Initial Commit

[packages] / xemacs-packages / xemacs-devel / texi / patcher.texi

\input texinfo

@c patcher.texi --- Patcher documentation

@c Copyright (C) 2002, 2003, 2004, 2007, 2008, 2009, 2010 Didier Verna.

@c Author:        Didier Verna <didier@xemacs.org>
@c Maintainer:    Didier Verna <didier@xemacs.org>
@c Created:       Sun Apr 21 21:34:06 2002
@c Last Revision: Wed Feb 10 18:00:51 2010

@c This file is part of Patcher.

@c Patcher is free software; you can redistribute it and/or modify
@c it under the terms of the GNU General Public License version 2,
@c as published by the Free Software Foundation.

@c Patcher is distributed in the hope that it will be useful,
@c but WITHOUT ANY WARRANTY; without even the implied warranty of
@c MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
@c GNU General Public License for more details.

@c You should have received a copy of the GNU General Public License
@c along with this program; if not, write to the Free Software
@c Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.


@c Commentary:

@c Contents management by FCM version 0.1.


@c %** start of header
@setfilename patcher.info
@settitle Patcher
@setchapternewpage odd
@setcontentsaftertitlepage
@dircategory Emacs
@direntry
* Patcher: (patcher). Automatic archive-based projects maintenance.
@end direntry
@c %** end of header


@c ====================================================================
@c Definitions
@c ====================================================================
@set VERSION 3.11
@set COPYRIGHT_DATE 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2007, 2008, 2009, 2010


@c ====================================================================
@c Copyright
@c ====================================================================
@ifinfo
This file contains the documentation for Patcher version
@value{VERSION}, an XEmacs package for automating the maintenance of
archive-based projects.

Copyright @copyright{} @value{COPYRIGHT_DATE} Didier Verna.

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 a 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 also that the
sections entitled ``Copying'' and ``GNU General Public License'' are
included exactly as in the original, and 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 Free Software Foundation.
@end ifinfo



@c ====================================================================
@c Title Page
@c ====================================================================
@titlepage
@title Patcher
@subtitle Automatic archive-based projects maintenance from within XEmacs
@subtitle Version @value{VERSION}
@author Didier Verna <@email{didier@@xemacs.org}>
@page
@vskip 0pt plus 1filll
Copyright @copyright{} @value{COPYRIGHT_DATE} Didier Verna.


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 also that the
sections entitled ``Copying'' and ``GNU General Public License'' are
included exactly as in the original, and 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 Free Software Foundation.
@end titlepage



@c ====================================================================
@c Master menu
@c ====================================================================
@ifnottex
@node Top, Copying, , (dir)
@comment  node-name,  next,  previous,  up
@top Patcher

Patcher is an XEmacs package designed to automate and ease the
maintenance of archive-based projects. It provides assistance in
building, reporting and committing patches, as well as in handling the
corresponding ChangeLog entries, for example by creating skeletons.

This info file documents Patcher version @value{VERSION}.

@menu
* Copying::          The GNU General Public License
* Introduction::     What Patcher is all about
* Installation::     How to get and install Patcher
* Quick Start::      For the brave and the impatient
* User Manual::      A step-by-step guide to using Patcher
* XEmacs Devel::     How to setup Patcher for XEmacs Development
* Variables Index::
* Functions Index::
* Keystrokes Index::
@end menu

Patcher is developed by Didier Verna @email{didier@@xemacs.org}.

@end ifnottex



@c ====================================================================
@c Copying
@c ====================================================================
@node Copying, Introduction, Top, Top
@comment  node-name,  next,  previous,  up
@unnumbered Copying

Patcher is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License version 2,
as published by the Software Foundation.

Patcher 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., 675 Mass Ave, Cambridge, MA 02139, USA.



@c ====================================================================
@c Introduction
@c ====================================================================
@node Introduction, Installation, Copying, Top
@comment  node-name,  next,  previous,  up
@chapter Introduction

When a project becomes important in size, or when the development is
performed cooperatively by several people across the Internet, it is a
common practice to help maintaining it by using a development control
system. Such tools (Git, Mercurial, Subversion, Darcs, CVS, PRCS to name
a few) usually work by maintaining a centralized or distributed project
archive (also called a repository) that keeps track of the history of
the changes, lets you develop different ``branches'' at the same time
and perform operations like merges between these different project
branches.

In such ``archive-based'' maintenance models, making the project evolve
usually involves repeatedly the same few steps, some of which can be
tedious: you work on your local copy of the project; once you're
satisfied with your changes, you create a patch by diffing your local
copy against the project's archive; then (or progressively), you
construct the ChangeLog entries. Finally, you propose your changes by
sending a mail to the developers list with your patch and the ChangeLog
entries included, hoping that your proposition will be accepted. If
you're one of the maintainers, you will still probably send the message
to the list, simply announcing the modification, and immediately commit
the patch with an informative log message.

Patcher is an XEmacs package designed to automate this process. Patcher
can't work on the project for you. However, as soon as you give it some
basic knowledge on the project structure and repository, it can
automatically build a patch by comparing your local copy with the
repository, create ChangeLog entries, prepare a mail announcing the
changes, and even commit the patch for you with a informative log
message. All of this is done in just a few keystrokes. Additionally,
Patcher can perform some sanity checks, like verifying that your local
copy is up-to-date, that you did not forget some ChangeLog entries, that
the commit operation went well and so on.

If you're brave and impatient, and want to start using the basics of
Patcher as soon as possible, see @ref{Quick Start}. It is recommended to
read it anyway, since it gives an overview of how Patcher works. If you
know the basics and want a more detailed guide, see @ref{User Manual}.

Enjoy using Patcher !



@c ====================================================================
@c Installation
@c ====================================================================
@node Installation, Quick Start, Introduction, Top
@comment  node-name,  next,  previous,  up
@chapter Installation

@menu
* Distribution::                How to get Patcher
* Requirements::                What you need to get Patcher running
* Insinuation::                 How to plug Patcher into other libraries
@end menu


@c Distribution =======================================================
@node Distribution, Requirements, , Installation
@comment  node-name,  next,  previous,  up
@section Distribution

Patcher is available either as a standalone package, or as part of the
standard @file{xemacs-devel} XEmacs package.

You can get @file{xemacs-devel} in three different ways:
@itemize @bullet
@item
You can install @file{xemacs-devel} from within XEmacs via the package
user interface (accessible from the @samp{Tools} menu or via @kbd{M-x
list-packages}).

@item
You can also download the @file{xemacs-devel-<version>-pkg.tar.gz}
tarball from @uref{ftp://ftp.xemacs.org/pub/xemacs/packages/}, and
unpack it where you put the other packages.

@item
@file{xemacs-devel} sources are available from the XEmacs CVS
repository. The module name is @samp{xemacs-devel}. See
@uref{http://www.xemacs.org/Develop/cvsaccess.html} for more information
about the XEmacs CVS repository.
@end itemize


I also provide Patcher as a standalone package. This is mainly for my
personal usage (when developing and debugging new versions) but you
might want to use it if you're not interested in the rest of
@file{xemacs-devel}. The standalone version of Patcher can be found at
@uref{http://www.lrde.epita.fr/~didier/software/xemacs.php}.
You will also find different inlined versions of this documentation at
that place. For installation instructions, please read the
@file{INSTALL} file.


@c Requirements =======================================================
@node Requirements, Insinuation, Distribution, Installation
@comment  node-name,  next,  previous,  up
@section Requirements

Patcher currently works only with XEmacs 21.4 or later. I'm not sure it
works with earlier versions of XEmacs, but I'm sure it does @strong{not}
work with GNU Emacs (see below).

Patcher might also have some other requirements, depending on how you use
it:

@itemize @bullet
@item
If you let Patcher create ChangeLogs for you (@pxref{ChangeLogs
Handling}), you will need the @file{add-log} library from the
@file{xemacs-base} package, version 2.21 or later, installed on your
system. That's actually the may reason for Patcher not being able to run
in GNU Emacs, though I have not tried to port it yet.

@item
If you want to send mails from Patcher (@pxref{Mail Methods}), you will
need a mail user agent. Patcher currently supports @file{sendmail},
@file{message} and @file{Gnus} natively and through the
@file{compose-mail} interface. Other MUA might be partly supported when
used with @code{compose-mail}. Patcher will probably suffer from non
critical deficiencies in that case however (it will issue warnings).
@end itemize


@c Insinuation ========================================================
@node Insinuation, , Requirements, Installation
@comment  node-name,  next,  previous,  up
@section Insinuation

With a proper installation of Patcher (either way), you don't need any
special trickery in your @file{.emacs} file because all entry points to
the library should be autoloaded.

However, Patcher has the ability to hook into external libraries, but
won't do so unless requested. Currently, Patcher has hooks for Gnus
only. If you're using Gnus as your MUA, you might want to add the
following line to your @file{gnusrc} file:

@example
(patcher-insinuate-gnus)
@end example

This will add some facilities described along the text.



@c ====================================================================
@c Quick Start
@c ====================================================================
@node Quick Start, User Manual, Installation, Top
@comment  node-name,  next,  previous,  up
@chapter Quick Start

In this chapter, we quickly setup the basic Patcher configuration for
hacking on XEmacs. Adapt the example as you wish. Let's make some
assumptions first:

@itemize @bullet
@item
You own a computer.

@item
You have the @file{add-log} library from the @file{xemacs-base} package,
version 1.48 or later, installed on your system.

@item
You're using XEmacs from the CVS repository.

@item
Your local copy of the repository is located in
@file{/usr/local/src/XEmacs-21.5}.

@item
Your local copy of the repository is up-to-date, but you've done some
hacking in the sources that you'd like to submit.

@item
Since you're lazy, you didn't write the ChangeLog entries yet.
@end itemize


@menu
* Setting up Patcher::     Making Patcher aware of your project
* Calling Patcher::        Preparing a patch and a message
* Filling the ChangeLogs:: Patcher only creates skeletons
* Filling the Message::    You should insert the ChangeLog entries
* Committing the Patch::   Applying your modifications to the archive
* Sending the Message::    Telling people about your modifications
@end menu


@c The Project Descriptor =============================================
@node Setting up Patcher, Calling Patcher, Quick Start, Quick Start
@comment  node-name,  next,  previous,  up
@section Setting up Patcher

The first thing to do is to make patcher aware of your ``XEmacs''
project. Put this in your @file{.emacs} file:

@lisp
(setq patcher-projects
      '(("XEmacs" "/usr/local/src/XEmacs-21.5"
         :to-address "xemacs-patches@@xemacs.org")
        ))
@end lisp

@vindex patcher-projects
As you can imagine, @code{patcher-projects} is a user option in which
you store information about the projects you want to manage with
Patcher. It is actually a list of what's called @dfn{project
descriptors}. Here's the meaning of the only project descriptor we have
in the example above: we have a project named ``XEmacs'', located in
@file{/usr/local/src/XEmacs-21.5} and for which emails should be sent to
@email{xemacs-patches@@xemacs.org}.

Note the particular syntax for specifying the mailing address. This is
what's called a @dfn{project option}. Contrary to the project's name and
directory, which are mandatory and always appear as the first and second
elements of a project descriptor, project options are optional and can
appear in any order.


@c Calling Patcher ====================================================
@node Calling Patcher, Filling the ChangeLogs, Setting up Patcher, Quick Start
@comment  node-name,  next,  previous,  up
@section Calling Patcher

Now you want to build a patch with your changes, and prepare a message
to submit them. The way Patcher works is currently to setup the message
first, and then to control all subsequent operations from there. In
other words, to create a patch, you actually ask Patcher to prepare a
mail. Type this:

@lisp
M-x patcher-mail
@end lisp

@findex patcher-mail
First, you're prompted (with completion) for a project name (the first
element of each project descriptor, remember ?). We currently only have
an ``XEmacs'' project, so hitting @kbd{TAB} will directly fill the
minibuffer in with this only choice. Then, you're prompted for a subject
line that will be used in the mail. Say something sensible.

Three operations are now executed in turn:

@enumerate
@item
@vindex patcher-default-to-address
@vindex :to-address
Patcher prepares a mail buffer. The message will be sent to the address
you specified with the @code{:to-address} project option, and the
subject line now reads ``[PATCH] something sensible''.

@item
Patcher now builds the patch. The command used to do this is @samp{cvs
-q diff -u} (this is also a project option). Upon successful completion
of this command (we assume that's indeed the case), the patch is
inserted into the mail buffer. Some information about the patch is
provided just above it (the command used, the files affected and so on).

@item
Finally, Patcher generates ChangeLog skeletons from what it understands
of the patch. This involves visiting the appropriate ChangeLog files,
and creating initial entries.
@end enumerate


@c Filling the ChangeLogs =============================================
@node Filling the ChangeLogs, Filling the Message, Calling Patcher, Quick Start
@comment  node-name,  next,  previous,  up
@section Filling the ChangeLogs

Patcher has just created initial ChangeLog entries for you. You must now
browse through the ChangeLog file(s) and fill the entries as you see
fit. Once you're done, you can very well save the ChangeLog buffers.
However, don't kill them ! Don't even think about it. Patcher still
needs them. Now, please return to the mail buffer.


@c Inserting the ChangeLog entries ====================================
@node Filling the Message, Committing the Patch, Filling the ChangeLogs, Quick Start
@comment  node-name,  next,  previous,  up
@section Filling the message

Now that you're satisfied with your ChangeLog entries and you've
returned to the mail buffer, you want to write some explanation text in
the message. I'll let you do that. You also want to insert the ChangeLog
entries corresponding to your patch, since they are usually much more
readable than the patch itself.

@findex patcher-insert-change-logs
@kindex C-c C-p i
Inserting your ChangeLog entries in the mail buffer is as simple as
typing @samp{C-c C-p i}. This actually calls the function
@code{patcher-insert-change-logs}, which places them just above the
patch, with a short information line (per ChangeLog file) on top.


@c Committing the Patch ===============================================
@node Committing the Patch, Sending the Message, Filling the Message, Quick Start
@comment  node-name,  next,  previous,  up
@section Committing the Patch

If you have commit access to your project, you should read this.
Otherwise, you may directly see @ref{Sending the Message}.

Committing your changes involves three steps: preparing the commit
command, preparing the commit log message, and actually committing the
changes. Although Patcher can do all of this in one shot, it lets you
control each step by default.

@findex patcher-commit-change
@kindex C-c C-p c
In order to start the commit process, simply type @samp{C-c C-p c}. This
calls the function @code{patcher-commit-change}. Congratulations. You've
just been transported to a new buffer, the @code{*Patcher Log Message*}
buffer. This buffer lets you edit the log message that will accompany
your commit. Note that the message is initialized with the subject line
of your mail. This is also a project option.

@findex patcher-logmsg-commit
@kindex C-c C-c
Once you're satisfied with the log message, type @samp{C-c C-c}. This
calls the function @code{patcher-logmsg-commit}, which computes the
commit command to use. The command is displayed in a special buffer, the
@code{*Patcher Commit Command*} buffer. The default commit command is
@samp{cvs commit}, which of course is also a project option. Note that
Patcher stores the log message in a temporary file and uses the
@option{-F} @kbd{cvs commit} option. Finally, note that Patcher has
automatically appended the affected ChangeLog files to the commit
command.

At this point, you have the opportunity to abort the commit operation.
Otherwise, answer positively and the commit will be executed. If you had
not previously saved the ChangeLog files, Patcher will do it for you
just before committing.


@c Sending the Message ================================================
@node Sending the Message, , Committing the Patch, Quick Start
@comment  node-name,  next,  previous,  up
@section Sending the Message

Sending the message has actually nothing to do with Patcher. It depends
on the method you use for sending mails, but will usually be done via a
@samp{C-c C-c} command of some sort. On thing to note however: if you've
committed your changes via Patcher, the message has been slightly
modified: the subject line now reads ``[COMMIT] something sensible''
instead of ``[PATCH] @dots{}'', and a short commit notice has been
inserted just at the beginning of the message's body.

That's it. That was easy. Congratulations on your first shot at Patcher,
anyway ! Of course, Patcher is much more powerful and customizable than
what has been described in this chapter. For a complete documentation on
how to use and customize Patcher, please refer to @ref{User Manual}.



@c ====================================================================
@c User Manual
@c ====================================================================
@node User Manual, XEmacs Devel, Quick Start, Top
@comment  node-name,  next,  previous,  up
@chapter User Manual

This chapter provides a step-by-step guide to using Patcher. Everything
there is to know about Patcher is here, though the features are
introduced progressively.

All user options that are going to be presented in this manual can be
found in the @code{patcher} customization group, or a subgroup of it.

@menu
* Project Descriptors::  The project specification mechanism
* Mail Preparation::     Customizing Patcher messages
* Patch Generation::     Specifying how a patch is to be constructed
* ChangeLogs Handling::  How Patcher behaves with respect to ChangeLogs
* Project Check In::     Committing your changes from Patcher
* Mail Sending::         Sending the message and cleaning up the place
* More On Commands::     Error handling and other generalities
* More On Subprojects::  Defining permanent subprojects
@end menu

@ignore
The manual is currently organized in a rather sequential way: one
chapter for source patches, one for ChangeLogs. This might not be the
best way to do it. Another way is to split the information in terms of
patch/ChangeLog creation, and Patch/ChangeLog appearance in messages.
@end ignore

@c Project Descriptors ================================================
@node Project Descriptors, Mail Preparation, User Manual, User Manual
@comment  node-name,  next,  previous,  up
@section Project Descriptors

@vindex patcher-projects
Projects specifications are stored in @code{patcher-projects}. This user
option is actually a list of @dfn{project descriptors}. Each project
descriptor has the following form: @samp{(@var{NAME} @var{DIR}
@var{:OPTION} @var{VALUE} @dots{})}

@vindex patcher-default-to-address
@vindex :to-address
@var{NAME} is a string naming your project, @var{DIR} is a string
specifying the directory in which it resides. The remainder of a project
descriptor is a sequence of zero or more option/value pairs, that we
call @dfn{project options}. All option names start with a colon. The
type of a value depends on the corresponding option. For example, there
is a project option named @code{:to-address}, whose value should be a
string giving the email address to which you want to send Patcher
messages.

When Patcher needs the value for a particular project option, it looks
for it directly in the project descriptor, but also in other places.
This process is described below.

@menu
* Themes::                 Collections of options
* Project inheritance::    Getting options from other projects
* Fallbacks::              Default values for project options
* Retrieval::              The process of getting an option's value
* Inheritance or theme ?:: The proper way to factor out option values
@end menu

@node Themes, Project inheritance, , Project Descriptors
@comment  node-name,  next,  previous,  up
@subsection Themes

If you have several projects sharing the same option set, you might want
to setup a theme. Themes are named collections of project options.

@vindex patcher-themes
Themes are stored in the @code{patcher-themes} user option. This option
is a list of themes. Each theme has the following form:
@samp{(@var{NAME} @var{:OPTION} @var{VALUE} @dots{})}.

@var{NAME} is the theme's name (a symbol). The remainder of the list is
a sequence of zero or more option/value pairs, just like in project
descriptors.

@vindex patcher-default-themes
@vindex :themes
In order to use a theme in a given project, a @code{:themes} project
option is provided. It is a list of theme names (symbols). Use this
option in your project descriptor, and the project will implicitly
inherit all options from the corresponding theme.

One important note: as @code{:themes} is a project option, it can appear
in a theme. In other words, themes can inherit from other themes. When
Patcher tries to retrieve an option, the themes tree is traversed in
depth first.

@vindex patcher-max-theme-depth
Because themes can contain themes, a bogus setting might lead to an
infinite loop (a cycle in a theme graph). To prevent this, the
@code{patcher-max-theme-depth} user option is provided. It represents
the expected maximum theme nesting level, and defaults to 8.

@vindex patcher-built-in-themes
Patcher comes with a set of built-in themes for several revision control
systems. These are Git, Mercurial (Hg), Darcs, Subversion (Svn) and CVS
and PRCS. Look at the value of @code{patcher-built-in-themes} to see
what's in them. Each of these themes have a @code{-ws} counterpart which
eliminates white-space differences in diff outputs. This comes in handy
if you perform some kind of automatic white-space cleanup in the files
you edit, especially when you let Patcher generate the ChangeLog
entries.

While you can't modify the value of @code{patcher-built-in-themes},
you're free to do whatever you want in @code{patcher-themes},
including creating a theme with the same name as a built-in one. This
new theme will take precedence over the other. Having this built-in
variable (a constant, actually) lets me modify its value from release
to release without risking to smash your own adjustments.

@node Project inheritance, Fallbacks, Themes, Project Descriptors
@comment  node-name,  next,  previous,  up
@subsection Project inheritance

When two projects are very similar, you might want to use the project
inheritance mechanism described below.

@vindex :inheritance
There is a special project option called @code{:inheritance}. This
option must be a list of project names (strings). The inheritance of a
project defines a list of projects from which to inherit options.

One important note: inherited projects might have their own
@code{:inheritance} option set to other projects in turn. In other
words, the project inheritance can be more than one level deep. When
Patcher tries to retrieve an option, the inheritance tree is traversed
in depth first.

@vindex patcher-max-inheritance-depth
Because inherited projects can inherit from projects, a bogus setting
might lead to an infinite loop (a cycle in a project graph). To prevent
this, the @code{patcher-max-inheritance-depth} user option is provided.
It represents the expected maximum project inheritance level, and
defaults to 8.

The @code{:inheritance} project option is somewhat special in the sense
that it can't appear in a theme. We will encounter other exceptions
later in this manual.


@node Fallbacks, Retrieval, Project inheritance, Project Descriptors
@comment  node-name,  next,  previous,  up
@subsection Fallbacks

@vindex patcher-default-to-address
@vindex :to-address
For each existing project option, Patcher also has a @dfn{fallback} user
option with a default value that would be shared among all (or some of)
your projects. The name of the fallback is obtained by replacing the
colon in the project option's name with the prefix
@code{patcher-default-}. For example, the fallback corresponding to the
@code{:to-address} project option is named
@code{patcher-default-to-address}.

The @code{:inheritance} project option is also special in the sense that
it doesn't have a corresponding fallback. We will encounter other
exceptions later in this manual.

In the remainder of this manual, we will rarely mention the fallbacks
again. When we introduce a new project option, just remember that it
always has a corresponding fallback (well, not always, as you just
discovered).


@node Retrieval, Inheritance or theme ?, Fallbacks, Project Descriptors
@comment  node-name,  next,  previous,  up
@subsection Retrieval

When Patcher needs the value of a particular project option, it looks
for it in the following manner:

@itemize @bullet
@item
First, it looks directly in the project descriptor to see if the option
is given.

@item
@vindex patcher-default-themes
@vindex :themes
@vindex patcher-themes
If that fails, it next tries the given themes, if any. This involves
recursively traversing the project's themes tree. Options successfully
retrieved in themes are said to be @dfn{themed}.

@item
@vindex :inheritance
If that still fails, it then tries the inherited projects, if any. This
involves recursively traversing the project's inheritance tree. Options
successfully retrieved in inherited projects are said to be
@dfn{inherited}. Note that in turn, such options could have been
actually themed in the inherited project.

@item
If that fails again, it finally falls back to the value given in the
corresponding fallback (if it exists). In such a case, the option is
said to be @dfn{fallbacked}.
@end itemize

Note that a value of @code{nil} for a project option @strong{is an
actual value}. It is not equivalent to an option being unset. As a
consequence, if Patcher finds a project option with a value of
@code{nil} somewhere, it will use it and stop the search, even if a non
nil value could be retrieved later from a theme, an inherited project or
a fallback. This provides you with a way to annihilate themed,
inherited or fallbacked options.


@node Inheritance or theme ?, , Retrieval, Project Descriptors
@comment  node-name,  next,  previous,  up
@subsection Inheritance or theme ?

At that point, you might be wondering why the themes and inheritance
concepts were both designed, since they actually perform very similar
tasks. Good question. Here is a good answer.

Projects might share options for different reasons. For example, my
``XEmacs'' (source) and ``XEmacs Packages'' projects share many options
(@code{To:} address, @code{From:} address, diff and commit commands and
so on) because they both relate to XEmacs. On the other hand I have
personal but totally unrelated projects that share the same commands
because they are all handled through a common system: Darcs.

In other words, you should rather use the inheritance mechanism when
projects relate to each other, and the theme mechanism for settings that
are orthogonal the projects they apply to.



@c Mail Preparation ===================================================
@node Mail Preparation, Patch Generation, Project Descriptors, User Manual
@comment  node-name,  next,  previous,  up
@section Mail Preparation

Patcher currently uses the mail buffers as ``master'' buffers for
controlling all operations: building a patch, creating the ChangeLog
entries, committing@dots{} all is done from the mail buffer. Note
however that you don't need to actually send mails to use Patcher
(@pxref{Fake Mail Method}).

To use Patcher on a certain project, you start by preparing a (possibly
fake) mail. There are several ways to do so: you could start a brand new
message, ``adapt'' a message already in preparation to Patcher, or even
compose some sort of a Patcher reply to another message.

@defun patcher-mail
Start composing a brand new Patcher message. This function interactively
prompts you for the name of the project and for a (mail) subject line.
@end defun

@defun patcher-mail-adapt
Assuming that you are already editing a message (with your usual MUA),
this function ``adapts'' it to Patcher by prompting you for the name of
a project. Note that this function does @emph{not} prompt you for a
subject.
@end defun

If you're using Gnus to read mail and have properly insinuated it
(@pxref{Insinuation}), Patcher offers different Gnus-like ways to answer
mails and adapt them to Patcher. All the functions below are available
from both the Gnus Summary and the Article buffer.

@defun patcher-gnus-summary-followup
@kindex C-c C-p f
Compose a followup to the article, and adapt it to Patcher. This
function is bound to @kbd{C-c C-p f}.
@end defun

@defun patcher-gnus-summary-followup-with-original
@kindex C-c C-p F
Idem, but also cite the original article. This function is bound to
@kbd{C-c C-p F}.
@end defun

@defun patcher-gnus-summary-reply
@kindex C-c C-p r
Like @code{patcher-gnus-summary-followup}, but compose a reply. This
function is bound to @kbd{C-c C-p r}.
@end defun

@defun patcher-gnus-summary-reply-with-original
@kindex C-c C-p R
Idem, but also cite the original article. This function is bound to
@kbd{C-c C-p R}.
@end defun


In any case, Patcher starts working on the project (by first creating
the patch) after the message is prepared. Because of this, we'll start
by reviewing the mail-related customizations you might want to setup.

@menu
* Mail Methods::          Using a particular mailer
* Message Customization:: Specifying initial contents for messages
@end menu

@node Mail Methods, Message Customization, Mail Preparation, Mail Preparation
@comment  node-name,  next,  previous,  up
@subsection Mail Methods

@vindex patcher-default-mail-method
@vindex :mail-method
Since there are different mail packages working in XEmacs, Patcher
supports different methods for preparing messages. You can specify the
method you prefer in the @code{:mail-method} project option. The value
must be a symbol.

@menu
* Standard Mail Methods:: Patcher supports standard mail packages
* Fake Mail Method::      Patcher supports not sending mails
* Other Mail Methods::    Patcher supports everything
@end menu

@node Standard Mail Methods, Fake Mail Method, Mail Methods, Mail Methods
@comment  node-name,  next,  previous,  up
@subsubsection Standard Mail Methods

Patcher currently supports @file{sendmail}, @file{message} and
@file{Gnus} natively and through the @file{compose-mail} interface.
Other MUA might be partly supported when used with @code{compose-mail}.
Patcher will probably suffer from non critical deficiencies in that case
however (it will issue warnings).

@table @code
@item compose-mail
@findex compose-mail
@findex patcher-mail-compose-mail
@vindex mail-user-agent
This is the default. It is implemented via the function
@code{patcher-mail-compose-mail} which calls @code{compose-mail} to
prepare the message. If you are not familiar with @samp{compose-mail},
you might also want to throw an eye to the user option
@code{mail-user-agent}. If your project does not specify an address to
send the message to (@pxref{Message Customization}), it is prompted for.

@item sendmail
@findex mail
@findex patcher-mail-sendmail
A direct interface to the @code{mail} function from the @code{sendmail}
package. It is implemented via the function
@code{patcher-mail-sendmail}. If your project does not specify an
address to send the message to (@pxref{Message Customization}), it is
prompted for.

@item message
@findex message-mail
@findex patcher-mail-message
A direct interface to the @code{message-mail} function from the
@code{message} library (it is part of @code{Gnus}). It is implemented
via the function @code{patcher-mail-message}. If your project does not
specify an address to send the message to (@pxref{Message
Customization}), it is prompted for.

@item gnus
@findex gnus-post-news
@findex patcher-mail-gnus
A direct interface to the @code{gnus-post-news} function from the
@code{Gnus} package (it can also send mails@dots{}). It is implemented
via the function @code{patcher-mail-gnus}. This mail method is
interesting when you maintain a special mail group for messages that you
send with Patcher, most probably because they are sent to some
mailing-list, such as @email{xemacs-patches@@xemacs.org}.

@vindex patcher-default-gnus-group
@vindex :gnus-group
This method uses a Gnus group name and acts as if you had type @samp{C-u
a} on that group in the @code{*Group*} buffer, hence honoring the group
parameters and posting-styles. If your project does not specify a Gnus
group name (@pxref{Message Customization}), it is prompted for.
@end table

This last mail method is special in the sense that it requires a running
Gnus session to work. If that's needed, Patcher can start Gnus for you
in several ways, according to the following user options:

@table @code
@item patcher-mail-run-gnus
@vindex patcher-mail-run-gnus
If @code{nil}, Patcher will never start Gnus and abort the operation
instead. If @code{t}, Patcher will always start Gnus when needed. If
@code{'prompt}, Patcher will ask you what (you want) to do. This is the
default behavior.

@item patcher-mail-run-gnus-other-frame
@vindex patcher-mail-run-gnus-other-frame
Used when Patcher has to start Gnus by itself. If @code{nil}, continue
using the current frame. If @code{t}, start Gnus in another frame (this
is the default). If @code{'follow}, start Gnus in another frame, and use
this new frame to prepare the Patcher mail.
@end table

@node Fake Mail Method, Other Mail Methods, Standard Mail Methods, Mail Methods
@comment  node-name,  next,  previous,  up
@subsubsection Fake Mail Method

@findex patcher-mail-fake
At that point, you might be wondering why the mail method is a project
option and not simply a user option, since you probably only use one
mail agent at all. Right. But you might one day work on projects for
which you don't need to send messages at all. This could happen if you
start using Patcher on a project of your own for instance. For that
reason, there is a @code{fake} mail method available. It is implemented
via the @code{patcher-mail-fake} function and calls no particular mail
user agent. Once you type @samp{C-c C-c} to virtually send the fake
message, it only performs some cleanup.

All right. But did we really need this fake method ? I mean, one could
use the usual mail method, and simply not send the message in the end.
Huh, yeah, ok@dots{} Actually, it is very probable that in a future
release of Patcher, the mail buffer won't be the master buffer anymore,
and mail sending will be just another optional step in the process. In
that case, the mail method is likely to move away from project option
to standard user option.

@node Other Mail Methods, , Fake Mail Method, Mail Methods
@comment  node-name,  next,  previous,  up
@subsubsection Other Mail Methods

@vindex patcher-default-mail-method
@vindex :mail-method
If you're not satisfied with the provided mail methods (want a @code{vm}
one ?), you can provide your own, more or less. Here's what to do: set
your @code{:mail-method} project option to, say, @code{foo}, and write
your own function which must be named @code{patcher-mail-foo}.

This function must take two arguments (a project descriptor and a string
containing the subject of the message), and prepare a mail buffer. If
you want to do this, you should see how it's done for the built-in
methods.

Note that the mail adaptation facility won't be available for your
custom method. For that, I would have to hack the internals of Patcher.


@node Message Customization, , Mail Methods, Mail Preparation
@comment  node-name,  next,  previous,  up
@subsection Message Customization

When preparing a message, Patcher can fill some parts of it for you.
Here's a list of mail-related project options.

@table @code
@item :user-name
@vindex patcher-default-user-name
@vindex :user-name
The name (your name) to use when composing the message. It will affect
the @code{From:} header. This option is used by all mail methods but
@code{fake}. If not given, @code{user-full-name} is used.

@item :user-mail
@vindex patcher-default-user-mail
@vindex :user-mail
The mail (your mail) address to use when composing the message. It will
affect the @code{From:} header. This option is used by all mail methods but
@code{fake}. If not given, @code{user-mail-address} is used.

@item :to-address
@vindex patcher-default-to-address
@vindex :to-address
The address to send messages to (a string). This option is used by all
mail methods but @code{gnus} and @code{fake}. If not given, it is
prompted for when calling @code{patcher-mail}.

@item :gnus-group
@vindex patcher-default-gnus-group
@vindex :gnus-group
The Gnus group name to use for posting messages (a string). This option
is used only by the @code{gnus} mail method. If not given, it is
prompted for when calling @code{patcher-mail}.

Note that if you configured your name and mail in Gnus, for instance
through posting styles, these configurations take precedence over the
corresponding Patcher options.

@item :subject-prefix
@vindex patcher-default-subject-prefix
@vindex :subject-prefix
A prefix for the subject line of messages. It can be @code{nil} or a
string. By default, ``[PATCH]'' is used. This part of subjects is never
prompted for. A @samp{%n} occurring in this string will be replaced with
the project name. Also, a space is inserted between the prefix and the
remainder of the subject, when appropriate.

@item :subject
@vindex patcher-default-subject
@vindex :subject
A default value for prompted subjects (a string). A @samp{%n} occurring
in this string will be replaced with the project's name
(@pxref{Subproject Naming}). Please note that this is used @strong{only}
to provide a default value for prompted subjects. Subjects are
@strong{always} prompted for.

@item :mail-prologue
A prologue to insert just at the top of a message body (a string).
@end table

@c Patch Generation ===================================================
@node Patch Generation, ChangeLogs Handling, Mail Preparation, User Manual
@comment  node-name,  next,  previous,  up
@section Patch Generation

Patcher creates patches by diffing your local copy of the project
against the repository. This is done automatically after preparing a
message, so you shouldn't normally bother to do it manually. That
situation might still happen from time to time, like, if you further
modify the sources, or if the initial operation failed for some reason
(@pxref{More On Commands}).

@findex patcher-generate-diff
@kindex C-c C-p d
The way to (re)generate the patch manually is to call
@code{patcher-generate-diff} from the mail buffer. This function is
bound to @samp{C-c C-p d} in this buffer. You can both customize the
diff command used to (re)generate the patch, and restrict the diff to a
subset of the project's files.

When possible, Patcher also tries to check that your project is
up-to-date with respect to the archive, and will inform you otherwise.

@menu
* Diff Command::        Specifying the command to generate the patch
* Diff Headers::        Associating hunks with files
* Diff Line Filter::    Omitting lines from the diff output
* Patch Restriction::   Specifying the files affected by the patch
* Patcher Instances::   You can have several instances of Patcher
* Diff Prologue::       A header is inserted above the patch
@end menu

By the way, (re)generating the patch does not necessarily mean that it
is directly inserted into the mail buffer. This also depends on the
ChangeLogs behavior (@pxref{ChangeLogs Handling}).

@node Diff Command, Diff Headers, Patch Generation, Patch Generation
@comment  node-name,  next,  previous,  up
@subsection Diff Command

@vindex patcher-default-diff-command
@vindex :diff-command
@findex patcher-mail
The diff command used to generate the patch is specified by the
@code{:diff-command} project option. You can also punctually change this
command by calling @code{patcher-mail} or @code{patcher-generate-diff}
with a prefix argument. Patcher will then prompt you for a new command
and use it exclusively for this particular patch.

For detail about the format of this option, see @ref{Patch Restriction}.
See also @ref{More On Commands}.

@node Diff Headers, Diff Line Filter, Diff Command, Patch Generation
@comment  node-name,  next,  previous,  up
@subsection Diff Headers

When Patcher generates a diff, it needs to associate every hunk with the
corresponding source file (and possibly with the corresponding ChangeLog
file as well). Unfortunately, different revision control systems might
generate different diff outputs, making this association difficult to
establish.

@vindex patcher-default-diff-header
@vindex :diff-header
Patcher provides a @code{:diff-header} project option to help. Its value
is of the form @code{(REGEXP NUMBER1 NUMBER2)}. @code{REGEXP} is used to
match the beginning of a diff output while NUMBER1 and NUMBER2 are the
parenthesized levels in which to find the corresponding old and new file
names.

When the change involves modifying a file's contents, the old and new
file names will be the same. However, they can be different in several
situations, like when a file is renamed, created or deleted. In case of
creation or deletion, some revision control systems use
``@code{/dev/null}'' to denote a virtual old or new file.

@vindex patcher-built-in-themes
If you want to see some examples, have a look at the built-in themes in
@code{patcher-built-in-themes} (@pxref{Themes}). They contain presets
for different revision control systems, along with suitable
@code{:diff-header} options.

Also, you should pay attention to the fact that the values of the
@code{:diff-header} and @code{:diff-command} options may depend on one
another to work properly. For instance, the diff output of Mercurial
looks different when you use the @code{--git} option.


@node Diff Line Filter, Patch Restriction, Diff Headers, Patch Generation
@comment  node-name,  next,  previous,  up
@subsection Diff Line Filter
When generating a global diff, that is, without specifying the files
affected by the patch explicitly (@pxref{Patch Restriction}), some
uninformative lines might be present in the output. A typical example is
files present in your local copy but unknown to the CVS server. Such
files are indicated by a question mark in diff outputs.

@vindex patcher-default-diff-line-filter
@vindex :diff-line-filter
Patcher has a project option named @code{:diff-line-filter} that lets
filter out such unwanted lines when inserting the diff in the mail
buffer. This must be a regular expression matching a whole line. Caution
however: do not put beginning or end of lines markers in your regexp.
Patcher will do it for you.

By default, the value is @code{"\\? .*"}, which excludes files unknown
to the CVS server.

@node Patch Restriction, Patcher Instances, Diff Line Filter, Patch Generation
@comment  node-name,  next,  previous,  up
@subsection Patch Restriction

Sometimes, you don't want to generate a global patch, but instead work
on a subset of the project's files. Patcher understands this concept as
working on a @dfn{subproject}.

As for working on whole projects, there are several alternatives to
start working on a subproject:

@defun patcher-mail-subproject
@findex patcher-mail
This function behaves exactly like @code{patcher-mail}, except that it
also prompts you for the files affected by the patch. You can specify
files as well as directories, use wildcards, just as you would construct
a command line diff.
@end defun

@defun patcher-mail-adapt-subproject
@findex patcher-mail-adapt
Idem, with respect to @code{patcher-mail-adapt}.
@end defun

@findex patcher-gnus-summary-followup
@findex patcher-gnus-summary-followup-with-original
@findex patcher-gnus-summary-reply
@findex patcher-gnus-summary-reply-with-original
Finally, all the Gnus-specific reply functions (@pxref{Mail
Preparation}) take an optional prefix argument that make them work on
subprojects.


Two important things are to be kept in mind when working on subprojects:
@itemize @bullet
@item
It is not necessary (it is even forbidden) to specify the ChangeLog
files. Patcher will automatically find them for you. In other words,
only specify source files, not ChangeLog files.

@item
@findex patcher-mail
@findex patcher-mail-subproject
@findex patcher-generate-diff
Don't mistake the prefix argument of @code{patcher-mail},
@code{patcher-mail-subproject} and @code{patcher-generate-diff} as a way
to specify files. It is not meant for that. It is meant only to
temporarily modify the diff command itself, not the files to which it
applies.
@end itemize

@vindex patcher-default-diff-command
@vindex :diff-command
When working on a subproject, Patcher uses the same diff command as the
one used for the whole project, that is, the value of the
@code{:diff-command} project option. In the case of a subproject
however, Patcher has to know where to put the files on the command line
(even if it's usually the last part of it). This is done with the special
construct @samp{%f}: a @samp{%f} appearing in the diff command will be
replaced with the specified files (and the computed ChangeLog files) for
a subproject, and will simply be removed when working on the whole
project.

For example, the built-in CVS theme (@pxref{Themes}) has the
@code{:diff-command} option set to @samp{cvs -q diff -u %f}.

There is another special construct available in this command: a
@samp{%n} will be replaced with the project's name (@pxref{Subproject
Naming}). This may be useful in commands with weird options syntax, like
PRCS.

For a more advanced way of handling subprojects, see @ref{More On
Subprojects}.

@node Patcher Instances, Diff Prologue, Patch Restriction, Patch Generation
@comment  node-name,  next,  previous,  up
@subsection Patcher Instances

The concept of subprojects brings up the question of having Patcher
working on different patches at the same time. It is possible under some
conditions:

@itemize @bullet
@item
You can have as many simultaneous Patcher instances as you want on
different projects (you would not define two projects sharing the same
directory, would you ?).

@item
You can have as many simultaneous Patcher instances as you want on the
same project, as long as there is no overlapping between each
subproject. This means that you can't have source files, or even
ChangeLog files in common.
@end itemize

This last point will be subject to improvements in the future.


@node Diff Prologue, , Patcher Instances, Patch Generation
@comment  node-name,  next,  previous,  up
@subsection Diff Prologue

Patcher can (and does) insert a special prologue just above a patch in
the message in preparation. This prologue gives information such as the
diff command used, the files affected and so on.

@vindex patcher-default-diff-prologue-function
@vindex :diff-prologue-function
@findex patcher-default-diff-prologue
The function used to generate this prologue can be specified with the
@code{:diff-prologue-function} project option. A value of @code{nil}
means don't insert any prologue. By default, the internal function
@code{patcher-default-diff-prologue} is used. If you want to provide
your own, here how to do it.

Your function should take one argument indicating the kind of diff (a
symbol), and perform insertion at current point in the current buffer.
The possible values for the argument are:

@table @code
@item sources
indicates a source diff only,

@item change-logs
indicates a ChangeLog diff only,

@item mixed
indicates a diff on both source and ChangeLog files.
@end table

The following variables are bound (when appropriate) when this function
is executed:

@table @code
@item name
the name of the current Patcher project,
@item source-diff
the command used to create a source diff,
@item change-log-diff
the command used to create a ChangeLog diff,
@item source-files
sources files affected by the current patch,
@item change-log-files
ChangeLog files affected by the current patch.
@end table

 In the case of a mixed diff, a @code{nil} value for
@code{change-log-diff} indicates that the same command was used for both
the source and ChangeLog files.



@c ChangeLog Handling =================================================
@node ChangeLogs Handling, Project Check In, Patch Generation, User Manual
@comment  node-name,  next,  previous,  up
@section ChangeLogs Handling

ChangeLogs management in Patcher involves two aspects: how ChangeLog
entries are created, and how they appear in the messages. Both aspects
can be customized beyond your craziest dreams.

@menu
* ChangeLogs Updating::   How Patcher deals with ChangeLogs
* ChangeLogs Appearance:: How ChangeLogs appear in messages
* ChangeLogs Prologue::   A header is inserted above the ChangeLogs
@end menu

@node ChangeLogs Updating, ChangeLogs Appearance, ChangeLogs Handling, ChangeLogs Handling
@comment  node-name,  next,  previous,  up
@subsection ChangeLogs Updating

@vindex patcher-default-change-logs-updating
@vindex :change-logs-updating
The way Patcher deals with ChangeLogs is controlled via the
@code{:change-logs-updating} project option. Its value (a symbol) must
be one of @code{automatic} (the default), @code{manual} or @code{none}.

@menu
* Automatic ChangeLogs::        Automatic creation of ChangeLog skeletons
* Manual ChangeLogs::           When ChangeLog entries already exist
* No ChangeLogs::               For project that don't use ChangeLog files
@end menu

@node Automatic ChangeLogs, Manual ChangeLogs, , ChangeLogs Updating
@comment  node-name,  next,  previous,  up
@subsubsection Automatic ChangeLogs

Automatic ChangeLogs mode is the default. Each time you (re)generate a
diff, Patcher (re)creates ChangeLog skeletons in the appropriate
ChangeLog files, by analyzing the generated diff. You then need to fill
the entries manually.

@menu
* Skeleton Generation:: How skeletons are created.
* Skeleton Parameters:: What the skeletons look like.
* ChangeLog Files::     Noticing and saving ChangeLogs.
@end menu

@node Skeleton Generation, Skeleton Parameters, Automatic ChangeLogs, Automatic ChangeLogs
@comment  node-name,  next,  previous,  up

@findex patch-to-change-log
@vindex patcher-default-diff-cleaner
@findex patcher-generic-diff-cleaner
@vindex :diff-cleaner
ChangeLog skeletons are not generated by Patcher directly, but rather by
the function @code{patch-to-change-log} of the @code{add-log} library
from the @code{xemacs-base} package. This function supports only
standard and CVS diff, in unified format. For revision control systems
that output something different, Patcher provides a @code{:diff-cleaner}
option. This option names a function that will be used to ``cleanup''
the diff (so that it looks like a standard one, just before calling
@code{patch-to-change-log}. Patcher also comes with a generic cleaner
function named @code{patcher-generic-diff-cleaner} which is used by
default and works correctly with Git, Mercurial, Darcs and PRCS, as long
as you use the corresponding built-in themes (@pxref{Themes}), or in
fact, as long as the corresponding @code{:diff-header} option is correct
(@pxref{Diff Headers}).

@vindex patcher-default-kill-source-files-after-diffing
@vindex :kill-source-files-after-diffing
In order to build the ChangeLog skeletons, the source files must be
visited. Patcher has a @code{:kill-source-files-after-diffing} project
option that lets you decide whether you want to kill the source buffers
just after diffing or not. The value can be either @code{nil} or
@code{t}. By default, the behavior is to preserve the source files,
since you might need them for reference when filling the ChangeLog
entries. If you don't use automatic ChangeLogs mode, it makes sense to
set this option to @code{t}.

@node Skeleton Parameters, ChangeLog Files, Skeleton Generation, Automatic ChangeLogs
@comment  node-name,  next,  previous,  up

@vindex patcher-default-change-logs-user-name
@vindex :change-logs-user-name
@vindex patcher-default-user-name
@vindex :user-name
@vindex patcher-default-change-logs-user-mail
@vindex :change-logs-user-mail
@vindex patcher-default-user-mail
@vindex :user-mail
@findex patch-to-change-log
@vindex user-full-name
@vindex user-mail-address
Patcher has two project options that give you some control on the
generated ChangeLog skeleton: @code{:change-logs-user-name} and
@code{:change-logs-user-mail}. As you might expect, these are strings
defining your name and mail address for ChangeLog entries'headers. When
@code{nil}, Patcher falls back to (respectively) the @code{:user-name}
and @code{:user-mail} project options. If in turn set to nil, Patcher
lets the function @code{patch-to-change-log} decide what to use (most
probably what the user options @code{user-full-name} and
@code{user-mail-address} say).

@node ChangeLog Files, , Skeleton Parameters, Automatic ChangeLogs
@comment  node-name,  next,  previous,  up

Normally, you don't modify source files when working with Patcher.
However, ChangeLog files need update and saving in automatic mode.
Patcher provides two hooks for plugging in additional processing on
ChangeLog files.

@itemize @bullet
@item @code{:notice-change-log-hook}
@vindex patcher-default-notice-change-log
@vindex :notice-change-log-hook
This hook is run every time Patcher ``notices'' a ChangeLog file. 
Noticing a ChangeLog file in this context means figuring out that it is
involved in the current patch. Every function in this hook will be given
the ChangeLog file name (relative to the project's directory) as
argument.
@item @code{:after-save-change-log-hook}
@vindex patcher-default-after-save-change-log-hook
@vindex :after-save-change-log-hook
This hook is run every right after you save a ChangeLog file. The
functions in this hook are executed in the ChangeLog's buffer.
@end itemize

@node Manual ChangeLogs, No ChangeLogs, Automatic ChangeLogs, ChangeLogs Updating
@comment  node-name,  next,  previous,  up
@subsubsection Manual ChangeLogs

In manual mode, Patcher assumes that you create ChangeLog entries
manually, as you write the code, so it won't create ChangeLog skeletons.
It is important to understand that in this situation, ChangeLog entries
@strong{must} have been written @strong{before} you call Patcher.
Patcher won't let you write them in the process.

Even in @code{manual} mode, Patcher might still need to know the
affected ChangeLog files (for the commit process) and your exact
ChangeLog entries in each of these files (for insertion in the message).
The ChangeLog files are automatically deduced from the patch. When
that's required, however, you will be presented with each ChangeLog file
in turn, and invited to precise the number of ChangeLog entries
concerning this patch. These entries must of course appear at the top of
the file.

@node No ChangeLogs, , Manual ChangeLogs, ChangeLogs Updating
@comment  node-name,  next,  previous,  up
@subsubsection No ChangeLogs

This mode is for projects that don't do ChangeLogs. Patcher won't try to
create ChangeLog entries, and won't expect that you have written
ChangeLog entries either.

Note that if you @emph{do} have ChangeLog files in this mode, they will
be regarded just as normal ones. As a consequence, this is a case where
it is not forbidden to list them explicitly as part of a subproject
(@pxref{Patch Restriction}), although I don't see why you would want to
do that.


@node ChangeLogs Appearance,  ChangeLogs Prologue, ChangeLogs Updating, ChangeLogs Handling
@comment  node-name,  next,  previous,  up
@subsection ChangeLogs Appearance

@findex patcher-insert-change-logs
@kindex C-c C-p i
When Patcher has to deal with ChangeLog files, you insert ChangeLog
entries in the message buffer by calling the function
@code{patcher-insert-change-logs}. It is bound to @samp{C-c C-p i} in
the mail buffer. Note that if ChangeLogs are written in advance, you
don't have to do that, as Patcher will do it automatically. You have to
call this function by hand only in @code{automatic} updating mode
(@pxref{ChangeLogs Updating}).

In all circumstances however, you can use this function to
@strong{reinsert} the ChangeLog entries into the mail buffer. That might
come in handy if you further modified them after the initial insertion
(it's never too late to fix a typo).

@vindex patcher-default-change-logs-appearance
@vindex :change-logs-appearance
The appearance of ChangeLog entries in the message is controlled by the
@code{:change-logs-appearance} project option. Its value must be a
symbol from the following:

@table @code
@item verbatim
This is the default. ChangeLog entries appear just as text in the
message, above the patch. Most people prefer this kind of appearance
since it is the most readable.

@item packed
ChangeLog entries appear as a patch (they are diff'ed against the
archive). This patch is however distinct from the source patch, and
appears above it.

@item patch
ChangeLog entries appear as a patch (they are diff'ed against the
archive), and this patch is integrated into the source patch. In other
words, the message looks like a global patch integrating both the
sources and the ChangeLogs.
@end table

@vindex patcher-default-change-logs-diff-command
@vindex :change-logs-diff-command
When the ChangeLogs appearance is either @code{packed} or @code{patch},
the diff command used to generate the patch is controlled by the
@code{:change-logs-diff-command} project option. The value can be the
symbol @code{diff}, meaning that the same diff command is to be used as
for the sources (@pxref{Diff Command}), or it can be a string specifying
an alternate command.

When diffing ChangeLog files, it is strongly recommended that you remove
contexts from the diff, because otherwise, ChangeLog patches often fail
to apply correctly.

@vindex patcher-default-diff-command
@vindex :diff-command
As in the case of the @code{:diff-command} project option (@pxref{Patch
Restriction}), a @code{%f} serves to indicate where to put the ChangeLog
files on the diff command line, and a @samp{%n} is replaced with the
project's name (@pxref{Subproject Naming}). See also @ref{More On
Commands}.

@vindex patcher-default-change-logs-diff-command
@vindex :change-logs-diff-command
Given these two remarks, you should now understand why the default value
for @code{patcher-default-change-logs-diff-command} is @samp{cvs -q diff
-U 0 %f}.

@node ChangeLogs Prologue, , ChangeLogs Appearance, ChangeLogs Handling
@comment  node-name,  next,  previous,  up
@subsection ChangeLogs Prologue

ChangeLog prologues are small pieces of informative text that Patcher
adds above each ChangeLog insertion in the mail buffer.

@vindex patcher-default-change-logs-prologue
@vindex :change-logs-prologue
When the ChangeLogs appearance is @code{verbatim}, Patcher inserts one
prologue per ChangeLog file. The prologue's contents is controlled by
the @code{:change-logs-prologue} project option (a string). A @samp{%f}
appearing in this string will be replaced with the ChangeLog filename.
The default value for @code{patcher-default-change-logs-prologue} is
@code{"%f addition:"}.

@vindex patcher-default-diff-prologue-function
@vindex :diff-prologue-function
@findex patcher-default-diff-prologue
When the ChangeLogs appearance is @code{packed}, Patcher inserts only
one prologue for the whole ChangeLogs patch. When @code{patch}, there is
a single prologue for both the ChangeLogs and the sources. For
customizing the prologue in both of these cases, see @ref{Diff
Prologue}.


@c Project Check In ===================================================
@node Project Check In, Mail Sending, ChangeLogs Handling, User Manual
@comment  node-name,  next,  previous,  up
@section Project Check In

@findex patcher-commit-change
@kindex C-c C-p c
If you have the privilege to commit your changes yourself, you might do
so directly from the mail buffer, as the last operation before actually
sending the message. This is done by calling the function
@code{patcher-commit-change} which is bound to @samp{C-c C-p c} in the
mail buffer.

Committing directly from Patcher has the advantage that both the commit
command line and the commit log message (@pxref{Log Message Handling})
are constructed automatically. Of course, you still have an in-depth
control on the commit process.

@menu
* Commit Command::       Specifying the command to commit the patch
* Log Message Handling:: Building a commit log message
* Commit Operation::     Actually committing your changes
@end menu

@node Commit Command, Log Message Handling, Project Check In, Project Check In
@comment  node-name,  next,  previous,  up
@subsection Commit Command

@findex patcher-commit-change
@vindex patcher-default-commit-command
@vindex :commit-command
The command used to to commit a patch is specified by the
@code{:commit-command} project option (a string). You can also
temporarily change the command in question by calling
@code{patcher-commit-change} with a prefix argument. As usual, note that
this prefix argument is not meant to modify the affected files on the
command line. It's meant only to punctually modify the commit command
itself. The affected files are computed automatically by Patcher.

The commit command accepts (and maybe requires) special constructs:

@itemize
@item
A @code{%n} occurring in the string will be replaced with the project's
name (@pxref{Subproject Naming}).

@item
A @code{%f} occurring in the string will be replaced with the files
affected by the patch (ChangeLogs included, this time), or discarded if
the patch is global.

@item
A @code{%?f@{blabla@}} occurring in the string will be replaced with
``blabla'' if @code{%f} would expand to something, or discarded
otherwise (this is an ``if @code{%f}'' form).

@item
A @code{%!f@{blabla@}} occurring in the string will be replaced with
``blabla'' if @code{%f} would expand to nothing, or discarded otherwise
(this is an ``if not @code{%f}'' form).

@item
A @code{%s} occurring in the string will be replaced with the name of
a file containing the commit log message (@pxref{Log Message Handling}).
This file is a temporary file handled by Patcher.

@item
A @code{%S} occurring in the string will be replaced with the commit log
message itself (@pxref{Log Message Handling}). Since the intent here is
to use the message as a command-line argument, it is quoted against
shell expansion.
@end itemize

Please note that a @code{%f} is required in your commit commands, unless
you know for sure that you will never ever work on a subproject
(@pxref{Patch Restriction}). But you never know that. Besides, you
should always also provide either a @code{%s} or a @code{%S}, unless
your archival software does not support log messages.

@vindex patcher-default-commit-command
@vindex :commit-command
Given that, you should now understand why the default value for
@code{patcher-default-commit-command} is @samp{cvs commit -F %s %f}.

@node Log Message Handling, Commit Operation, Commit Command, Project Check In
@comment  node-name,  next,  previous,  up
@subsection Log Message Handling

Most project management tools understand the concept of a @dfn{log
message}: a short yet informative message that accompany the commit
operation, which is also stored in the repository.

Before a commit operation, Patcher always builds an initial log message,
based on certain elements under your control (@pxref{Log Message
Elements}). Afterwards, you can decide on whether you want some further
(manual) edition of log message, or whether you want to commit
immediately.

@vindex patcher-default-edit-log-message
@vindex :edit-log-message
The @code{:edit-log-message} project option lets you specify this
behavior. If @code{t} (the default), you will be able to manually edit
the log message. If @code{nil}, Patcher will proceed directly to the
commit operation.

Please note that Patcher stores log messages in temporary files that
will be used later by the commit command.

@menu
* Log Message Elements:: Elements Patcher can add automatically
* Log Message Editing::  Manually modifying the log message
@end menu

@node Log Message Elements, Log Message Editing, Log Message Handling, Log Message Handling
@comment  node-name,  next,  previous,  up
@subsubsection Log Message Elements

@vindex patcher-default-log-message-items
@vindex :log-message-items
Patcher has the ability to initialize the log message with different
elements. These elements are specified with the
@code{:log-message-items} project option. Its value is either
@code{nil}, meaning that you don't want any initialization, or a list of
symbols specifying the elements you desire. The available items are:

@table @code
@item subject
The subject of the message. The subject's prefix is automatically
removed.

@item compressed-change-logs
The ``compressed'' ChangeLog entries. Only the most important part of
the ChangeLogs is preserved, so the entries appear in a more compact
fashion.

@item change-logs
The raw ChangeLog entries.
@end table

@vindex patcher-default-change-logs-separator
@vindex :change-logs-separator
By default, only the message's subject is used. When using more than one
item, they appear in the order specified above. If anything appears
before the raw ChangeLog entries, a separator string is used. This
string is specified by the @code{:change-logs-separator} project option.
By default the string looks like ``--- ChangeLog entries follow: ---''.


@node Log Message Editing, , Log Message Elements, Log Message Handling
@comment  node-name,  next,  previous,  up
@subsubsection Log Message Editing

If so required, Patcher lets you manually edit the log message after
having initialized it. Log message edition happens in a special buffer,
the @code{*Patcher Log Message*} buffer.

While editing this buffer, commands to insert the items described in
@ref{Log Message Elements} are at your disposal. These commands perform
insertion at point:

@table @code
@item patcher-logmsg-insert-subject
@findex patcher-logmsg-insert-subject
@kindex C-c C-p s
Bound to @samp{C-c C-p s}. Insert the message's subject (sans the
prefix).

@item patcher-logmsg-insert-compressed-change-logs
@findex patcher-logmsg-insert-compressed-change-logs
@kindex C-c C-p c
Bound to @samp{C-c C-p c}. Insert the compressed ChangeLog entries.

@item patcher-logmsg-insert-change-logs
@findex patcher-logmsg-insert-change-logs
@kindex C-c C-p l
Bound to @samp{C-c C-p l}. Insert the raw ChangeLog entries. Use a
prefix argument if you also want the ChangeLogs separator string to be
inserted.
@end table

@findex patcher-logmsg-init-message
@kindex C-c C-p i
In addition to these commands, you can also completely reinitialize the
log message according to the specified items, by calling the function
@code{patcher-logmsg-init-message}, bound to @samp{C-c C-p i}. Caution:
this command first erases the buffer.

@findex patcher-logmsg-commit
@kindex C-c C-c
Finally, once you're happy with your log message, you proceed to the
commit operation by calling the function @code{patcher-logmsg-commit},
bound to @samp{C-c C-c}.

@node Commit Operation, , Log Message Handling, Project Check In
@comment  node-name,  next,  previous,  up
@subsection Commit Operation

@kindex C-c C-p c
@kindex C-c C-c
The commit operation occurs after typing @samp{C-c C-p c} from the mail
buffer, if you have not required log message edition, or after typing
@samp{C-c C-c} from the @code{*Patcher Log Message*} buffer otherwise.

@vindex patcher-default-confirm-commits
@vindex :confirm-commits
At that point, Patcher has constructed a proper commit command. You
still have an option to abort the process, though: the
@code{:confirm-commits} project option lets you specify whether to
require confirmation before committing. If @code{t} (the default),
Patcher will display the constructed commit command and ask for a
confirmation before proceeding. If @code{nil}, Patcher will commit
immediately.

After the commit operation, Patcher changes some parts of the mail
buffer in the following manner:

@itemize @bullet
@item
@vindex patcher-default-subject-committed-prefix
@vindex :subject-committed-prefix
The subject prefix is changed to that specified by the
@code{:subject-committed-prefix} project option (a string), unless it is
@code{nil}. By default, ``[COMMIT]'' is used.

@item
@vindex patcher-default-committed-notice
@vindex :committed-notice
A commit notice is added right at the beginning of the message's body.
This notice is specified by the @code{:committed-notice} project option.
It can be @code{nil} or a string. By default, it reads ``NOTE: this
patch has been committed.''.
@end itemize


@c Mail Sending =======================================================
@node Mail Sending, More On Commands, Project Check In, User Manual
@comment  node-name,  next,  previous,  up
@section Mail Sending

Sending the message will most probably be done by typing @samp{C-c C-c}
in the mail buffer. This is also the case when you're using the fake
mail method, by the way.

@menu
* Before Sending:: Patcher does some checkings
* After Sending::  patcher does some cleanup
@end menu

@node Before Sending, After Sending, Mail Sending, Mail Sending
@comment  node-name,  next,  previous,  up
@subsection Before Sending

There are circumstances in which Patcher will perform some checkings on
your message when you send it, just before it is actually sent:

@itemize @bullet
@item ChangeLogs insertion

@vindex patcher-mail-check-change-logs-insertion
In case of manual ChangeLog insertion (@pxref{ChangeLogs Appearance}),
Patcher can check that you have indeed inserted the ChangeLog entries
before sending the message. This behavior is controlled by the
@code{patcher-mail-check-change-logs-insertion} user option. A value of
@code{nil} means never check (the message will be sent as-is). A value
of t means check, and abort the sending if the ChangeLog entries are
missing. A value of @code{ask} (the default) means ask for your opinion
on this terrible matter.

@item Commit Operation

@vindex patcher-default-commit-privilege
@vindex :commit-privilege
Patcher has a @code{:commit-privilege} project option; a boolean
specifying whether you're likely to commit your changes by yourself.

@vindex patcher-mail-check-commit-action
In case of commit privilege, Patcher can check that you have indeed
committed your changes before sending the message. This behavior is
controlled by the @code{patcher-mail-check-commit-action} user option. A
value of @code{nil} means never check (the message will be sent as-is).
A value of t means check, and abort the sending if the commit operation
has not been performed. A value of @code{ask} (the default) means ask
for your opinion on this rather unfortunate situation.
@end itemize


Two notes on these checkings:

@itemize @bullet
@item
For uninteresting technical reasons, Patcher does not currently (and
will probably never) offer you an automatic ChangeLogs insertion or
commit operation, at mail sending time, but just abort the sending
process in some circumstances. That's not a big deal though.

@item
For other uninteresting technical reasons, these checkings require a
native knowledge of your mail user agent. Patcher does not currently
support all mail user agents on earth (I'll add them on demand however).
If that's the case, you will be warned and invited to send me a message.
Also, you can send me one on April 21st: it's my birthday.
@end itemize

@node After Sending, , Before Sending, Mail Sending
@comment  node-name,  next,  previous,  up
@subsection After Sending

After sending the message, Patcher also performs some cleanup
operations, that you can customize. The cleanup is controlled by the
following project options. Each one is a boolean option which defaults
to @code{t}.

@table @code
@item :kill-source-files-after-sending
@vindex patcher-default-kill-source-files-after-sending
@vindex :kill-source-files-after-sending
Whether to kill source files after sending the message. If @code{nil},
the source files will remain visited.

@item :kill-change-logs-after-sending
@vindex patcher-default-kill-change-logs-after-sending
@vindex :kill-change-logs-after-sending
Whether to kill ChangeLog files after sending the message. If
@code{nil}, the ChangeLog files will remain visited.
@end table


@c More On Commands ===================================================
@node More On Commands, More On Subprojects, Mail Sending, User Manual
@comment  node-name,  next,  previous,  up
@section More On Commands

This section deals with information that apply to all commands used by
Patcher (diff and commit operations).

@menu
* Prefixing Commands::  Bypassing a firewall
* Error Handling::      Dealing with errors
@end menu

@node Prefixing Commands, Error Handling, , More On Commands
@comment  node-name,  next,  previous,  up
@subsection Prefixing Commands

If you're working on a distant archive and you're behind a firewall, you
might need to prefix all your commands with something like
@code{runsocks}. Of course, this can be done manually in all your
command settings, but Patcher offers you a simpler way to do it.

@vindex patcher-default-pre-command
@vindex :pre-command
There is a project option named @code{:pre-command} which can be used
for this kind of thing. It must be a string that will be prepended to
all operations performed by Patcher.


@node Error Handling, , Prefixing Commands, More On Commands
@comment  node-name,  next,  previous,  up
@subsection Error Handling

@vindex patcher-default-failed-command-regexp
@vindex :failed-command-regexp
From time to time, commands may fail for different reasons. Patcher
tracks the failures in these commands and lets you know when that
happens. The @code{:failed-command-regexp} project option lets you
specify a regular expression to match with the output of an aborted
command. By default, the value is @code{"^cvs \\[[^]]* aborted\\]"}.


@c More On Subprojects ===============================================
@node More On Subprojects, , More On Commands, User Manual
@comment  node-name,  next,  previous,  up
@section More On Subprojects

We've already seen how to work punctually on a subproject, that is, on a
subset of a project's files (@pxref{Patch Restriction}). If you happen
to work more than once on the same project subset, it will quickly
become annoying to have to specify explicitly the same files over and
over again. Patcher offers you a way to permanently define subprojects.

@menu
* Defining Subprojects::  Storing subprojects definitions
* Subproject Naming::     When different projects have the same name
* Command Directory::     For non local revision control systems
@end menu


@node Defining Subprojects, Subproject Naming, , More On Subprojects
@comment  node-name,  next,  previous,  up
@subsection Defining Subprojects

@vindex patcher-subprojects
The user option @code{patcher-subprojects} stores a list of
@dfn{subproject descriptors}. A subproject descriptor is almost the same
as a project descriptor, with a few exceptions:

@itemize @bullet
@item
Instead of the project directory field (the second field in a project
descriptor), you rather specify the name of the project this subproject
is based on.

@item
In addition to the standard project options we've already seen, two
subproject options are available:
@table @code
@item :subdirectory
@vindex :subdirectory
This lets you specify a subdirectory of the original project's directory
in which the whole subproject resides.

@item :files
@vindex :files
This lets you specify a list of files or directories composing the
subproject. Each file specification can contain wildcards. If a
@code{:subdirectory} subproject option is given, these files or
directories should be relative to this subdirectory. Otherwise, they
should be relative to the parent project's directory.
@end table

Please note that these subproject options have no corresponding fallback
(that would be meaningless). They can't appear in a theme either.

@item
@vindex :inheritance
Subprojects don't have an @code{:inheritance} mechanism. Instead, they
implicitly inherit from their base project (which in turn can inherit
from other projects).
@end itemize

Here are some important remarks about subprojects:

@itemize @bullet
@item
@vindex :subdirectory
@vindex :files
A subproject with neither a @code{:subdirectory} nor a @code{:files}
option is exactly the same as the base project, apart from project
options that you would override.

@item
@findex patcher-mail
@findex patcher-mail-subproject
@findex patcher-mail-adapt
@findex patcher-mail-adapt-subproject
Since permanent subprojects are defined statically, they should be used
with the normal Patcher functions like @code{patcher-mail}. You don't
need to use the @code{*-subproject} versions any more.

@item
Because of that, projects and subprojects can't have names in common. If
that happens to be the case, Patcher will blindly use a subproject
rather than a project of that name

@item
So what happens if you persist in using a @code{*-subproject} function
on a permanent subproject ? Well, you punctually work on a
subsubproject@dots{}
@end itemize


@node Subproject Naming, Command Directory, Defining Subprojects, More On Subprojects
@comment  node-name,  next,  previous,  up
@subsection Subproject Naming

@vindex patcher-projects
@vindex patcher-subprojects
As you already know, Patcher distinguishes projects by their @var{NAME}
field in the @code{patcher-projects} and @code{patcher-subprojects} user
options. This is problematic, however, if you define different patcher
projects for a single physical project, and the commands require the
project's name. That might for instance be the case with PRCS
(decidedly) weird options syntax.

@vindex patcher-default-name
@vindex :name
To remedy this problem, patcher provides a @code{:name} project option.
If set, it will actually be used for all @code{%n} substitutions where
the project's name (in the Patcher sense) would normally be used.

If you want to use the project's name in the Patcher sense, regardless
of the value of the @code{:name} option, use @code{%N} instead of
@code{%n}.


@node Command Directory, , Subproject Naming, More On Subprojects
@comment  node-name,  next,  previous,  up
@subsection Command Directory

Some revision control systems, like CVS, work on a per-file basis. Some
others, like PRCS, work only globally. This has an important implication
for subprojects being sudirectories of other projects. Indeed, while
CVS, for instance, can be executed anywhere, PRCS can only be run in the
base project's root directory.

@vindex :command-directory
@vindex patcher-default-command-directory
If you want to define projects for which the revision control system can
be executed in only one directory, Patcher provides you with the
@code{:command-directory} project option (a string). This directory can
be specified relative to the project's directory (but note that it must
go upwards).

All commands (diff and commit ones) will be executed from there. Also,
note that the command directory does not change the way you might
specify files. Patcher modifies all needed paths automatically to handle
the command directory properly.


@c ====================================================================
@c XEmacs Devel
@c ====================================================================
@node XEmacs Devel, Variables Index, User Manual, Top
@comment  node-name,  next,  previous,  up
@chapter XEmacs Development

Here is some specific information on how to setup Patcher for XEmacs
developers. This assumes you're a committer to the mercurial
repository of the 21.5 branch.

The way I do it, I have a directory named
@code{/usr/local/src/xemacs-21.5} which is a clone of the Debian
Mercurial repository. Let's call it the @emph{local trunk}. I have set
the proper @code{default} and @code{default-push} options in
@code{/usr/local/src/xemacs-21.5/.hg/hgrc} so that I can push, pull,
fetch or whatever without any argument on the command line. I also
have set the @code{username} option to my XEmacs identity (that's
mainly for the commit log messages).

When I work on a new feature, I do it in a directory named
@code{/usr/local/src/xemacs-21.5-new-feature}, in which I commit
regularly and occasionally merge with the main repository to keep up
to date. When I'm satisfied with the feature, it's time to integrate
it, but I don't want to import my whole personal history; so I just
apply the final diff to my local trunk (which @emph{must} be up to
date), commit locally with Patcher and then push it up by hand.

Under those conditions, the Patcher setup is quite simple:

@itemize @bullet
@item
Set the @code{commit-privilege} option to @code{t}.
@item
Use the @code{mercurial} or @code{mercurial-ws} built-in theme.
@item
Use the @code{fake} mail method, because messages are now sent to
@code{xemacs-patches} automatically.
@item
Put @code{subject} in the @code{:log-message-items} project options.
This will serve as the summary line, and then you can put whatever
longer description you want on the lines below it.
@end itemize

That's it. Patcher generates the ChangeLog entries and performs the
local commit. I just finish by pushing by hand. Maybe Patcher could
also do that for me one day...



@c ====================================================================
@c Variables Index
@c ====================================================================
@node Variables Index, Functions Index, XEmacs Devel, Top
@comment  node-name,  next,  previous,  up
@unnumbered Variables Index
@printindex vr


@c ====================================================================
@c Functions Index
@c ====================================================================
@node Functions Index,  Keystrokes Index, Variables Index, Top
@comment  node-name,  next,  previous,  up
@unnumbered Functions Index
@printindex fn


@c ====================================================================
@c Keystrokes Index
@c ====================================================================
@node Keystrokes Index,  , Functions Index, Top
@comment  node-name,  next,  previous,  up
@unnumbered Keystrokes Index
@printindex ky


@bye

@c Don't add stuff under this line !!
@c patcher.texi ends here