RFC822
@itemx STD11
Standard for the Format of ARPA Internet Text Messages.

@item RFC1036
Standard for Interchange of USENET Messages

@item RFC1524
A User Agent Configuration Mechanism For Multimedia Mail Format
Information

@item RFC2045
Format of Internet Message Bodies

@item RFC2046
Media Types

@item RFC2047
Message Header Extensions for Non-ASCII Text

@item RFC2048
Registration Procedures

@item RFC2049
Conformance Criteria and Examples

@item RFC2231
MIME Parameter Value and Encoded Word Extensions: Character Sets,
Languages, and Continuations

@item RFC1843
HZ - A Data Format for Exchanging Files of Arbitrarily Mixed Chinese and
ASCII characters

@item draft-ietf-drums-msg-fmt-05.txt
Draft for the successor of RFC822

@item RFC2112
The MIME Multipart/Related Content-type

@item RFC1892
The Multipart/Report Content Type for the Reporting of Mail System
Administrative Messages

@item RFC2183
Communicating Presentation Information in Internet Messages: The
Content-Disposition Header Field

@end table


@node Index
@chapter Index
@printindex cp
@printindex fn

@summarycontents
@contents
@bye

@c End: