Sophie

Sophie

distrib > Mandriva > 2010.1 > x86_64 > media > main-release > by-pkgid > fb4b31bc010bcc6f53ea6f824c60d7fd > files > 34

mutt-1.5.20-8mdv2010.1.x86_64.rpm

Required tools
--------------

If you are planning to hack on mutt, please subscribe to the
mutt-dev mailing list (mutt-dev@mutt.org, contact
majordomo@mutt.org).  Announcements about recent development
versions go to that mailing list, as go technical discussions and
patches.

Patches should, if possible, be made using Mercurial against
the latest revision.

You'll need several GNU development utilities for working on mutt:

- autoconf (versions less than 2.59 are unsupported)
  (this package includes autoheader and autoreconf)

  If the build fails during any of the auto* stages, first of all try if
  re-running the ./prepare script fixes things.  Remember to give the
  same options you passed to it or to the configure it generated the
  last time, you can query them with:
    ./config.status --version

- automake (versions less than 1.9 are not officially supported)
  (this package includes aclocal)

  Note that you MUST re-run ./prepare (with the original arguments)
  if you change the automake version between builds for the same source
  directory.

- GNU make may be needed for the dependency tricks

- The internationalization (i18n) stuff requires GNU gettext.
  See intl/VERSION for the version we are currently relying on.
  Please note that using gettext-0.10 will most probably not work -
  get the latest test release from alpha.gnu.org, it's the recommended
  version of gettext anyway.

  If you are experiencing problems with unknown "dcgettext" symbols,
  the autoconf/automake macros from your gettext package are broken.
  Apply the following patch to that macro file (usually found under
  /usr/share/aclocal/gettext.m4):

--- gettext.m4.bak	Thu Jul  2 18:46:08 1998
+++ gettext.m4	Mon Oct  5 23:32:54 1998
@@ -46,12 +46,13 @@
 
 	   if test "$gt_cv_func_gettext_libc" != "yes"; then
 	     AC_CHECK_LIB(intl, bindtextdomain,
-	       [AC_CACHE_CHECK([for gettext in libintl],
-		 gt_cv_func_gettext_libintl,
-		 [AC_CHECK_LIB(intl, gettext,
-		  gt_cv_func_gettext_libintl=yes,
-		  gt_cv_func_gettext_libintl=no)],
+	       [AC_CHECK_LIB(intl, gettext,
+	         gt_cv_func_gettext_libintl=yes,
 		 gt_cv_func_gettext_libintl=no)])
+	   fi
+
+	   if test "$gt_cv_func_gettext_libintl" = "yes" ; then
+	     LIBS="-lintl $LIBS"
 	   fi
 
 	   if test "$gt_cv_func_gettext_libc" = "yes" \


Generating Mutt Documentation From Source
-----------------------------------------

To translate Mutt's Docbook XML documentation into HTML (and then text),
you'll need one tool and two sets of data which you may need to download
and install.  The tool is xsltproc (part of the libxslt package), and
it's a command-line program for performing XSL transformations on XML
documents.  The data sets are the Docbook XML and Docbook XSL libraries.

Whenever your operating system provides packages or pkgsrc or ports of
these, you should install them. Some systems, for instance SUSE Linux
and FreeBSD's ports system, automatically set up a registry of installed
XML/XSL and SGML catalogs so that the user does not need to care about
what to install where, how to set environment variables, and so on.

If your system does not provide these libraries and data sets,
you can download them from:

  . xsltproc
    http://xmlsoft.org/
    ftp://xmlsoft.org/libxslt/libxslt-1.1.17.tar.gz

  . docbook-xml-4.2
    http://www.docbook.org/
    http://www.docbook.org/xml/4.2/docbook-xml-4.2.zip

  . docbook-xsl-1.70.1
    http://docbook.sourceforge.net/
    http://prdownloads.sourceforge.net/docbook/docbook-xsl-1.70.1.zip

First, if you don't already have xsltproc, build and install libxslt,
which will provide xsltproc, too.

Next, obtain and unpack the two docbook archives.  You can unpack these
anywhere that you want to have them installed -- there's no installation
procedure other than unarchival.  On my Solaris system, I install
packages under /opt/pkgs/packagename-version, so I unpacked these ZIP
archives to /opt/pkgs/docbook-xml-4.2 and /opt/pkgs/docbook-xsl-1.70.1.

Now you need to create (and export) an environment variable to process
the manuals.  The environment variable will contain a space-separated
list of "catalog" files for the two docbook archives, so substitute
the path where you unpacked them below:

    sh$ XML_CATALOG_FILES="/path/to/docbook-xml-4.2/catalog.xml /path/to/docbook-xsl-1.70.1/catalog.xml"; export XML_CATALOG_FILES
or
    csh$ setenv XML_CATALOG_FILES "/path/to/docbook-xml-4.2/catalog.xml /path/to/docbook-xsl-1.70.1/catalog.xml"

Once all these are installed and XML_CATALOG_FILES is set, you should be
able to generate manual.html with a simple "make" -- all as a part of
the mutt compilation.

The Makefile depends upon lynx (or any other text-mode web browser)
to turn the HTML into text, so if that fails you may need to install
something else.


Getting started from Mercurial
------------------------------

The official Mercurial repository is located at:
http://dev.mutt.org/hg/mutt/. You can get a fresh clone via:

  $ hg clone http://dev.mutt.org/hg/mutt/ mutt

As a result of CVS-to-Mercurial conversion, you need to do:

  $ hg update -C HEAD

in the cloned directory.

Once you've checked out a copy of the source, or changed your
automake version, you'll need to run the script called './prepare' that
is in the root directory.  The script does all the automake/autoconf
magic that needs to be done with a fresh checkout.


Contributing patches
--------------------

As Mercurial is a distributed version control system, it's easy to
commit changes locally without impacting anybody else's work, starting
over again, or turn several commit and backouts into a new single patch
ready for submission.

These so-called "changesets" (a diff with a reasonable message
describing the change) can be exported using Mercurial through the
"patchbomb" extension shipped with Mercurial (please see the hg
documentation for details) which also is the preferred format for
submission to the mutt-dev mailing list for discussion and review.

In order to ease later bisecting in case of bugs and code history,
changes should be grouped logically, feature by feature or bugfix by
bugfix. Escpecially a single patch fixing several problems at once
should be avoided.


A word about warnings
---------------------

Mutt's default build process sets some pretty restrictive compiler
flags which may lead to lots of warnings.  Generally, warnings are
something which should be eliminated.

Nevertheless, the code in intl/ is said to generate some warnings with
the compiler settings we usually rely upon.  This code is not
maintained by the mutt developers, so please redirect any comments to
the GNU gettext library's developers.


Style Guide
-----------

- global functions should have the prefix "mutt_".  All
  other functions should be declared "static".

- avoid global variables where possible.  If one is required,
  try to contain it to a single source file and declare it
  "static".  Global variables should have the first letter of
  each word capitalized, and no underscores should be used
  (e.g., MailGid, LastFolder, MailDir).

- re-use code as much as possible.  There are a lot of
  "library" functions. One of the biggest causes of bloat
  in ELM and PINE is the tremendous duplication of code...
  Help keep Mutt small!

- when adding new options, make the old behavior the
  default.

- try to keep mutt as portable as possible.

- special characters should be in utf-8.  If you find remnants
  from the times when this was an iso-8859-1 source code tree,
  please feel free to fix them.

Documentation
-------------

Please document your changes.  Note that there are several places
where you may have to add documentation:

- doc/manual.xml.{head,tail} contain The Manual.

- doc/muttrc.man.{head,tail} contain an abridged version of The
  Manual in nroff format (see man(7)), which deals with
  configuration file commands.

- UPDATING includes short documentation of user-visible
  changes, i.e., any incompatibilities should go here.

Configuration _variables_ are documented directly in init.h.   Note
that this includes documentation for possibly added format flags!

The parts of The Manual and the muttrc manual page dealing with
these variables, and the global Muttrc, are generated automatically
from that documentation.  To start this process, type "make
update-doc" in the top-level source directory.

Note that you may have to update the makedoc utility (makedoc.c)
when adding new data types to init.h.

More precisely, variable name, type, and default value are directly
extracted from the initializer for the MuttVars array. Documentation
is expected in special comments which _follow_ the initializer.
For a line to be included with the documentation, it must (after,
possibly, some white space) begin with either "/**" or "**".
Any following white space is ignored. The rest of the line is
expected to be plain text, with some formatting instructions roughly
similar to [ntg]roff:

 - \fI switches to italics

 - \fB switches to boldface

 - \fT switches to monospace

 - \fP switches to normal display after \fI, \fB or \fT

 - \(as can be used to represent an asterisk (*).  This is intended
   to help avoiding character sequences such as /* or */ inside
   comments.

 - \(rs can be used to represent a backslash (\).  This is intended
   to help avoiding problems when trying to represent any of the \
   sequences used by makedoc.

 - .dl on a line starts a "definition list" environment (name taken
    from HTML) where terms and definitions alternate.

 - .dt marks a term in a definition list.

 - .dd marks a definition in a definition list.

 - .de on a line finishes a definition list environment.

 - .ts on a line starts a "verbose tscreen" environment (name taken from
   SGML).  Please try to keep lines inside such an environment
   short; a length of about 40 characters should be OK.  This is
   necessary to avoid a really bad-looking muttrc (5) manual page.

 - .te on a line finishes this environment.

 - .pp on a line starts a paragraph.

 - $word will be converted to a reference to word, where appropriate.
   Note that $$word is possible as well.
   Use $$$ to get a literal $ without making a reference.

 - '. ' in the beginning of a line expands to two space characters.
   This is used to protect indentations in tables.

Do _not_ use any other SGML or nroff formatting instructions here!