Make set-auto-mode respect mode: entries at the end of the file (bug#8586)

[bpt/emacs.git] / doc / lispref / modes.texi

@c -*-texinfo-*-
@c This is part of the GNU Emacs Lisp Reference Manual.
@c Copyright (C) 1990-1995, 1998-1999, 2001-2011  Free Software Foundation, Inc.
@c See the file elisp.texi for copying conditions.
@setfilename ../../info/modes
@node Modes, Documentation, Keymaps, Top
@chapter Major and Minor Modes
@cindex mode

  A @dfn{mode} is a set of definitions that customize Emacs and can be
turned on and off while you edit.  There are two varieties of modes:
@dfn{major modes}, which are mutually exclusive and used for editing
particular kinds of text, and @dfn{minor modes}, which provide features
that users can enable individually.

  This chapter describes how to write both major and minor modes, how to
indicate them in the mode line, and how they run hooks supplied by the
user.  For related topics such as keymaps and syntax tables, see
@ref{Keymaps}, and @ref{Syntax Tables}.

@menu
* Hooks::                       How to use hooks; how to write code that provides hooks.
* Major Modes::                 Defining major modes.
* Minor Modes::                 Defining minor modes.
* Mode Line Format::            Customizing the text that appears in the mode line.
* Imenu::                       How a mode can provide a menu
                         of definitions in the buffer.
* Font Lock Mode::              How modes can highlight text according to syntax.
* Auto-Indentation::            How to teach Emacs to indent for a major mode.
* Desktop Save Mode::           How modes can have buffer state saved between
                         Emacs sessions.
@end menu

@node Hooks
@section Hooks
@cindex hooks

  A @dfn{hook} is a variable where you can store a function or functions
to be called on a particular occasion by an existing program.  Emacs
provides hooks for the sake of customization.  Most often, hooks are set
up in the init file (@pxref{Init File}), but Lisp programs can set them also.
@xref{Standard Hooks}, for a list of standard hook variables.

@cindex normal hook
  Most of the hooks in Emacs are @dfn{normal hooks}.  These variables
contain lists of functions to be called with no arguments.  By
convention, whenever the hook name ends in @samp{-hook}, that tells
you it is normal.  We try to make all hooks normal, as much as
possible, so that you can use them in a uniform way.

  Every major mode function is supposed to run a normal hook called
the @dfn{mode hook} as the one of the last steps of initialization.
This makes it easy for a user to customize the behavior of the mode,
by overriding the buffer-local variable assignments already made by
the mode.  Most minor mode functions also run a mode hook at the end.
But hooks are used in other contexts too.  For example, the hook
@code{suspend-hook} runs just before Emacs suspends itself
(@pxref{Suspending Emacs}).

  The recommended way to add a hook function to a normal hook is by
calling @code{add-hook} (see below).  The hook functions may be any of
the valid kinds of functions that @code{funcall} accepts (@pxref{What
Is a Function}).  Most normal hook variables are initially void;
@code{add-hook} knows how to deal with this.  You can add hooks either
globally or buffer-locally with @code{add-hook}.

@cindex abnormal hook
  If the hook variable's name does not end with @samp{-hook}, that
indicates it is probably an @dfn{abnormal hook}.  That means the hook
functions are called with arguments, or their return values are used
in some way.  The hook's documentation says how the functions are
called.  You can use @code{add-hook} to add a function to an abnormal
hook, but you must write the function to follow the hook's calling
convention.

  By convention, abnormal hook names end in @samp{-functions} or
@samp{-hooks}.  If the variable's name ends in @samp{-function}, then
its value is just a single function, not a list of functions.

@menu
* Running Hooks::               How to run a hook.
* Setting Hooks::               How to put functions on a hook, or remove them.
@end menu

@node Running Hooks
@subsection Running Hooks

  At the appropriate times, Emacs uses the @code{run-hooks} function
and the other functions below to run particular hooks.

@defun run-hooks &rest hookvars
This function takes one or more normal hook variable names as
arguments, and runs each hook in turn.  Each argument should be a
symbol that is a normal hook variable.  These arguments are processed
in the order specified.

If a hook variable has a non-@code{nil} value, that value should be a
list of functions.  @code{run-hooks} calls all the functions, one by
one, with no arguments.

The hook variable's value can also be a single function---either a
lambda expression or a symbol with a function definition---which
@code{run-hooks} calls.  But this usage is obsolete.
@end defun

@defun run-hook-with-args hook &rest args
This function is the way to run an abnormal hook and always call all
of the hook functions.  It calls each of the hook functions one by
one, passing each of them the arguments @var{args}.
@end defun

@defun run-hook-with-args-until-failure hook &rest args
This function is the way to run an abnormal hook until one of the hook
functions fails.  It calls each of the hook functions, passing each of
them the arguments @var{args}, until some hook function returns
@code{nil}.  It then stops and returns @code{nil}.  If none of the
hook functions return @code{nil}, it returns a non-@code{nil} value.
@end defun

@defun run-hook-with-args-until-success hook &rest args
This function is the way to run an abnormal hook until a hook function
succeeds.  It calls each of the hook functions, passing each of them
the arguments @var{args}, until some hook function returns
non-@code{nil}.  Then it stops, and returns whatever was returned by
the last hook function that was called.  If all hook functions return
@code{nil}, it returns @code{nil} as well.
@end defun

@node Setting Hooks
@subsection Setting Hooks

  Here's an example that uses a mode hook to turn on Auto Fill mode when
in Lisp Interaction mode:

@example
(add-hook 'lisp-interaction-mode-hook 'turn-on-auto-fill)
@end example

@defun add-hook hook function &optional append local
This function is the handy way to add function @var{function} to hook
variable @var{hook}.  You can use it for abnormal hooks as well as for
normal hooks.  @var{function} can be any Lisp function that can accept
the proper number of arguments for @var{hook}.  For example,

@example
(add-hook 'text-mode-hook 'my-text-hook-function)
@end example

@noindent
adds @code{my-text-hook-function} to the hook called @code{text-mode-hook}.

If @var{function} is already present in @var{hook} (comparing using
@code{equal}), then @code{add-hook} does not add it a second time.

If @var{function} has a non-@code{nil} property
@code{permanent-local-hook}, then @code{kill-all-local-variables} (or
changing major modes) won't delete it from the hook variable's local
value.

It is best to design your hook functions so that the order in which
they are executed does not matter.  Any dependence on the order is
asking for trouble.  However, the order is predictable: normally,
@var{function} goes at the front of the hook list, so it will be
executed first (barring another @code{add-hook} call).  If the
optional argument @var{append} is non-@code{nil}, the new hook
function goes at the end of the hook list and will be executed last.

@code{add-hook} can handle the cases where @var{hook} is void or its
value is a single function; it sets or changes the value to a list of
functions.

If @var{local} is non-@code{nil}, that says to add @var{function} to
the buffer-local hook list instead of to the global hook list.  If
needed, this makes the hook buffer-local and adds @code{t} to the
buffer-local value.  The latter acts as a flag to run the hook
functions in the default value as well as in the local value.
@end defun

@defun remove-hook hook function &optional local
This function removes @var{function} from the hook variable
@var{hook}.  It compares @var{function} with elements of @var{hook}
using @code{equal}, so it works for both symbols and lambda
expressions.

If @var{local} is non-@code{nil}, that says to remove @var{function}
from the buffer-local hook list instead of from the global hook list.
@end defun

@node Major Modes
@section Major Modes
@cindex major mode

  Major modes specialize Emacs for editing particular kinds of text.
Each buffer has only one major mode at a time.  For each major mode
there is a function to switch to that mode in the current buffer; its
name should end in @samp{-mode}.  These functions work by setting
buffer-local variable bindings and other data associated with the
buffer, such as a local keymap.  The effect lasts until you switch
to another major mode in the same buffer.

@menu
* Major Mode Basics::           
* Major Mode Conventions::      Coding conventions for keymaps, etc.
* Auto Major Mode::             How Emacs chooses the major mode automatically.
* Mode Help::                   Finding out how to use a mode.
* Derived Modes::               Defining a new major mode based on another major
                              mode.
* Generic Modes::               Defining a simple major mode that supports
                              comment syntax and Font Lock mode.
* Mode Hooks::                  Hooks run at the end of major mode functions.
* Example Major Modes::         Text mode and Lisp modes.
@end menu

@node Major Mode Basics
@subsection Major Mode Basics
@cindex Fundamental mode

  The least specialized major mode is called @dfn{Fundamental mode}.
This mode has no mode-specific definitions or variable settings, so each
Emacs command behaves in its default manner, and each option is in its
default state.  All other major modes redefine various keys and options.
For example, Lisp Interaction mode provides special key bindings for
@kbd{C-j} (@code{eval-print-last-sexp}), @key{TAB}
(@code{lisp-indent-line}), and other keys.

  When you need to write several editing commands to help you perform a
specialized editing task, creating a new major mode is usually a good
idea.  In practice, writing a major mode is easy (in contrast to
writing a minor mode, which is often difficult).

  If the new mode is similar to an old one, it is often unwise to
modify the old one to serve two purposes, since it may become harder
to use and maintain.  Instead, copy and rename an existing major mode
definition and alter the copy---or use the @code{define-derived-mode}
macro to define a @dfn{derived mode} (@pxref{Derived Modes}).  For
example, Rmail Edit mode is a major mode that is very similar to Text
mode except that it provides two additional commands.  Its definition
is distinct from that of Text mode, but uses that of Text mode.

  Even if the new mode is not an obvious derivative of any other mode,
we recommend to use @code{define-derived-mode}, since it automatically
enforces the most important coding conventions for you.

  For a very simple programming language major mode that handles
comments and fontification, you can use @code{define-generic-mode}.
@xref{Generic Modes}.

  Rmail Edit mode offers an example of changing the major mode
temporarily for a buffer, so it can be edited in a different way (with
ordinary Emacs commands rather than Rmail commands).  In such cases, the
temporary major mode usually provides a command to switch back to the
buffer's usual mode (Rmail mode, in this case).  You might be tempted to
present the temporary redefinitions inside a recursive edit and restore
the usual ones when the user exits; but this is a bad idea because it
constrains the user's options when it is done in more than one buffer:
recursive edits must be exited most-recently-entered first.  Using an
alternative major mode avoids this limitation.  @xref{Recursive
Editing}.

  The standard GNU Emacs Lisp library directory tree contains the code
for several major modes, in files such as @file{text-mode.el},
@file{texinfo.el}, @file{lisp-mode.el}, @file{c-mode.el}, and
@file{rmail.el}.  They are found in various subdirectories of the
@file{lisp} directory.  You can study these libraries to see how modes
are written.  Text mode is perhaps the simplest major mode aside from
Fundamental mode.  Rmail mode is a complicated and specialized mode.

@node Major Mode Conventions
@subsection Major Mode Conventions
@cindex major mode conventions
@cindex conventions for writing major modes

  The code for existing major modes follows various coding conventions,
including conventions for local keymap and syntax table initialization,
global names, and hooks.  Please follow these conventions when you
define a new major mode.  (Fundamental mode is an exception to many
of these conventions, because its definition is to present the global
state of Emacs.)

  This list of conventions is only partial, because each major mode
should aim for consistency in general with other Emacs major modes.
This makes Emacs as a whole more coherent.  It is impossible to list
here all the possible points where this issue might come up; if the
Emacs developers point out an area where your major mode deviates from
the usual conventions, please make it compatible.

@itemize @bullet
@item
Define a command whose name ends in @samp{-mode}, with no arguments,
that switches to the new mode in the current buffer.  This command
should set up the keymap, syntax table, and buffer-local variables in an
existing buffer, without changing the buffer's contents.

@item
Write a documentation string for this command that describes the
special commands available in this mode.  @kbd{C-h m}
(@code{describe-mode}) in your mode will display this string.

The documentation string may include the special documentation
substrings, @samp{\[@var{command}]}, @samp{\@{@var{keymap}@}}, and
@samp{\<@var{keymap}>}, which enable the documentation to adapt
automatically to the user's own key bindings.  @xref{Keys in
Documentation}.

@item
The major mode command should start by calling
@code{kill-all-local-variables}.  This runs the normal hook
@code{change-major-mode-hook}, then gets rid of the buffer-local
variables of the major mode previously in effect.  @xref{Creating
Buffer-Local}.

@item
The major mode command should set the variable @code{major-mode} to the
major mode command symbol.  This is how @code{describe-mode} discovers
which documentation to print.

@item
The major mode command should set the variable @code{mode-name} to the
``pretty'' name of the mode, usually a string (but see @ref{Mode Line
Data}, for other possible forms).  The name of the mode appears
in the mode line.

@item
@cindex functions in modes
Since all global names are in the same name space, all the global
variables, constants, and functions that are part of the mode should
have names that start with the major mode name (or with an abbreviation
of it if the name is long).  @xref{Coding Conventions}.

@item
In a major mode for editing some kind of structured text, such as a
programming language, indentation of text according to structure is
probably useful.  So the mode should set @code{indent-line-function}
to a suitable function, and probably customize other variables
for indentation.  @xref{Auto-Indentation}.

@item
@cindex keymaps in modes
The major mode should usually have its own keymap, which is used as the
local keymap in all buffers in that mode.  The major mode command should
call @code{use-local-map} to install this local map.  @xref{Active
Keymaps}, for more information.

This keymap should be stored permanently in a global variable named
@code{@var{modename}-mode-map}.  Normally the library that defines the
mode sets this variable.

@xref{Tips for Defining}, for advice about how to write the code to set
up the mode's keymap variable.

@item
The key sequences bound in a major mode keymap should usually start with
@kbd{C-c}, followed by a control character, a digit, or @kbd{@{},
@kbd{@}}, @kbd{<}, @kbd{>}, @kbd{:} or @kbd{;}.  The other punctuation
characters are reserved for minor modes, and ordinary letters are
reserved for users.

A major mode can also rebind the keys @kbd{M-n}, @kbd{M-p} and
@kbd{M-s}.  The bindings for @kbd{M-n} and @kbd{M-p} should normally
be some kind of ``moving forward and backward,'' but this does not
necessarily mean cursor motion.

It is legitimate for a major mode to rebind a standard key sequence if
it provides a command that does ``the same job'' in a way better
suited to the text this mode is used for.  For example, a major mode
for editing a programming language might redefine @kbd{C-M-a} to
``move to the beginning of a function'' in a way that works better for
that language.

It is also legitimate for a major mode to rebind a standard key
sequence whose standard meaning is rarely useful in that mode.  For
instance, minibuffer modes rebind @kbd{M-r}, whose standard meaning is
rarely of any use in the minibuffer.  Major modes such as Dired or
Rmail that do not allow self-insertion of text can reasonably redefine
letters and other printing characters as special commands.

@item
Major modes for editing text should not define @key{RET} to do
anything other than insert a newline.  However, it is ok for
specialized modes for text that users don't directly edit, such as
Dired and Info modes, to redefine @key{RET} to do something entirely
different.

@item
Major modes should not alter options that are primarily a matter of user
preference, such as whether Auto-Fill mode is enabled.  Leave this to
each user to decide.  However, a major mode should customize other
variables so that Auto-Fill mode will work usefully @emph{if} the user
decides to use it.

@item
@cindex syntax tables in modes
The mode may have its own syntax table or may share one with other
related modes.  If it has its own syntax table, it should store this in
a variable named @code{@var{modename}-mode-syntax-table}.  @xref{Syntax
Tables}.

@item
If the mode handles a language that has a syntax for comments, it should
set the variables that define the comment syntax.  @xref{Options for
Comments,, Options Controlling Comments, emacs, The GNU Emacs Manual}.

@item
@cindex abbrev tables in modes
The mode may have its own abbrev table or may share one with other
related modes.  If it has its own abbrev table, it should store this
in a variable named @code{@var{modename}-mode-abbrev-table}.  If the
major mode command defines any abbrevs itself, it should pass @code{t}
for the @var{system-flag} argument to @code{define-abbrev}.
@xref{Defining Abbrevs}.

@item
The mode should specify how to do highlighting for Font Lock mode, by
setting up a buffer-local value for the variable
@code{font-lock-defaults} (@pxref{Font Lock Mode}).

@item
Each face that the mode defines should, if possible, inherit from an
existing Emacs face.  This reduces the chance of conflicting with a
user's face customizations.  Useful faces include:

@table @asis
@item @code{highlight}
for stretches of text that should temporarily stand out.

@item @code{match}
for text matching a search command.

@item @code{link} and @code{link-visited}
for clickable text buttons that send the user to a different buffer or
``location''.

@item @code{button}
for clickable text buttons that perform other actions.

@item @asis{Font Lock faces}
for other kinds of syntactic highlighting, if highlighting is not
handled by Font Lock mode or some Font Lock faces are not in use.
@xref{Faces for Font Lock}, for how to assign Font Lock faces.
@end table

@item
The mode should specify how Imenu should find the definitions or
sections of a buffer, by setting up a buffer-local value for the
variable @code{imenu-generic-expression}, for the two variables
@code{imenu-prev-index-position-function} and
@code{imenu-extract-index-name-function}, or for the variable
@code{imenu-create-index-function} (@pxref{Imenu}).

@item
The mode can specify a local value for
@code{eldoc-documentation-function} to tell ElDoc mode how to handle
this mode.

@item
The mode can specify how to complete various keywords by adding
to the special hook @code{completion-at-point-functions}.

@item
Use @code{defvar} or @code{defcustom} to set mode-related variables, so
that they are not reinitialized if they already have a value.  (Such
reinitialization could discard customizations made by the user.)

@item
@cindex buffer-local variables in modes
To make a buffer-local binding for an Emacs customization variable, use
@code{make-local-variable} in the major mode command, not
@code{make-variable-buffer-local}.  The latter function would make the
variable local to every buffer in which it is subsequently set, which
would affect buffers that do not use this mode.  It is undesirable for a
mode to have such global effects.  @xref{Buffer-Local Variables}.

With rare exceptions, the only reasonable way to use
@code{make-variable-buffer-local} in a Lisp package is for a variable
which is used only within that package.  Using it on a variable used by
other packages would interfere with them.

@item
@cindex mode hook
@cindex major mode hook
Each major mode should have a normal @dfn{mode hook} named
@code{@var{modename}-mode-hook}.  The very last thing the major mode command
should do is to call @code{run-mode-hooks}.  This runs the mode hook,
and then runs the normal hook @code{after-change-major-mode-hook}.
@xref{Mode Hooks}.

@item
The major mode command may start by calling some other major mode
command (called the @dfn{parent mode}) and then alter some of its
settings.  A mode that does this is called a @dfn{derived mode}.  The
recommended way to define one is to use the @code{define-derived-mode}
macro, but this is not required.  Such a mode should call the parent
mode command inside a @code{delay-mode-hooks} form.  (Using
@code{define-derived-mode} does this automatically.)  @xref{Derived
Modes}, and @ref{Mode Hooks}.

@item
If something special should be done if the user switches a buffer from
this mode to any other major mode, this mode can set up a buffer-local
value for @code{change-major-mode-hook} (@pxref{Creating Buffer-Local}).

@item
If this mode is appropriate only for specially-prepared text, then the
major mode command symbol should have a property named @code{mode-class}
with value @code{special}, put on as follows:

@kindex mode-class @r{(property)}
@cindex @code{special}
@example
(put 'funny-mode 'mode-class 'special)
@end example

@noindent
This tells Emacs that new buffers created while the current buffer is
in Funny mode should not inherit Funny mode, in case the default value
of @code{major-mode} is @code{nil}.  Modes such as Dired, Rmail,
and Buffer List use this feature.

The @code{define-derived-mode} macro automatically marks the derived
mode as special if the parent mode is special.  The special mode
@code{special-mode} provides a convenient parent for other special
modes to inherit from; it sets @code{buffer-read-only} to @code{t},
and does little else.

@item
If you want to make the new mode the default for files with certain
recognizable names, add an element to @code{auto-mode-alist} to select
the mode for those file names (@pxref{Auto Major Mode}).  If you
define the mode command to autoload, you should add this element in
the same file that calls @code{autoload}.  If you use an autoload
cookie for the mode command, you can also use an autoload cookie for
the form that adds the element (@pxref{autoload cookie}).  If you do
not autoload the mode command, it is sufficient to add the element in
the file that contains the mode definition.

@item
In the comments that document the file, you should provide a sample
@code{autoload} form and an example of how to add to
@code{auto-mode-alist}, that users can include in their init files
(@pxref{Init File}).

@item
@cindex mode loading
The top-level forms in the file defining the mode should be written so
that they may be evaluated more than once without adverse consequences.
Even if you never load the file more than once, someone else will.
@end itemize

@node Auto Major Mode
@subsection How Emacs Chooses a Major Mode
@cindex major mode, automatic selection

  Based on information in the file name or in the file itself, Emacs
automatically selects a major mode for the new buffer when a file is
visited.  It also processes local variables specified in the file text.

@deffn Command fundamental-mode
  Fundamental mode is a major mode that is not specialized for anything
in particular.  Other major modes are defined in effect by comparison
with this one---their definitions say what to change, starting from
Fundamental mode.  The @code{fundamental-mode} function does @emph{not}
run any mode hooks; you're not supposed to customize it.  (If you want Emacs
to behave differently in Fundamental mode, change the @emph{global}
state of Emacs.)
@end deffn

@deffn Command normal-mode &optional find-file
This function establishes the proper major mode and buffer-local variable
bindings for the current buffer.  First it calls @code{set-auto-mode}
(see below), then it runs @code{hack-local-variables} to parse, and
bind or evaluate as appropriate, the file's local variables
(@pxref{File Local Variables}).

If the @var{find-file} argument to @code{normal-mode} is non-@code{nil},
@code{normal-mode} assumes that the @code{find-file} function is calling
it.  In this case, it may process local variables in the @samp{-*-}
line or at the end of the file.  The variable
@code{enable-local-variables} controls whether to do so.  @xref{File
Variables, , Local Variables in Files, emacs, The GNU Emacs Manual},
for the syntax of the local variables section of a file.

If you run @code{normal-mode} interactively, the argument
@var{find-file} is normally @code{nil}.  In this case,
@code{normal-mode} unconditionally processes any file local variables.

The function calls @code{set-auto-mode} to choose a major mode.  If it
does not specify a mode, the buffer stays in the major mode determined
by the default value of @code{major-mode} (see below).

@cindex file mode specification error
@code{normal-mode} uses @code{condition-case} around the call to the
major mode function, so errors are caught and reported as a @samp{File
mode specification error},  followed by the original error message.
@end deffn

@defun set-auto-mode &optional keep-mode-if-same
@cindex visited file mode
  This function selects the major mode that is appropriate for the
current buffer.  It bases its decision (in order of precedence) on
the @w{@samp{-*-}} line, on any @samp{mode:} local variable near the
end of a file, on the @w{@samp{#!}} line (using
@code{interpreter-mode-alist}), on the text at the beginning of the
buffer (using @code{magic-mode-alist}), and finally on the visited
file name (using @code{auto-mode-alist}).  @xref{Choosing Modes, , How
Major Modes are Chosen, emacs, The GNU Emacs Manual}. 
If @code{enable-local-variables} is @code{nil}, @code{set-auto-mode}
does not check the @w{@samp{-*-}} line, or near the end of the file,
for any mode tag.

If @var{keep-mode-if-same} is non-@code{nil}, this function does not
call the mode command if the buffer is already in the proper major
mode.  For instance, @code{set-visited-file-name} sets this to
@code{t} to avoid killing buffer local variables that the user may
have set.
@end defun

@defopt major-mode
The buffer-local value of this variable holds the major mode
currently active.  The default value of this variable holds the
default major mode for new buffers.  The standard default value is
@code{fundamental-mode}.

If the default value of @code{major-mode} is @code{nil}, Emacs uses
the (previously) current buffer's major mode as the default major mode
of a new buffer.  However, if that major mode symbol has a @code{mode-class}
property with value @code{special}, then it is not used for new buffers;
Fundamental mode is used instead.  The modes that have this property are
those such as Dired and Rmail that are useful only with text that has
been specially prepared.
@end defopt

@defun set-buffer-major-mode buffer
This function sets the major mode of @var{buffer} to the default value of
@code{major-mode}; if that is @code{nil}, it uses the
current buffer's major mode (if that is suitable).  As an exception,
if @var{buffer}'s name is @samp{*scratch*}, it sets the mode to
@code{initial-major-mode}.

The low-level primitives for creating buffers do not use this function,
but medium-level commands such as @code{switch-to-buffer} and
@code{find-file-noselect} use it whenever they create buffers.
@end defun

@defopt initial-major-mode
@cindex @samp{*scratch*}
The value of this variable determines the major mode of the initial
@samp{*scratch*} buffer.  The value should be a symbol that is a major
mode command.  The default value is @code{lisp-interaction-mode}.
@end defopt

@defvar interpreter-mode-alist
This variable specifies major modes to use for scripts that specify a
command interpreter in a @samp{#!} line.  Its value is an alist with
elements of the form @code{(@var{interpreter} . @var{mode})}; for
example, @code{("perl" . perl-mode)} is one element present by
default.  The element says to use mode @var{mode} if the file
specifies an interpreter which matches @var{interpreter}.
@end defvar

@defvar magic-mode-alist
This variable's value is an alist with elements of the form
@code{(@var{regexp} .  @var{function})}, where @var{regexp} is a
regular expression and @var{function} is a function or @code{nil}.
After visiting a file, @code{set-auto-mode} calls @var{function} if
the text at the beginning of the buffer matches @var{regexp} and
@var{function} is non-@code{nil}; if @var{function} is @code{nil},
@code{auto-mode-alist} gets to decide the mode.
@end defvar

@defvar magic-fallback-mode-alist
This works like @code{magic-mode-alist}, except that it is handled
only if @code{auto-mode-alist} does not specify a mode for this file.
@end defvar

@defvar auto-mode-alist
This variable contains an association list of file name patterns
(regular expressions) and corresponding major mode commands.  Usually,
the file name patterns test for suffixes, such as @samp{.el} and
@samp{.c}, but this need not be the case.  An ordinary element of the
alist looks like @code{(@var{regexp} .  @var{mode-function})}.

For example,

@smallexample
@group
(("\\`/tmp/fol/" . text-mode)
 ("\\.texinfo\\'" . texinfo-mode)
 ("\\.texi\\'" . texinfo-mode)
@end group
@group
 ("\\.el\\'" . emacs-lisp-mode)
 ("\\.c\\'" . c-mode)
 ("\\.h\\'" . c-mode)
 @dots{})
@end group
@end smallexample

When you visit a file whose expanded file name (@pxref{File Name
Expansion}), with version numbers and backup suffixes removed using
@code{file-name-sans-versions} (@pxref{File Name Components}), matches
a @var{regexp}, @code{set-auto-mode} calls the corresponding
@var{mode-function}.  This feature enables Emacs to select the proper
major mode for most files.

If an element of @code{auto-mode-alist} has the form @code{(@var{regexp}
@var{function} t)}, then after calling @var{function}, Emacs searches
@code{auto-mode-alist} again for a match against the portion of the file
name that did not match before.  This feature is useful for
uncompression packages: an entry of the form @code{("\\.gz\\'"
@var{function} t)} can uncompress the file and then put the uncompressed
file in the proper mode according to the name sans @samp{.gz}.

Here is an example of how to prepend several pattern pairs to
@code{auto-mode-alist}.  (You might use this sort of expression in your
init file.)

@smallexample
@group
(setq auto-mode-alist
  (append
   ;; @r{File name (within directory) starts with a dot.}
   '(("/\\.[^/]*\\'" . fundamental-mode)
     ;; @r{File name has no dot.}
     ("/[^\\./]*\\'" . fundamental-mode)
     ;; @r{File name ends in @samp{.C}.}
     ("\\.C\\'" . c++-mode))
   auto-mode-alist))
@end group
@end smallexample
@end defvar

@node Mode Help
@subsection Getting Help about a Major Mode
@cindex mode help
@cindex help for major mode
@cindex documentation for major mode

  The @code{describe-mode} function is used to provide information
about major modes.  It is normally called with @kbd{C-h m}.  The
@code{describe-mode} function uses the value of @code{major-mode},
which is why every major mode function needs to set the
@code{major-mode} variable.

@deffn Command describe-mode
This function displays the documentation of the current major mode.

The @code{describe-mode} function calls the @code{documentation}
function using the value of @code{major-mode} as an argument.  Thus, it
displays the documentation string of the major mode function.
(@xref{Accessing Documentation}.)
@end deffn

@defvar major-mode
This buffer-local variable holds the symbol for the current buffer's
major mode.  This symbol should have a function definition that is the
command to switch to that major mode.  The @code{describe-mode}
function uses the documentation string of the function as the
documentation of the major mode.
@end defvar

@node Derived Modes
@subsection Defining Derived Modes
@cindex derived mode

  The recommended way to define a new major mode is to derive it
from an existing one using @code{define-derived-mode}.  If there is no
closely related mode, you can inherit from @code{text-mode},
@code{special-mode}, @code{prog-mode}, or in the worst case
@code{fundamental-mode}.

@defmac define-derived-mode variant parent name docstring keyword-args@dots{} body@dots{}
This macro defines @var{variant} as a major mode command, using
@var{name} as the string form of the mode name.  @var{variant} and
@var{parent} should be unquoted symbols.

The new command @var{variant} is defined to call the function
@var{parent}, then override certain aspects of that parent mode:

@itemize @bullet
@item
The new mode has its own sparse keymap, named
@code{@var{variant}-map}.  @code{define-derived-mode}
makes the parent mode's keymap the parent of the new map, unless
@code{@var{variant}-map} is already set and already has a parent.

@item
The new mode has its own syntax table, kept in the variable
@code{@var{variant}-syntax-table}, unless you override this using the
@code{:syntax-table} keyword (see below).  @code{define-derived-mode}
makes the parent mode's syntax-table the parent of
@code{@var{variant}-syntax-table}, unless the latter is already set
and already has a parent different from the standard syntax table.

@item
The new mode has its own abbrev table, kept in the variable
@code{@var{variant}-abbrev-table}, unless you override this using the
@code{:abbrev-table} keyword (see below).

@item
The new mode has its own mode hook, @code{@var{variant}-hook}.  It
runs this hook, after running the hooks of its ancestor modes, with
@code{run-mode-hooks}, as the last thing it does. @xref{Mode Hooks}.
@end itemize

In addition, you can specify how to override other aspects of
@var{parent} with @var{body}.  The command @var{variant}
evaluates the forms in @var{body} after setting up all its usual
overrides, just before running the mode hooks.

If @var{parent} has a non-@code{nil} @code{mode-class} symbol
property, then @code{define-derived-mode} sets the @code{mode-class}
property of @var{variant} to the same value.  This ensures, for
example, that if @var{parent} is a special mode, then @var{variant} is
also a special mode (@pxref{Major Mode Conventions}).

You can also specify @code{nil} for @var{parent}.  This gives the new
mode no parent.  Then @code{define-derived-mode} behaves as described
above, but, of course, omits all actions connected with @var{parent}.

The argument @var{docstring} specifies the documentation string for
the new mode.  @code{define-derived-mode} adds some general
information about the mode's hook, followed by the mode's keymap, at
the end of this docstring.  If you omit @var{docstring},
@code{define-derived-mode} generates a documentation string.

The @var{keyword-args} are pairs of keywords and values.  The values
are evaluated.  The following keywords are currently supported:

@table @code
@item :syntax-table
You can use this to explicitly specify a syntax table for the new
mode.  If you specify a @code{nil} value, the new mode uses the same
syntax table as @var{parent}, or the standard syntax table if
@var{parent} is @code{nil}.  (Note that this does @emph{not} follow
the convention used for non-keyword arguments that a @code{nil} value
is equivalent with not specifying the argument.)

@item :abbrev-table
You can use this to explicitly specify an abbrev table for the new
mode.  If you specify a @code{nil} value, the new mode uses the same
abbrev table as @var{parent}, or @code{fundamental-mode-abbrev-table}
if @var{parent} is @code{nil}.  (Again, a @code{nil} value is
@emph{not} equivalent to not specifying this keyword.)

@item :group
If this is specified, the value should be the customization group for
this mode.  (Not all major modes have one.)  Only the (still
experimental and unadvertised) command @code{customize-mode} currently
uses this.  @code{define-derived-mode} does @emph{not} automatically
define the specified customization group.
@end table

Here is a hypothetical example:

@example
(define-derived-mode hypertext-mode
  text-mode "Hypertext"
  "Major mode for hypertext.
\\@{hypertext-mode-map@}"
  (setq case-fold-search nil))

(define-key hypertext-mode-map
  [down-mouse-3] 'do-hyper-link)
@end example

Do not write an @code{interactive} spec in the definition;
@code{define-derived-mode} does that automatically.
@end defmac

@node Generic Modes
@subsection Generic Modes
@cindex generic mode

  @dfn{Generic modes} are simple major modes with basic support for
comment syntax and Font Lock mode.  To define a generic mode, use the
macro @code{define-generic-mode}.  See the file @file{generic-x.el}
for some examples of the use of @code{define-generic-mode}.

@defmac define-generic-mode mode comment-list keyword-list font-lock-list auto-mode-list function-list &optional docstring
This macro defines a generic mode command named @var{mode} (a symbol,
not quoted).  The optional argument @var{docstring} is the
documentation for the mode command.  If you do not supply it,
@code{define-generic-mode} generates one by default.

The argument @var{comment-list} is a list in which each element is
either a character, a string of one or two characters, or a cons cell.
A character or a string is set up in the mode's syntax table as a
``comment starter.''  If the entry is a cons cell, the @sc{car} is set
up as a ``comment starter'' and the @sc{cdr} as a ``comment ender.''
(Use @code{nil} for the latter if you want comments to end at the end
of the line.)  Note that the syntax table mechanism has limitations
about what comment starters and enders are actually possible.
@xref{Syntax Tables}.

The argument @var{keyword-list} is a list of keywords to highlight
with @code{font-lock-keyword-face}.  Each keyword should be a string.
Meanwhile, @var{font-lock-list} is a list of additional expressions to
highlight.  Each element of this list should have the same form as an
element of @code{font-lock-keywords}.  @xref{Search-based
Fontification}.

The argument @var{auto-mode-list} is a list of regular expressions to
add to the variable @code{auto-mode-alist}.  They are added by the execution
of the @code{define-generic-mode} form, not by expanding the macro call.

Finally, @var{function-list} is a list of functions for the mode
command to call for additional setup.  It calls these functions just
before it runs the mode hook variable @code{@var{mode}-hook}.
@end defmac

@node Mode Hooks
@subsection Mode Hooks

  Every major mode function should finish by running its mode hook and
the mode-independent normal hook @code{after-change-major-mode-hook}.
It does this by calling @code{run-mode-hooks}.  If the major mode is a
derived mode, that is if it calls another major mode (the parent mode)
in its body, it should do this inside @code{delay-mode-hooks} so that
the parent won't run these hooks itself.  Instead, the derived mode's
call to @code{run-mode-hooks} runs the parent's mode hook too.
@xref{Major Mode Conventions}.

  Emacs versions before Emacs 22 did not have @code{delay-mode-hooks}.
When user-implemented major modes have not been updated to use it,
they won't entirely follow these conventions: they may run the
parent's mode hook too early, or fail to run
@code{after-change-major-mode-hook}.  If you encounter such a major
mode, please correct it to follow these conventions.

  When you defined a major mode using @code{define-derived-mode}, it
automatically makes sure these conventions are followed.  If you
define a major mode ``by hand,'' not using @code{define-derived-mode},
use the following functions to handle these conventions automatically.

@defun run-mode-hooks &rest hookvars
Major modes should run their mode hook using this function.  It is
similar to @code{run-hooks} (@pxref{Hooks}), but it also runs
@code{after-change-major-mode-hook}.

When this function is called during the execution of a
@code{delay-mode-hooks} form, it does not run the hooks immediately.
Instead, it arranges for the next call to @code{run-mode-hooks} to run
them.
@end defun

@defmac delay-mode-hooks body@dots{}
When one major mode command calls another, it should do so inside of
@code{delay-mode-hooks}.

This macro executes @var{body}, but tells all @code{run-mode-hooks}
calls during the execution of @var{body} to delay running their hooks.
The hooks will actually run during the next call to
@code{run-mode-hooks} after the end of the @code{delay-mode-hooks}
construct.
@end defmac

@defvar after-change-major-mode-hook
This is a normal hook run by @code{run-mode-hooks}.  It is run at the
very end of every properly-written major mode function.
@end defvar

@node Example Major Modes
@subsection Major Mode Examples

  Text mode is perhaps the simplest mode besides Fundamental mode.
Here are excerpts from  @file{text-mode.el} that illustrate many of
the conventions listed above:

@smallexample
@group
;; @r{Create the syntax table for this mode.}
(defvar text-mode-syntax-table
  (let ((st (make-syntax-table)))
    (modify-syntax-entry ?\" ".   " st)
    (modify-syntax-entry ?\\ ".   " st)
    ;; Add `p' so M-c on `hello' leads to `Hello', not `hello'.
    (modify-syntax-entry ?' "w p" st)
    st)
  "Syntax table used while in `text-mode'.")
@end group

;; @r{Create the keymap for this mode.}
@group
(defvar text-mode-map
  (let ((map (make-sparse-keymap)))
    (define-key map "\e\t" 'ispell-complete-word)
    (define-key map "\es" 'center-line)
    (define-key map "\eS" 'center-paragraph)
    map)
  "Keymap for `text-mode'.
Many other modes, such as Mail mode, Outline mode
and Indented Text mode, inherit all the commands
defined in this map.")
@end group
@end smallexample

  Here is how the actual mode command is defined now:

@smallexample
@group
(define-derived-mode text-mode nil "Text"
  "Major mode for editing text written for humans to read.
In this mode, paragraphs are delimited only by blank or white lines.
You can thus get the full benefit of adaptive filling
 (see the variable `adaptive-fill-mode').
\\@{text-mode-map@}
Turning on Text mode runs the normal hook `text-mode-hook'."
@end group
@group
  (set (make-local-variable 'text-mode-variant) t)
  ;; @r{These two lines are a feature added recently.}
  (set (make-local-variable 'require-final-newline)
       mode-require-final-newline)
  (set (make-local-variable 'indent-line-function) 'indent-relative))
@end group
@end smallexample

@noindent
(The last line is redundant nowadays, since @code{indent-relative} is
the default value, and we'll delete it in a future version.)

  Here is how it was defined formerly, before
@code{define-derived-mode} existed:

@smallexample
@group
;; @r{This isn't needed nowadays, since @code{define-derived-mode} does it.}
(define-abbrev-table 'text-mode-abbrev-table ()
  "Abbrev table used while in text mode.")
@end group

@group
(defun text-mode ()
  "Major mode for editing text intended for humans to read...
 Special commands: \\@{text-mode-map@}
@end group
@group
Turning on text-mode runs the hook `text-mode-hook'."
  (interactive)
  (kill-all-local-variables)
  (use-local-map text-mode-map)
@end group
@group
  (setq local-abbrev-table text-mode-abbrev-table)
  (set-syntax-table text-mode-syntax-table)
@end group
@group
  ;; @r{These four lines are absent from the current version}
  ;; @r{not because this is done some other way, but rather}
  ;; @r{because nowadays Text mode uses the normal definition of paragraphs.}
  (set (make-local-variable 'paragraph-start)
       (concat "[ \t]*$\\|" page-delimiter))
  (set (make-local-variable 'paragraph-separate) paragraph-start)
  (set (make-local-variable 'indent-line-function) 'indent-relative-maybe)
@end group
@group
  (setq mode-name "Text")
  (setq major-mode 'text-mode)
  (run-mode-hooks 'text-mode-hook)) ; @r{Finally, this permits the user to}
                                    ;   @r{customize the mode with a hook.}
@end group
@end smallexample

@cindex @file{lisp-mode.el}
  The three Lisp modes (Lisp mode, Emacs Lisp mode, and Lisp
Interaction mode) have more features than Text mode and the code is
correspondingly more complicated.  Here are excerpts from
@file{lisp-mode.el} that illustrate how these modes are written.

@cindex syntax table example
@smallexample
@group
;; @r{Create mode-specific table variables.}
(defvar lisp-mode-syntax-table nil "")
(defvar lisp-mode-abbrev-table nil "")
@end group

@group
(defvar emacs-lisp-mode-syntax-table
  (let ((table (make-syntax-table)))
    (let ((i 0))
@end group

@group
      ;; @r{Set syntax of chars up to @samp{0} to say they are}
      ;;   @r{part of symbol names but not words.}
      ;;   @r{(The digit @samp{0} is @code{48} in the @acronym{ASCII} character set.)}
      (while (< i ?0)
        (modify-syntax-entry i "_   " table)
        (setq i (1+ i)))
      ;; @r{@dots{} similar code follows for other character ranges.}
@end group
@group
      ;; @r{Then set the syntax codes for characters that are special in Lisp.}
      (modify-syntax-entry ?  "    " table)
      (modify-syntax-entry ?\t "    " table)
      (modify-syntax-entry ?\f "    " table)
      (modify-syntax-entry ?\n ">   " table)
@end group
@group
      ;; @r{Give CR the same syntax as newline, for selective-display.}
      (modify-syntax-entry ?\^m ">   " table)
      (modify-syntax-entry ?\; "<   " table)
      (modify-syntax-entry ?` "'   " table)
      (modify-syntax-entry ?' "'   " table)
      (modify-syntax-entry ?, "'   " table)
@end group
@group
      ;; @r{@dots{}likewise for many other characters@dots{}}
      (modify-syntax-entry ?\( "()  " table)
      (modify-syntax-entry ?\) ")(  " table)
      (modify-syntax-entry ?\[ "(]  " table)
      (modify-syntax-entry ?\] ")[  " table))
    table))
@end group
@group
;; @r{Create an abbrev table for lisp-mode.}
(define-abbrev-table 'lisp-mode-abbrev-table ())
@end group
@end smallexample

  The three modes for Lisp share much of their code.  For instance,
each calls the following function to set various variables:

@smallexample
@group
(defun lisp-mode-variables (lisp-syntax)
  (when lisp-syntax
    (set-syntax-table lisp-mode-syntax-table))
  (setq local-abbrev-table lisp-mode-abbrev-table)
  @dots{}
@end group
@end smallexample

  In Lisp and most programming languages, we want the paragraph
commands to treat only blank lines as paragraph separators.  And the
modes should understand the Lisp conventions for comments.  The rest of
@code{lisp-mode-variables} sets this up:

@smallexample
@group
  (set (make-local-variable 'paragraph-start) (concat page-delimiter "\\|$" ))
  (set (make-local-variable 'paragraph-separate) paragraph-start)
  @dots{}
@end group
@group
  (set (make-local-variable 'comment-indent-function) 'lisp-comment-indent))
  @dots{}
@end group
@end smallexample

  Each of the different Lisp modes has a slightly different keymap.  For
example, Lisp mode binds @kbd{C-c C-z} to @code{run-lisp}, but the other
Lisp modes do not.  However, all Lisp modes have some commands in
common.  The following code sets up the common commands:

@smallexample
@group
(defvar shared-lisp-mode-map
  (let ((map (make-sparse-keymap)))
    (define-key shared-lisp-mode-map "\e\C-q" 'indent-sexp)
    (define-key shared-lisp-mode-map "\177"
                'backward-delete-char-untabify)
    map)
  "Keymap for commands shared by all sorts of Lisp modes.")
@end group
@end smallexample

@noindent
And here is the code to set up the keymap for Lisp mode:

@smallexample
@group
(defvar lisp-mode-map
  (let ((map (make-sparse-keymap)))
    (set-keymap-parent map shared-lisp-mode-map)
    (define-key map "\e\C-x" 'lisp-eval-defun)
    (define-key map "\C-c\C-z" 'run-lisp)
    map)
  "Keymap for ordinary Lisp mode...")
@end group
@end smallexample

  Finally, here is the complete major mode function definition for
Lisp mode.

@smallexample
@group
(defun lisp-mode ()
  "Major mode for editing Lisp code for Lisps other than GNU Emacs Lisp.
Commands:
Delete converts tabs to spaces as it moves back.
Blank lines separate paragraphs.  Semicolons start comments.
\\@{lisp-mode-map@}
Note that `run-lisp' may be used either to start an inferior Lisp job
or to switch back to an existing one.
@end group

@group
Entry to this mode calls the value of `lisp-mode-hook'
if that value is non-nil."
  (interactive)
  (kill-all-local-variables)
@end group
@group
  (use-local-map lisp-mode-map)          ; @r{Select the mode's keymap.}
  (setq major-mode 'lisp-mode)           ; @r{This is how @code{describe-mode}}
                                         ;   @r{finds out what to describe.}
  (setq mode-name "Lisp")                ; @r{This goes into the mode line.}
  (lisp-mode-variables t)                ; @r{This defines various variables.}
  (set (make-local-variable 'comment-start-skip)
       "\\(\\(^\\|[^\\\\\n]\\)\\(\\\\\\\\\\)*\\)\\(;+\\|#|\\) *")
  (set (make-local-variable 'font-lock-keywords-case-fold-search) t)
@end group
@group
  (setq imenu-case-fold-search t)
  (set-syntax-table lisp-mode-syntax-table)
  (run-mode-hooks 'lisp-mode-hook))      ; @r{This permits the user to use a}
                                         ;   @r{hook to customize the mode.}
@end group
@end smallexample

@node Minor Modes
@section Minor Modes
@cindex minor mode

  A @dfn{minor mode} provides features that users may enable or disable
independently of the choice of major mode.  Minor modes can be enabled
individually or in combination.  Minor modes would be better named
``generally available, optional feature modes,'' except that such a name
would be unwieldy.

  A minor mode is not usually meant as a variation of a single major mode.
Usually they are general and can apply to many major modes.  For
example, Auto Fill mode works with any major mode that permits text
insertion.  To be general, a minor mode must be effectively independent
of the things major modes do.

  A minor mode is often much more difficult to implement than a major
mode.  One reason is that you should be able to activate and deactivate
minor modes in any order.  A minor mode should be able to have its
desired effect regardless of the major mode and regardless of the other
minor modes in effect.

  Often the biggest problem in implementing a minor mode is finding a
way to insert the necessary hook into the rest of Emacs.  Minor mode
keymaps make this easier than it used to be.

@defvar minor-mode-list
The value of this variable is a list of all minor mode commands.
@end defvar

@menu
* Minor Mode Conventions::      Tips for writing a minor mode.
* Keymaps and Minor Modes::     How a minor mode can have its own keymap.
* Defining Minor Modes::        A convenient facility for defining minor modes.
@end menu

@node Minor Mode Conventions
@subsection Conventions for Writing Minor Modes
@cindex minor mode conventions
@cindex conventions for writing minor modes

  There are conventions for writing minor modes just as there are for
major modes.  Several of the major mode conventions apply to minor
modes as well: those regarding the name of the mode initialization
function, the names of global symbols, the use of a hook at the end of
the initialization function, and the use of keymaps and other tables.

  In addition, there are several conventions that are specific to
minor modes.  (The easiest way to follow all the conventions is to use
the macro @code{define-minor-mode}; @ref{Defining Minor Modes}.)

@itemize @bullet
@item
@cindex mode variable
Make a variable whose name ends in @samp{-mode} to control the minor
mode.  We call this the @dfn{mode variable}.  The minor mode command
should set this variable (@code{nil} to disable; anything else to
enable).

If possible, implement the mode so that setting the variable
automatically enables or disables the mode.  Then the minor mode command
does not need to do anything except set the variable.

This variable is used in conjunction with the @code{minor-mode-alist} to
display the minor mode name in the mode line.  It can also enable
or disable a minor mode keymap.  Individual commands or hooks can also
check the variable's value.

If you want the minor mode to be enabled separately in each buffer,
make the variable buffer-local.

@item
Define a command whose name is the same as the mode variable.
Its job is to enable and disable the mode by setting the variable.

The command should accept one optional argument.  If the argument is
@code{nil}, it should toggle the mode (turn it on if it is off, and
off if it is on).  It should turn the mode on if the argument is a
positive integer, the symbol @code{t}, or a list whose @sc{car} is one
of those.  It should turn the mode off if the argument is a negative
integer or zero, the symbol @code{-}, or a list whose @sc{car} is a
negative integer or zero.  The meaning of other arguments is not
specified.

Here is an example taken from the definition of @code{transient-mark-mode}.
It shows the use of @code{transient-mark-mode} as a variable that enables or
disables the mode's behavior, and also shows the proper way to toggle,
enable or disable the minor mode based on the raw prefix argument value.

@smallexample
@group
(setq transient-mark-mode
      (if (null arg) (not transient-mark-mode)
        (> (prefix-numeric-value arg) 0)))
@end group
@end smallexample

@item
Add an element to @code{minor-mode-alist} for each minor mode
(@pxref{Definition of minor-mode-alist}), if you want to indicate the
minor mode in the mode line.  This element should be a list of the
following form:

@smallexample
(@var{mode-variable} @var{string})
@end smallexample

Here @var{mode-variable} is the variable that controls enabling of the
minor mode, and @var{string} is a short string, starting with a space,
to represent the mode in the mode line.  These strings must be short so
that there is room for several of them at once.

When you add an element to @code{minor-mode-alist}, use @code{assq} to
check for an existing element, to avoid duplication.  For example:

@smallexample
@group
(unless (assq 'leif-mode minor-mode-alist)
  (setq minor-mode-alist
        (cons '(leif-mode " Leif") minor-mode-alist)))
@end group
@end smallexample

@noindent
or like this, using @code{add-to-list} (@pxref{List Variables}):

@smallexample
@group
(add-to-list 'minor-mode-alist '(leif-mode " Leif"))
@end group
@end smallexample
@end itemize

  Global minor modes distributed with Emacs should if possible support
enabling and disabling via Custom (@pxref{Customization}).  To do this,
the first step is to define the mode variable with @code{defcustom}, and
specify @code{:type 'boolean}.

  If just setting the variable is not sufficient to enable the mode, you
should also specify a @code{:set} method which enables the mode by
invoking the mode command.  Note in the variable's documentation string that
setting the variable other than via Custom may not take effect.

  Also mark the definition with an autoload cookie (@pxref{autoload cookie}),
and specify a @code{:require} so that customizing the variable will load
the library that defines the mode.  This will copy suitable definitions
into @file{loaddefs.el} so that users can use @code{customize-option} to
enable the mode.  For example:

@smallexample
@group

;;;###autoload
(defcustom msb-mode nil
  "Toggle msb-mode.
Setting this variable directly does not take effect;
use either \\[customize] or the function `msb-mode'."
  :set 'custom-set-minor-mode
  :initialize 'custom-initialize-default
  :version "20.4"
  :type    'boolean
  :group   'msb
  :require 'msb)
@end group
@end smallexample

@node Keymaps and Minor Modes
@subsection Keymaps and Minor Modes

  Each minor mode can have its own keymap, which is active when the mode
is enabled.  To set up a keymap for a minor mode, add an element to the
alist @code{minor-mode-map-alist}.  @xref{Definition of minor-mode-map-alist}.

@cindex @code{self-insert-command}, minor modes
  One use of minor mode keymaps is to modify the behavior of certain
self-inserting characters so that they do something else as well as
self-insert.  In general, this is the only way to do that, since the
facilities for customizing @code{self-insert-command} are limited to
special cases (designed for abbrevs and Auto Fill mode).  (Do not try
substituting your own definition of @code{self-insert-command} for the
standard one.  The editor command loop handles this function specially.)

The key sequences bound in a minor mode should consist of @kbd{C-c}
followed by one of @kbd{.,/?`'"[]\|~!#$%^&*()-_+=}.  (The other
punctuation characters are reserved for major modes.)

@node Defining Minor Modes
@subsection Defining Minor Modes

  The macro @code{define-minor-mode} offers a convenient way of
implementing a mode in one self-contained definition.

@defmac define-minor-mode mode doc [init-value [lighter [keymap]]] keyword-args@dots{} body@dots{}
This macro defines a new minor mode whose name is @var{mode} (a
symbol).  It defines a command named @var{mode} to toggle the minor
mode, with @var{doc} as its documentation string.  It also defines a
variable named @var{mode}, which is set to @code{t} or @code{nil} by
enabling or disabling the mode.  The variable is initialized to
@var{init-value}.  Except in unusual circumstances (see below), this
value must be @code{nil}.

The string @var{lighter} says what to display in the mode line
when the mode is enabled; if it is @code{nil}, the mode is not displayed
in the mode line.

The optional argument @var{keymap} specifies the keymap for the minor
mode.  If non-@code{nil}, it should be a variable name (whose value is
a keymap), a keymap, or an alist of the form

@example
(@var{key-sequence} . @var{definition})
@end example

@noindent
where each @var{key-sequence} and @var{definition} are arguments
suitable for passing to @code{define-key} (@pxref{Changing Key
Bindings}).  If @var{keymap} is a keymap or an alist, this also
defines the variable @code{@var{mode}-map}.

The above three arguments @var{init-value}, @var{lighter}, and
@var{keymap} can be (partially) omitted when @var{keyword-args} are
used.  The @var{keyword-args} consist of keywords followed by
corresponding values.  A few keywords have special meanings:

@table @code
@item :group @var{group}
Custom group name to use in all generated @code{defcustom} forms.
Defaults to @var{mode} without the possible trailing @samp{-mode}.
@strong{Warning:} don't use this default group name unless you have
written a @code{defgroup} to define that group properly.  @xref{Group
Definitions}.

@item :global @var{global}
If non-@code{nil}, this specifies that the minor mode should be global
rather than buffer-local.  It defaults to @code{nil}.

One of the effects of making a minor mode global is that the
@var{mode} variable becomes a customization variable.  Toggling it
through the Custom interface turns the mode on and off, and its value
can be saved for future Emacs sessions (@pxref{Saving
Customizations,,, emacs, The GNU Emacs Manual}.  For the saved
variable to work, you should ensure that the @code{define-minor-mode}
form is evaluated each time Emacs starts; for packages that are not
part of Emacs, the easiest way to do this is to specify a
@code{:require} keyword.

@item :init-value @var{init-value}
This is equivalent to specifying @var{init-value} positionally.

@item :lighter @var{lighter}
This is equivalent to specifying @var{lighter} positionally.

@item :keymap @var{keymap}
This is equivalent to specifying @var{keymap} positionally.
@end table

Any other keyword arguments are passed directly to the
@code{defcustom} generated for the variable @var{mode}.

The command named @var{mode} first performs the standard actions such
as setting the variable named @var{mode} and then executes the
@var{body} forms, if any.  It finishes by running the mode hook
variable @code{@var{mode}-hook}.
@end defmac

  The initial value must be @code{nil} except in cases where (1) the
mode is preloaded in Emacs, or (2) it is painless for loading to
enable the mode even though the user did not request it.  For
instance, if the mode has no effect unless something else is enabled,
and will always be loaded by that time, enabling it by default is
harmless.  But these are unusual circumstances.  Normally, the
initial value must be @code{nil}.

@findex easy-mmode-define-minor-mode
  The name @code{easy-mmode-define-minor-mode} is an alias
for this macro.

  Here is an example of using @code{define-minor-mode}:

@smallexample
(define-minor-mode hungry-mode
  "Toggle Hungry mode.
With no argument, this command toggles the mode.
Non-null prefix argument turns on the mode.
Null prefix argument turns off the mode.

When Hungry mode is enabled, the control delete key
gobbles all preceding whitespace except the last.
See the command \\[hungry-electric-delete]."
 ;; The initial value.
 nil
 ;; The indicator for the mode line.
 " Hungry"
 ;; The minor mode bindings.
 '(([C-backspace] . hungry-electric-delete))
 :group 'hunger)
@end smallexample

@noindent
This defines a minor mode named ``Hungry mode,'' a command named
@code{hungry-mode} to toggle it, a variable named @code{hungry-mode}
which indicates whether the mode is enabled, and a variable named
@code{hungry-mode-map} which holds the keymap that is active when the
mode is enabled.  It initializes the keymap with a key binding for
@kbd{C-@key{DEL}}.  It puts the variable @code{hungry-mode} into
custom group @code{hunger}.  There are no @var{body} forms---many
minor modes don't need any.

  Here's an equivalent way to write it:

@smallexample
(define-minor-mode hungry-mode
  "Toggle Hungry mode.
With no argument, this command toggles the mode.
Non-null prefix argument turns on the mode.
Null prefix argument turns off the mode.

When Hungry mode is enabled, the control delete key
gobbles all preceding whitespace except the last.
See the command \\[hungry-electric-delete]."
 ;; The initial value.
 :init-value nil
 ;; The indicator for the mode line.
 :lighter " Hungry"
 ;; The minor mode bindings.
 :keymap
 '(([C-backspace] . hungry-electric-delete)
   ([C-M-backspace]
    . (lambda ()
        (interactive)
        (hungry-electric-delete t))))
 :group 'hunger)
@end smallexample

@defmac define-globalized-minor-mode global-mode mode turn-on keyword-args@dots{}
This defines a global toggle named @var{global-mode} whose meaning is
to enable or disable the buffer-local minor mode @var{mode} in all
buffers.  To turn on the minor mode in a buffer, it uses the function
@var{turn-on}; to turn off the minor mode, it calls @code{mode} with
@minus{}1 as argument.

Globally enabling the mode also affects buffers subsequently created
by visiting files, and buffers that use a major mode other than
Fundamental mode; but it does not detect the creation of a new buffer
in Fundamental mode.

This defines the customization option @var{global-mode} (@pxref{Customization}),
which can be toggled in the Custom interface to turn the minor mode on
and off.  As with @code{define-minor-mode}, you should ensure that the
@code{define-globalized-minor-mode} form is evaluated each time Emacs
starts, for example by providing a @code{:require} keyword.

Use @code{:group @var{group}} in @var{keyword-args} to specify the
custom group for the mode variable of the global minor mode.
@end defmac

@node Mode Line Format
@section Mode-Line Format
@cindex mode line

  Each Emacs window (aside from minibuffer windows) typically has a mode
line at the bottom, which displays status information about the buffer
displayed in the window.  The mode line contains information about the
buffer, such as its name, associated file, depth of recursive editing,
and major and minor modes.  A window can also have a @dfn{header
line}, which is much like the mode line but appears at the top of the
window.

  This section describes how to control the contents of the mode line
and header line.  We include it in this chapter because much of the
information displayed in the mode line relates to the enabled major and
minor modes.

@menu
* Base: Mode Line Basics.       Basic ideas of mode line control.
* Data: Mode Line Data.         The data structure that controls the mode line.
* Top: Mode Line Top.           The top level variable, mode-line-format.
* Mode Line Variables::         Variables used in that data structure.
* %-Constructs::                Putting information into a mode line.
* Properties in Mode::          Using text properties in the mode line.
* Header Lines::                Like a mode line, but at the top.
* Emulating Mode Line::         Formatting text as the mode line would.
@end menu

@node Mode Line Basics
@subsection Mode Line Basics

  @code{mode-line-format} is a buffer-local variable that holds a
@dfn{mode line construct}, a kind of template, which controls what is
displayed on the mode line of the current buffer.  The value of
@code{header-line-format} specifies the buffer's header line in the
same way.  All windows for the same buffer use the same
@code{mode-line-format} and @code{header-line-format}.

  For efficiency, Emacs does not continuously recompute the mode
line and header line of a window.  It does so when circumstances
appear to call for it---for instance, if you change the window
configuration, switch buffers, narrow or widen the buffer, scroll, or
change the buffer's modification status.  If you modify any of the
variables referenced by @code{mode-line-format} (@pxref{Mode Line
Variables}), or any other variables and data structures that affect
how text is displayed (@pxref{Display}), you may want to force an
update of the mode line so as to display the new information or
display it in the new way.

@defun force-mode-line-update &optional all
Force redisplay of the current buffer's mode line and header line.
The next redisplay will update the mode line and header line based on
the latest values of all relevant variables.  With optional
non-@code{nil} @var{all}, force redisplay of all mode lines and header
lines.

This function also forces recomputation of the menu bar menus
and the frame title.
@end defun

  The selected window's mode line is usually displayed in a different
color using the face @code{mode-line}.  Other windows' mode lines
appear in the face @code{mode-line-inactive} instead.  @xref{Faces}.

@node Mode Line Data
@subsection The Data Structure of the Mode Line
@cindex mode-line construct

  The mode-line contents are controlled by a data structure called a
@dfn{mode-line construct}, made up of lists, strings, symbols, and
numbers kept in buffer-local variables.  Each data type has a specific
meaning for the mode-line appearance, as described below.  The same
data structure is used for constructing frame titles (@pxref{Frame
Titles}) and header lines (@pxref{Header Lines}).

  A mode-line construct may be as simple as a fixed string of text,
but it usually specifies how to combine fixed strings with variables'
values to construct the text.  Many of these variables are themselves
defined to have mode-line constructs as their values.

  Here are the meanings of various data types as mode-line constructs:

@table @code
@cindex percent symbol in mode line
@item @var{string}
A string as a mode-line construct appears verbatim except for
@dfn{@code{%}-constructs} in it.  These stand for substitution of
other data; see @ref{%-Constructs}.

If parts of the string have @code{face} properties, they control
display of the text just as they would text in the buffer.  Any
characters which have no @code{face} properties are displayed, by
default, in the face @code{mode-line} or @code{mode-line-inactive}
(@pxref{Standard Faces,,, emacs, The GNU Emacs Manual}).  The
@code{help-echo} and @code{local-map} properties in @var{string} have
special meanings.  @xref{Properties in Mode}.

@item @var{symbol}
A symbol as a mode-line construct stands for its value.  The value of
@var{symbol} is used as a mode-line construct, in place of @var{symbol}.
However, the symbols @code{t} and @code{nil} are ignored, as is any
symbol whose value is void.

There is one exception: if the value of @var{symbol} is a string, it is
displayed verbatim: the @code{%}-constructs are not recognized.

Unless @var{symbol} is marked as ``risky'' (i.e., it has a
non-@code{nil} @code{risky-local-variable} property), all text
properties specified in @var{symbol}'s value are ignored.  This
includes the text properties of strings in @var{symbol}'s value, as
well as all @code{:eval} and @code{:propertize} forms in it.  (The
reason for this is security: non-risky variables could be set
automatically from file variables without prompting the user.)

@item (@var{string} @var{rest}@dots{})
@itemx (@var{list} @var{rest}@dots{})
A list whose first element is a string or list means to process all the
elements recursively and concatenate the results.  This is the most
common form of mode-line construct.

@item (:eval @var{form})
A list whose first element is the symbol @code{:eval} says to evaluate
@var{form}, and use the result as a string to display.  Make sure this
evaluation cannot load any files, as doing so could cause infinite
recursion.

@item (:propertize @var{elt} @var{props}@dots{})
A list whose first element is the symbol @code{:propertize} says to
process the mode-line construct @var{elt} recursively, then add the text
properties specified by @var{props} to the result.  The argument
@var{props} should consist of zero or more pairs @var{text-property}
@var{value}.  (This feature is new as of Emacs 22.1.)

@item (@var{symbol} @var{then} @var{else})
A list whose first element is a symbol that is not a keyword specifies
a conditional.  Its meaning depends on the value of @var{symbol}.  If
@var{symbol} has a non-@code{nil} value, the second element,
@var{then}, is processed recursively as a mode-line element.
Otherwise, the third element, @var{else}, is processed recursively.
You may omit @var{else}; then the mode-line element displays nothing
if the value of @var{symbol} is @code{nil} or void.

@item (@var{width} @var{rest}@dots{})
A list whose first element is an integer specifies truncation or
padding of the results of @var{rest}.  The remaining elements
@var{rest} are processed recursively as mode-line constructs and
concatenated together.  When @var{width} is positive, the result is
space filled on the right if its width is less than @var{width}.  When
@var{width} is negative, the result is truncated on the right to
@minus{}@var{width} columns if its width exceeds @minus{}@var{width}.

For example, the usual way to show what percentage of a buffer is above
the top of the window is to use a list like this: @code{(-3 "%p")}.
@end table

@node Mode Line Top
@subsection The Top Level of Mode Line Control

  The variable in overall control of the mode line is
@code{mode-line-format}.

@defopt mode-line-format
The value of this variable is a mode-line construct that controls the
contents of the mode-line.  It is always buffer-local in all buffers.

If you set this variable to @code{nil} in a buffer, that buffer does
not have a mode line.  (A window that is just one line tall never
displays a mode line.)
@end defopt

  The default value of @code{mode-line-format} is designed to use the
values of other variables such as @code{mode-line-position} and
@code{mode-line-modes} (which in turn incorporates the values of the
variables @code{mode-name} and @code{minor-mode-alist}).  Very few
modes need to alter @code{mode-line-format} itself.  For most
purposes, it is sufficient to alter some of the variables that
@code{mode-line-format} either directly or indirectly refers to.

  If you do alter @code{mode-line-format} itself, the new value should
use the same variables that appear in the default value (@pxref{Mode
Line Variables}), rather than duplicating their contents or displaying
the information in another fashion.  This way, customizations made by
the user or by Lisp programs (such as @code{display-time} and major
modes) via changes to those variables remain effective.

  Here is an example of a @code{mode-line-format} that might be
useful for @code{shell-mode}, since it contains the host name and default
directory.

@example
@group
(setq mode-line-format
  (list "-"
   'mode-line-mule-info
   'mode-line-modified
   'mode-line-frame-identification
   "%b--"
@end group
@group
   ;; @r{Note that this is evaluated while making the list.}
   ;; @r{It makes a mode-line construct which is just a string.}
   (getenv "HOST")
@end group
   ":"
   'default-directory
   "   "
   'global-mode-string
   "   %[("
   '(:eval (mode-line-mode-name))
   'mode-line-process
   'minor-mode-alist
   "%n"
   ")%]--"
@group
   '(which-func-mode ("" which-func-format "--"))
   '(line-number-mode "L%l--")
   '(column-number-mode "C%c--")
   '(-3 "%p")
   "-%-"))
@end group
@end example

@noindent
(The variables @code{line-number-mode}, @code{column-number-mode}
and @code{which-func-mode} enable particular minor modes; as usual,
these variable names are also the minor mode command names.)

@node Mode Line Variables
@subsection Variables Used in the Mode Line

  This section describes variables incorporated by the standard value
of @code{mode-line-format} into the text of the mode line.  There is
nothing inherently special about these variables; any other variables
could have the same effects on the mode line if
@code{mode-line-format}'s value were changed to use them.  However,
various parts of Emacs set these variables on the understanding that
they will control parts of the mode line; therefore, practically
speaking, it is essential for the mode line to use them.

@defvar mode-line-mule-info
This variable holds the value of the mode-line construct that displays
information about the language environment, buffer coding system, and
current input method.  @xref{Non-ASCII Characters}.
@end defvar

@defvar mode-line-modified
This variable holds the value of the mode-line construct that displays
whether the current buffer is modified.  Its default value displays
@samp{**} if the buffer is modified, @samp{--} if the buffer is not
modified, @samp{%%} if the buffer is read only, and @samp{%*} if the
buffer is read only and modified.

Changing this variable does not force an update of the mode line.
@end defvar

@defvar mode-line-frame-identification
This variable identifies the current frame.  Its default value
displays @code{" "} if you are using a window system which can show
multiple frames, or @code{"-%F "} on an ordinary terminal which shows
only one frame at a time.
@end defvar

@defvar mode-line-buffer-identification
This variable identifies the buffer being displayed in the window.
Its default value displays the buffer name, padded with spaces to at
least 12 columns.
@end defvar

@defopt mode-line-position
This variable indicates the position in the buffer.  Its default value
displays the buffer percentage and, optionally, the buffer size, the
line number and the column number.
@end defopt

@defvar vc-mode
The variable @code{vc-mode}, buffer-local in each buffer, records
whether the buffer's visited file is maintained with version control,
and, if so, which kind.  Its value is a string that appears in the mode
line, or @code{nil} for no version control.
@end defvar

@defopt mode-line-modes
This variable displays the buffer's major and minor modes.  Its
default value also displays the recursive editing level, information
on the process status, and whether narrowing is in effect.
@end defopt

  The following three variables are used in @code{mode-line-modes}:

@defvar mode-name
This buffer-local variable holds the ``pretty'' name of the current
buffer's major mode.  Each major mode should set this variable so that
the mode name will appear in the mode line.  The value does not have
to be a string, but can use any of the data types valid in a mode-line
construct (@pxref{Mode Line Data}).  To compute the string that will
identify the mode name in the mode line, use @code{format-mode-line}
(@pxref{Emulating Mode Line}).
@end defvar

@defvar mode-line-process
This buffer-local variable contains the mode-line information on process
status in modes used for communicating with subprocesses.  It is
displayed immediately following the major mode name, with no intervening
space.  For example, its value in the @samp{*shell*} buffer is
@code{(":%s")}, which allows the shell to display its status along
with the major mode as: @samp{(Shell:run)}.  Normally this variable
is @code{nil}.
@end defvar

@defvar minor-mode-alist
@anchor{Definition of minor-mode-alist}
This variable holds an association list whose elements specify how the
mode line should indicate that a minor mode is active.  Each element of
the @code{minor-mode-alist} should be a two-element list:

@example
(@var{minor-mode-variable} @var{mode-line-string})
@end example

More generally, @var{mode-line-string} can be any mode-line spec.  It
appears in the mode line when the value of @var{minor-mode-variable}
is non-@code{nil}, and not otherwise.  These strings should begin with
spaces so that they don't run together.  Conventionally, the
@var{minor-mode-variable} for a specific mode is set to a
non-@code{nil} value when that minor mode is activated.

@code{minor-mode-alist} itself is not buffer-local.  Each variable
mentioned in the alist should be buffer-local if its minor mode can be
enabled separately in each buffer.
@end defvar

@defvar global-mode-string
This variable holds a mode-line spec that, by default, appears in the
mode line just after the @code{which-func-mode} minor mode if set,
else after @code{mode-line-modes}.  The command @code{display-time}
sets @code{global-mode-string} to refer to the variable
@code{display-time-string}, which holds a string containing the time
and load information.

The @samp{%M} construct substitutes the value of
@code{global-mode-string}, but that is obsolete, since the variable is
included in the mode line from @code{mode-line-format}.
@end defvar

Here is a simplified version of the default value of
@code{mode-line-format}.  The real default value also
specifies addition of text properties.

@example
@group
("-"
 mode-line-mule-info
 mode-line-modified
 mode-line-frame-identification
 mode-line-buffer-identification
@end group
 "   "
 mode-line-position
 (vc-mode vc-mode)
 "   "
@group
 mode-line-modes
 (which-func-mode ("" which-func-format "--"))
 (global-mode-string ("--" global-mode-string))
 "-%-")
@end group
@end example

@node %-Constructs
@subsection @code{%}-Constructs in the Mode Line

  Strings used as mode-line constructs can use certain
@code{%}-constructs to substitute various kinds of data.  Here is a
list of the defined @code{%}-constructs, and what they mean.  In any
construct except @samp{%%}, you can add a decimal integer after the
@samp{%} to specify a minimum field width.  If the width is less, the
field is padded with spaces to the right.

@table @code
@item %b
The current buffer name, obtained with the @code{buffer-name} function.
@xref{Buffer Names}.

@item %c
The current column number of point.

@item %e
When Emacs is nearly out of memory for Lisp objects, a brief message
saying so.  Otherwise, this is empty.

@item %f
The visited file name, obtained with the @code{buffer-file-name}
function.  @xref{Buffer File Name}.

@item %F
The title (only on a window system) or the name of the selected frame.
@xref{Basic Parameters}.

@item %i
The size of the accessible part of the current buffer; basically
@code{(- (point-max) (point-min))}.

@item %I
Like @samp{%i}, but the size is printed in a more readable way by using
@samp{k} for 10^3, @samp{M} for 10^6, @samp{G} for 10^9, etc., to
abbreviate.

@item %l
The current line number of point, counting within the accessible portion
of the buffer.

@item %n
@samp{Narrow} when narrowing is in effect; nothing otherwise (see
@code{narrow-to-region} in @ref{Narrowing}).

@item %p
The percentage of the buffer text above the @strong{top} of window, or
@samp{Top}, @samp{Bottom} or @samp{All}.  Note that the default
mode-line specification truncates this to three characters.

@item %P
The percentage of the buffer text that is above the @strong{bottom} of
the window (which includes the text visible in the window, as well as
the text above the top), plus @samp{Top} if the top of the buffer is
visible on screen; or @samp{Bottom} or @samp{All}.

@item %s
The status of the subprocess belonging to the current buffer, obtained with
@code{process-status}.  @xref{Process Information}.

@item %t
Whether the visited file is a text file or a binary file.  This is a
meaningful distinction only on certain operating systems (@pxref{MS-DOS
File Types}).

@item %z
The mnemonics of keyboard, terminal, and buffer coding systems.

@item %Z
Like @samp{%z}, but including the end-of-line format.

@item %*
@samp{%} if the buffer is read only (see @code{buffer-read-only}); @*
@samp{*} if the buffer is modified (see @code{buffer-modified-p}); @*
@samp{-} otherwise.  @xref{Buffer Modification}.

@item %+
@samp{*} if the buffer is modified (see @code{buffer-modified-p}); @*
@samp{%} if the buffer is read only (see @code{buffer-read-only}); @*
@samp{-} otherwise.  This differs from @samp{%*} only for a modified
read-only buffer.  @xref{Buffer Modification}.

@item %&
@samp{*} if the buffer is modified, and @samp{-} otherwise.

@item %[
An indication of the depth of recursive editing levels (not counting
minibuffer levels): one @samp{[} for each editing level.
@xref{Recursive Editing}.

@item %]
One @samp{]} for each recursive editing level (not counting minibuffer
levels).

@item %-
Dashes sufficient to fill the remainder of the mode line.

@item %%
The character @samp{%}---this is how to include a literal @samp{%} in a
string in which @code{%}-constructs are allowed.
@end table

The following two @code{%}-constructs are still supported, but they are
obsolete, since you can get the same results with the variables
@code{mode-name} and @code{global-mode-string}.

@table @code
@item %m
The value of @code{mode-name}.

@item %M
The value of @code{global-mode-string}.
@end table

@node Properties in Mode
@subsection Properties in the Mode Line
@cindex text properties in the mode line

  Certain text properties are meaningful in the
mode line.  The @code{face} property affects the appearance of text; the
@code{help-echo} property associates help strings with the text, and
@code{local-map} can make the text mouse-sensitive.

  There are four ways to specify text properties for text in the mode
line:

@enumerate
@item
Put a string with a text property directly into the mode-line data
structure.

@item
Put a text property on a mode-line %-construct such as @samp{%12b}; then
the expansion of the %-construct will have that same text property.

@item
Use a @code{(:propertize @var{elt} @var{props}@dots{})} construct to
give @var{elt} a text property specified by @var{props}.

@item
Use a list containing @code{:eval @var{form}} in the mode-line data
structure, and make @var{form} evaluate to a string that has a text
property.
@end enumerate

  You can use the @code{local-map} property to specify a keymap.  This
keymap only takes real effect for mouse clicks; binding character keys
and function keys to it has no effect, since it is impossible to move
point into the mode line.

  When the mode line refers to a variable which does not have a
non-@code{nil} @code{risky-local-variable} property, any text
properties given or specified within that variable's values are
ignored.  This is because such properties could otherwise specify
functions to be called, and those functions could come from file
local variables.

@node Header Lines
@subsection Window Header Lines
@cindex header line (of a window)
@cindex window header line

  A window can have a @dfn{header line} at the
top, just as it can have a mode line at the bottom.  The header line
feature works just like the mode-line feature, except that it's
controlled by different variables.

@defvar header-line-format
This variable, local in every buffer, specifies how to display the
header line, for windows displaying the buffer.  The format of the value
is the same as for @code{mode-line-format} (@pxref{Mode Line Data}).
It is normally @code{nil}, so that ordinary buffers have no header line.
@end defvar

  A window that is just one line tall never displays a header line.  A
window that is two lines tall cannot display both a mode line and a
header line at once; if it has a mode line, then it does not display a
header line.

@node Emulating Mode Line
@subsection Emulating Mode-Line Formatting

  You can use the function @code{format-mode-line} to compute
the text that would appear in a mode line or header line
based on a certain mode-line specification.

@defun format-mode-line format &optional face window buffer
This function formats a line of text according to @var{format} as if it
were generating the mode line for @var{window}, but it also returns the
text as a string.  The argument @var{window} defaults to the selected
window.  If @var{buffer} is non-@code{nil}, all the information used is
taken from @var{buffer}; by default, it comes from @var{window}'s
buffer.

The value string normally has text properties that correspond to the
faces, keymaps, etc., that the mode line would have.  Any character for
which no @code{face} property is specified by @var{format} gets a
default value determined by @var{face}.  If @var{face} is @code{t}, that
stands for either @code{mode-line} if @var{window} is selected,
otherwise @code{mode-line-inactive}.  If @var{face} is @code{nil} or
omitted, that stands for the default face.  If @var{face} is an integer,
the value returned by this function will have no text properties.

You can also specify other valid faces as the value of @var{face}.
If specified, that face provides the @code{face} property for characters
whose face is not specified by @var{format}.

Note that using @code{mode-line}, @code{mode-line-inactive}, or
@code{header-line} as @var{face} will actually redisplay the mode line
or the header line, respectively, using the current definitions of the
corresponding face, in addition to returning the formatted string.
(Other faces do not cause redisplay.)

For example, @code{(format-mode-line header-line-format)} returns the
text that would appear in the selected window's header line (@code{""}
if it has no header line).  @code{(format-mode-line header-line-format
'header-line)} returns the same text, with each character
carrying the face that it will have in the header line itself, and also
redraws the header line.
@end defun

@node Imenu
@section Imenu

@cindex Imenu
  @dfn{Imenu} is a feature that lets users select a definition or
section in the buffer, from a menu which lists all of them, to go
directly to that location in the buffer.  Imenu works by constructing
a buffer index which lists the names and buffer positions of the
definitions, or other named portions of the buffer; then the user can
choose one of them and move point to it.  Major modes can add a menu
bar item to use Imenu using @code{imenu-add-to-menubar}.

@defun imenu-add-to-menubar name
This function defines a local menu bar item named @var{name}
to run Imenu.
@end defun

  The user-level commands for using Imenu are described in the Emacs
Manual (@pxref{Imenu,, Imenu, emacs, the Emacs Manual}).  This section
explains how to customize Imenu's method of finding definitions or
buffer portions for a particular major mode.

  The usual and simplest way is to set the variable
@code{imenu-generic-expression}:

@defvar imenu-generic-expression
This variable, if non-@code{nil}, is a list that specifies regular
expressions for finding definitions for Imenu.  Simple elements of
@code{imenu-generic-expression} look like this:

@example
(@var{menu-title} @var{regexp} @var{index})
@end example

Here, if @var{menu-title} is non-@code{nil}, it says that the matches
for this element should go in a submenu of the buffer index;
@var{menu-title} itself specifies the name for the submenu.  If
@var{menu-title} is @code{nil}, the matches for this element go directly
in the top level of the buffer index.

The second item in the list, @var{regexp}, is a regular expression
(@pxref{Regular Expressions}); anything in the buffer that it matches
is considered a definition, something to mention in the buffer index.
The third item, @var{index}, is a non-negative integer that indicates
which subexpression in @var{regexp} matches the definition's name.

An element can also look like this:

@example
(@var{menu-title} @var{regexp} @var{index} @var{function} @var{arguments}@dots{})
@end example

Each match for this element creates an index item, and when the index
item is selected by the user, it calls @var{function} with arguments
consisting of the item name, the buffer position, and @var{arguments}.

For Emacs Lisp mode, @code{imenu-generic-expression} could look like
this:

@c should probably use imenu-syntax-alist and \\sw rather than [-A-Za-z0-9+]
@example
@group
((nil "^\\s-*(def\\(un\\|subst\\|macro\\|advice\\)\
\\s-+\\([-A-Za-z0-9+]+\\)" 2)
@end group
@group
 ("*Vars*" "^\\s-*(def\\(var\\|const\\)\
\\s-+\\([-A-Za-z0-9+]+\\)" 2)
@end group
@group
 ("*Types*"
  "^\\s-*\
(def\\(type\\|struct\\|class\\|ine-condition\\)\
\\s-+\\([-A-Za-z0-9+]+\\)" 2))
@end group
@end example

Setting this variable makes it buffer-local in the current buffer.
@end defvar

@defvar imenu-case-fold-search
This variable controls whether matching against the regular
expressions in the value of @code{imenu-generic-expression} is
case-sensitive: @code{t}, the default, means matching should ignore
case.

Setting this variable makes it buffer-local in the current buffer.
@end defvar

@defvar imenu-syntax-alist
This variable is an alist of syntax table modifiers to use while
processing @code{imenu-generic-expression}, to override the syntax table
of the current buffer.  Each element should have this form:

@example
(@var{characters} . @var{syntax-description})
@end example

The @sc{car}, @var{characters}, can be either a character or a string.
The element says to give that character or characters the syntax
specified by @var{syntax-description}, which is passed to
@code{modify-syntax-entry} (@pxref{Syntax Table Functions}).

This feature is typically used to give word syntax to characters which
normally have symbol syntax, and thus to simplify
@code{imenu-generic-expression} and speed up matching.
For example, Fortran mode uses it this way:

@example
(setq imenu-syntax-alist '(("_$" . "w")))
@end example

The @code{imenu-generic-expression} regular expressions can then use
@samp{\\sw+} instead of @samp{\\(\\sw\\|\\s_\\)+}.  Note that this
technique may be inconvenient when the mode needs to limit the initial
character of a name to a smaller set of characters than are allowed in
the rest of a name.

Setting this variable makes it buffer-local in the current buffer.
@end defvar

  Another way to customize Imenu for a major mode is to set the
variables @code{imenu-prev-index-position-function} and
@code{imenu-extract-index-name-function}:

@defvar imenu-prev-index-position-function
If this variable is non-@code{nil}, its value should be a function that
finds the next ``definition'' to put in the buffer index, scanning
backward in the buffer from point.  It should return @code{nil} if it
doesn't find another ``definition'' before point.  Otherwise it should
leave point at the place it finds a ``definition'' and return any
non-@code{nil} value.

Setting this variable makes it buffer-local in the current buffer.
@end defvar

@defvar imenu-extract-index-name-function
If this variable is non-@code{nil}, its value should be a function to
return the name for a definition, assuming point is in that definition
as the @code{imenu-prev-index-position-function} function would leave
it.

Setting this variable makes it buffer-local in the current buffer.
@end defvar

  The last way to customize Imenu for a major mode is to set the
variable @code{imenu-create-index-function}:

@defvar imenu-create-index-function
This variable specifies the function to use for creating a buffer
index.  The function should take no arguments, and return an index
alist for the current buffer.  It is called within
@code{save-excursion}, so where it leaves point makes no difference.

The index alist can have three types of elements.  Simple elements
look like this:

@example
(@var{index-name} . @var{index-position})
@end example

Selecting a simple element has the effect of moving to position
@var{index-position} in the buffer.  Special elements look like this:

@example
(@var{index-name} @var{index-position} @var{function} @var{arguments}@dots{})
@end example

Selecting a special element performs:

@example
(funcall @var{function}
         @var{index-name} @var{index-position} @var{arguments}@dots{})
@end example

A nested sub-alist element looks like this:

@example
(@var{menu-title} @var{sub-alist})
@end example

It creates the submenu @var{menu-title} specified by @var{sub-alist}.

The default value of @code{imenu-create-index-function} is
@code{imenu-default-create-index-function}.  This function calls the
value of @code{imenu-prev-index-position-function} and the value of
@code{imenu-extract-index-name-function} to produce the index alist.
However, if either of these two variables is @code{nil}, the default
function uses @code{imenu-generic-expression} instead.

Setting this variable makes it buffer-local in the current buffer.
@end defvar

@node Font Lock Mode
@section Font Lock Mode
@cindex Font Lock mode

  @dfn{Font Lock mode} is a feature that automatically attaches
@code{face} properties to certain parts of the buffer based on their
syntactic role.  How it parses the buffer depends on the major mode;
most major modes define syntactic criteria for which faces to use in
which contexts.  This section explains how to customize Font Lock for a
particular major mode.

  Font Lock mode finds text to highlight in two ways: through
syntactic parsing based on the syntax table, and through searching
(usually for regular expressions).  Syntactic fontification happens
first; it finds comments and string constants and highlights them.
Search-based fontification happens second.

@menu
* Font Lock Basics::            Overview of customizing Font Lock.
* Search-based Fontification::  Fontification based on regexps.
* Customizing Keywords::        Customizing search-based fontification.
* Other Font Lock Variables::   Additional customization facilities.
* Levels of Font Lock::         Each mode can define alternative levels
                                  so that the user can select more or less.
* Precalculated Fontification::  How Lisp programs that produce the buffer
                                  contents can also specify how to fontify it.
* Faces for Font Lock::         Special faces specifically for Font Lock.
* Syntactic Font Lock::         Fontification based on syntax tables.
* Setting Syntax Properties::   Defining character syntax based on context
                                  using the Font Lock mechanism.
* Multiline Font Lock::         How to coerce Font Lock into properly
                                  highlighting multiline constructs.
@end menu

@node Font Lock Basics
@subsection Font Lock Basics

  There are several variables that control how Font Lock mode highlights
text.  But major modes should not set any of these variables directly.
Instead, they should set @code{font-lock-defaults} as a buffer-local
variable.  The value assigned to this variable is used, if and when Font
Lock mode is enabled, to set all the other variables.

@defvar font-lock-defaults
This variable is set by major modes, as a buffer-local variable, to
specify how to fontify text in that mode.  It automatically becomes
buffer-local when you set it.  If its value is @code{nil}, Font-Lock
mode does no highlighting, and you can use the @samp{Faces} menu
(under @samp{Edit} and then @samp{Text Properties} in the menu bar) to
assign faces explicitly to text in the buffer.

If non-@code{nil}, the value should look like this:

@example
(@var{keywords} [@var{keywords-only} [@var{case-fold}
 [@var{syntax-alist} [@var{syntax-begin} @var{other-vars}@dots{}]]]])
@end example

The first element, @var{keywords}, indirectly specifies the value of
@code{font-lock-keywords} which directs search-based fontification.
It can be a symbol, a variable or a function whose value is the list
to use for @code{font-lock-keywords}.  It can also be a list of
several such symbols, one for each possible level of fontification.
The first symbol specifies the @samp{mode default} level of
fontification, the next symbol level 1 fontification, the next level 2,
and so on.  The @samp{mode default} level is normally the same as level
1.  It is used when @code{font-lock-maximum-decoration} has a @code{nil}
value.  @xref{Levels of Font Lock}.

The second element, @var{keywords-only}, specifies the value of the
variable @code{font-lock-keywords-only}.  If this is omitted or
@code{nil}, syntactic fontification (of strings and comments) is also
performed.  If this is non-@code{nil}, such fontification is not
performed.  @xref{Syntactic Font Lock}.

The third element, @var{case-fold}, specifies the value of
@code{font-lock-keywords-case-fold-search}.  If it is non-@code{nil},
Font Lock mode ignores case when searching as directed by
@code{font-lock-keywords}.

If the fourth element, @var{syntax-alist}, is non-@code{nil}, it
should be a list of cons cells of the form @code{(@var{char-or-string}
. @var{string})}.  These are used to set up a syntax table for
syntactic fontification (@pxref{Syntax Table Functions}).  The
resulting syntax table is stored in @code{font-lock-syntax-table}.

The fifth element, @var{syntax-begin}, specifies the value of
@code{font-lock-beginning-of-syntax-function}.  We recommend setting
this variable to @code{nil} and using @code{syntax-begin-function}
instead.

All the remaining elements (if any) are collectively called
@var{other-vars}.  Each of these elements should have the form
@code{(@var{variable} . @var{value})}---which means, make
@var{variable} buffer-local and then set it to @var{value}.  You can
use these @var{other-vars} to set other variables that affect
fontification, aside from those you can control with the first five
elements.  @xref{Other Font Lock Variables}.
@end defvar

  If your mode fontifies text explicitly by adding
@code{font-lock-face} properties, it can specify @code{(nil t)} for
@code{font-lock-defaults} to turn off all automatic fontification.
However, this is not required; it is possible to fontify some things
using @code{font-lock-face} properties and set up automatic
fontification for other parts of the text.

@node Search-based Fontification
@subsection Search-based Fontification

  The most important variable for customizing Font Lock mode is
@code{font-lock-keywords}.  It specifies the search criteria for
search-based fontification.  You should specify the value of this
variable with @var{keywords} in @code{font-lock-defaults}.

@defvar font-lock-keywords
This variable's value is a list of the keywords to highlight.  Be
careful when composing regular expressions for this list; a poorly
written pattern can dramatically slow things down!
@end defvar

  Each element of @code{font-lock-keywords} specifies how to find
certain cases of text, and how to highlight those cases.  Font Lock mode
processes the elements of @code{font-lock-keywords} one by one, and for
each element, it finds and handles all matches.  Ordinarily, once
part of the text has been fontified already, this cannot be overridden
by a subsequent match in the same text; but you can specify different
behavior using the @var{override} element of a @var{subexp-highlighter}.

  Each element of @code{font-lock-keywords} should have one of these
forms:

@table @code
@item @var{regexp}
Highlight all matches for @var{regexp} using
@code{font-lock-keyword-face}.  For example,

@example
;; @r{Highlight occurrences of the word @samp{foo}}
;; @r{using @code{font-lock-keyword-face}.}
"\\<foo\\>"
@end example

The function @code{regexp-opt} (@pxref{Regexp Functions}) is useful
for calculating optimal regular expressions to match a number of
different keywords.

@item @var{function}
Find text by calling @var{function}, and highlight the matches
it finds using @code{font-lock-keyword-face}.

When @var{function} is called, it receives one argument, the limit of
the search; it should begin searching at point, and not search beyond the
limit.  It should return non-@code{nil} if it succeeds, and set the
match data to describe the match that was found.  Returning @code{nil}
indicates failure of the search.

Fontification will call @var{function} repeatedly with the same limit,
and with point where the previous invocation left it, until
@var{function} fails.  On failure, @var{function} need not reset point
in any particular way.

@item (@var{matcher} . @var{subexp})
In this kind of element, @var{matcher} is either a regular
expression or a function, as described above.  The @sc{cdr},
@var{subexp}, specifies which subexpression of @var{matcher} should be
highlighted (instead of the entire text that @var{matcher} matched).

@example
;; @r{Highlight the @samp{bar} in each occurrence of @samp{fubar},}
;; @r{using @code{font-lock-keyword-face}.}
("fu\\(bar\\)" . 1)
@end example

If you use @code{regexp-opt} to produce the regular expression
@var{matcher}, you can use @code{regexp-opt-depth} (@pxref{Regexp
Functions}) to calculate the value for @var{subexp}.

@item (@var{matcher} . @var{facespec})
In this kind of element, @var{facespec} is an expression whose value
specifies the face to use for highlighting.  In the simplest case,
@var{facespec} is a Lisp variable (a symbol) whose value is a face
name.

@example
;; @r{Highlight occurrences of @samp{fubar},}
;; @r{using the face which is the value of @code{fubar-face}.}
("fubar" . fubar-face)
@end example

However, @var{facespec} can also evaluate to a list of this form:

@example
(face @var{face} @var{prop1} @var{val1} @var{prop2} @var{val2}@dots{})
@end example

@noindent
to specify the face @var{face} and various additional text properties
to put on the text that matches.  If you do this, be sure to add the
other text property names that you set in this way to the value of
@code{font-lock-extra-managed-props} so that the properties will also
be cleared out when they are no longer appropriate.  Alternatively,
you can set the variable @code{font-lock-unfontify-region-function} to
a function that clears these properties.  @xref{Other Font Lock
Variables}.

@item (@var{matcher} . @var{subexp-highlighter})
In this kind of element, @var{subexp-highlighter} is a list
which specifies how to highlight matches found by @var{matcher}.
It has the form:

@example
(@var{subexp} @var{facespec} [@var{override} [@var{laxmatch}]])
@end example

The @sc{car}, @var{subexp}, is an integer specifying which subexpression
of the match to fontify (0 means the entire matching text).  The second
subelement, @var{facespec}, is an expression whose value specifies the
face, as described above.

The last two values in @var{subexp-highlighter}, @var{override} and
@var{laxmatch}, are optional flags.  If @var{override} is @code{t},
this element can override existing fontification made by previous
elements of @code{font-lock-keywords}.  If it is @code{keep}, then
each character is fontified if it has not been fontified already by
some other element.  If it is @code{prepend}, the face specified by
@var{facespec} is added to the beginning of the @code{font-lock-face}
property.  If it is @code{append}, the face is added to the end of the
@code{font-lock-face} property.

If @var{laxmatch} is non-@code{nil}, it means there should be no error
if there is no subexpression numbered @var{subexp} in @var{matcher}.
Obviously, fontification of the subexpression numbered @var{subexp} will
not occur.  However, fontification of other subexpressions (and other
regexps) will continue.  If @var{laxmatch} is @code{nil}, and the
specified subexpression is missing, then an error is signaled which
terminates search-based fontification.

Here are some examples of elements of this kind, and what they do:

@smallexample
;; @r{Highlight occurrences of either @samp{foo} or @samp{bar}, using}
;; @r{@code{foo-bar-face}, even if they have already been highlighted.}
;; @r{@code{foo-bar-face} should be a variable whose value is a face.}
("foo\\|bar" 0 foo-bar-face t)

;; @r{Highlight the first subexpression within each occurrence}
;; @r{that the function @code{fubar-match} finds,}
;; @r{using the face which is the value of @code{fubar-face}.}
(fubar-match 1 fubar-face)
@end smallexample

@item (@var{matcher} . @var{anchored-highlighter})
In this kind of element, @var{anchored-highlighter} specifies how to
highlight text that follows a match found by @var{matcher}.  So a
match found by @var{matcher} acts as the anchor for further searches
specified by @var{anchored-highlighter}.  @var{anchored-highlighter}
is a list of the following form:

@example
(@var{anchored-matcher} @var{pre-form} @var{post-form}
                        @var{subexp-highlighters}@dots{})
@end example

Here, @var{anchored-matcher}, like @var{matcher}, is either a regular
expression or a function.  After a match of @var{matcher} is found,
point is at the end of the match.  Now, Font Lock evaluates the form
@var{pre-form}.  Then it searches for matches of
@var{anchored-matcher} and uses @var{subexp-highlighters} to highlight
these.  A @var{subexp-highlighter} is as described above.  Finally,
Font Lock evaluates @var{post-form}.

The forms @var{pre-form} and @var{post-form} can be used to initialize
before, and cleanup after, @var{anchored-matcher} is used.  Typically,
@var{pre-form} is used to move point to some position relative to the
match of @var{matcher}, before starting with @var{anchored-matcher}.
@var{post-form} might be used to move back, before resuming with
@var{matcher}.

After Font Lock evaluates @var{pre-form}, it does not search for
@var{anchored-matcher} beyond the end of the line.  However, if
@var{pre-form} returns a buffer position that is greater than the
position of point after @var{pre-form} is evaluated, then the position
returned by @var{pre-form} is used as the limit of the search instead.
It is generally a bad idea to return a position greater than the end
of the line; in other words, the @var{anchored-matcher} search should
not span lines.

For example,

@smallexample
;; @r{Highlight occurrences of the word @samp{item} following}
;; @r{an occurrence of the word @samp{anchor} (on the same line)}
;; @r{in the value of @code{item-face}.}
("\\<anchor\\>" "\\<item\\>" nil nil (0 item-face))
@end smallexample

Here, @var{pre-form} and @var{post-form} are @code{nil}.  Therefore
searching for @samp{item} starts at the end of the match of
@samp{anchor}, and searching for subsequent instances of @samp{anchor}
resumes from where searching for @samp{item} concluded.

@item (@var{matcher} @var{highlighters}@dots{})
This sort of element specifies several @var{highlighter} lists for a
single @var{matcher}.  A @var{highlighter} list can be of the type
@var{subexp-highlighter} or @var{anchored-highlighter} as described
above.

For example,

@smallexample
;; @r{Highlight occurrences of the word @samp{anchor} in the value}
;; @r{of @code{anchor-face}, and subsequent occurrences of the word}
;; @r{@samp{item} (on the same line) in the value of @code{item-face}.}
("\\<anchor\\>" (0 anchor-face)
                ("\\<item\\>" nil nil (0 item-face)))
@end smallexample

@item (eval . @var{form})
Here @var{form} is an expression to be evaluated the first time
this value of @code{font-lock-keywords} is used in a buffer.
Its value should have one of the forms described in this table.
@end table

@strong{Warning:} Do not design an element of @code{font-lock-keywords}
to match text which spans lines; this does not work reliably.
For details, see @xref{Multiline Font Lock}.

You can use @var{case-fold} in @code{font-lock-defaults} to specify
the value of @code{font-lock-keywords-case-fold-search} which says
whether search-based fontification should be case-insensitive.

@defvar font-lock-keywords-case-fold-search
Non-@code{nil} means that regular expression matching for the sake of
@code{font-lock-keywords} should be case-insensitive.
@end defvar

@node Customizing Keywords
@subsection Customizing Search-Based Fontification

  You can use @code{font-lock-add-keywords} to add additional
search-based fontification rules to a major mode, and
@code{font-lock-remove-keywords} to remove rules.

@defun font-lock-add-keywords mode keywords &optional how
This function adds highlighting @var{keywords}, for the current buffer
or for major mode @var{mode}.  The argument @var{keywords} should be a
list with the same format as the variable @code{font-lock-keywords}.

If @var{mode} is a symbol which is a major mode command name, such as
@code{c-mode}, the effect is that enabling Font Lock mode in
@var{mode} will add @var{keywords} to @code{font-lock-keywords}.
Calling with a non-@code{nil} value of @var{mode} is correct only in
your @file{~/.emacs} file.

If @var{mode} is @code{nil}, this function adds @var{keywords} to
@code{font-lock-keywords} in the current buffer.  This way of calling
@code{font-lock-add-keywords} is usually used in mode hook functions.

By default, @var{keywords} are added at the beginning of
@code{font-lock-keywords}.  If the optional argument @var{how} is
@code{set}, they are used to replace the value of
@code{font-lock-keywords}.  If @var{how} is any other non-@code{nil}
value, they are added at the end of @code{font-lock-keywords}.

Some modes provide specialized support you can use in additional
highlighting patterns.  See the variables
@code{c-font-lock-extra-types}, @code{c++-font-lock-extra-types},
and @code{java-font-lock-extra-types}, for example.

@strong{Warning:} major mode functions must not call
@code{font-lock-add-keywords} under any circumstances, either directly
or indirectly, except through their mode hooks.  (Doing so would lead
to incorrect behavior for some minor modes.)  They should set up their
rules for search-based fontification by setting
@code{font-lock-keywords}.
@end defun

@defun font-lock-remove-keywords mode keywords
This function removes @var{keywords} from @code{font-lock-keywords}
for the current buffer or for major mode @var{mode}.  As in
@code{font-lock-add-keywords}, @var{mode} should be a major mode
command name or @code{nil}.  All the caveats and requirements for
@code{font-lock-add-keywords} apply here too.
@end defun

  For example, this code

@smallexample
(font-lock-add-keywords 'c-mode
 '(("\\<\\(FIXME\\):" 1 font-lock-warning-face prepend)
   ("\\<\\(and\\|or\\|not\\)\\>" . font-lock-keyword-face)))
@end smallexample

@noindent
adds two fontification patterns for C mode: one to fontify the word
@samp{FIXME}, even in comments, and another to fontify the words
@samp{and}, @samp{or} and @samp{not} as keywords.

@noindent
That example affects only C mode proper.  To add the same patterns to
C mode @emph{and} all modes derived from it, do this instead:

@smallexample
(add-hook 'c-mode-hook
 (lambda ()
  (font-lock-add-keywords nil
   '(("\\<\\(FIXME\\):" 1 font-lock-warning-face prepend)
     ("\\<\\(and\\|or\\|not\\)\\>" .
      font-lock-keyword-face)))))
@end smallexample

@node Other Font Lock Variables
@subsection Other Font Lock Variables

  This section describes additional variables that a major mode can
set by means of @var{other-vars} in @code{font-lock-defaults}
(@pxref{Font Lock Basics}).

@defvar font-lock-mark-block-function
If this variable is non-@code{nil}, it should be a function that is
called with no arguments, to choose an enclosing range of text for
refontification for the command @kbd{M-o M-o}
(@code{font-lock-fontify-block}).

The function should report its choice by placing the region around it.
A good choice is a range of text large enough to give proper results,
but not too large so that refontification becomes slow.  Typical values
are @code{mark-defun} for programming modes or @code{mark-paragraph} for
textual modes.
@end defvar

@defvar font-lock-extra-managed-props
This variable specifies additional properties (other than
@code{font-lock-face}) that are being managed by Font Lock mode.  It
is used by @code{font-lock-default-unfontify-region}, which normally
only manages the @code{font-lock-face} property.  If you want Font
Lock to manage other properties as well, you must specify them in a
@var{facespec} in @code{font-lock-keywords} as well as add them to
this list.  @xref{Search-based Fontification}.
@end defvar

@defvar font-lock-fontify-buffer-function
Function to use for fontifying the buffer.  The default value is
@code{font-lock-default-fontify-buffer}.
@end defvar

@defvar font-lock-unfontify-buffer-function
Function to use for unfontifying the buffer.  This is used when
turning off Font Lock mode.  The default value is
@code{font-lock-default-unfontify-buffer}.
@end defvar

@defvar font-lock-fontify-region-function
Function to use for fontifying a region.  It should take two
arguments, the beginning and end of the region, and an optional third
argument @var{verbose}.  If @var{verbose} is non-@code{nil}, the
function should print status messages.  The default value is
@code{font-lock-default-fontify-region}.
@end defvar

@defvar font-lock-unfontify-region-function
Function to use for unfontifying a region.  It should take two
arguments, the beginning and end of the region.  The default value is
@code{font-lock-default-unfontify-region}.
@end defvar

@defun jit-lock-register function &optional contextual
This function tells Font Lock mode to run the Lisp function
@var{function} any time it has to fontify or refontify part of the
current buffer.  It calls @var{function} before calling the default
fontification functions, and gives it two arguments, @var{start} and
@var{end}, which specify the region to be fontified or refontified.

The optional argument @var{contextual}, if non-@code{nil}, forces Font
Lock mode to always refontify a syntactically relevant part of the
buffer, and not just the modified lines.  This argument can usually be
omitted.
@end defun

@defun jit-lock-unregister function
If @var{function} was previously registered as a fontification
function using @code{jit-lock-register}, this function unregisters it.
@end defun

@node Levels of Font Lock
@subsection Levels of Font Lock

  Many major modes offer three different levels of fontification.  You
can define multiple levels by using a list of symbols for @var{keywords}
in @code{font-lock-defaults}.  Each symbol specifies one level of
fontification; it is up to the user to choose one of these levels,
normally by setting @code{font-lock-maximum-decoration} (@pxref{Font
Lock,,, emacs, the GNU Emacs Manual}).  The chosen level's symbol
value is used to initialize @code{font-lock-keywords}.

  Here are the conventions for how to define the levels of
fontification:

@itemize @bullet
@item
Level 1: highlight function declarations, file directives (such as include or
import directives), strings and comments.  The idea is speed, so only
the most important and top-level components are fontified.

@item
Level 2: in addition to level 1, highlight all language keywords,
including type names that act like keywords, as well as named constant
values.  The idea is that all keywords (either syntactic or semantic)
should be fontified appropriately.

@item
Level 3: in addition to level 2, highlight the symbols being defined in
function and variable declarations, and all builtin function names,
wherever they appear.
@end itemize

@node Precalculated Fontification
@subsection Precalculated Fontification

  Some major modes such as @code{list-buffers} and @code{occur}
construct the buffer text programmatically.  The easiest way for them
to support Font Lock mode is to specify the faces of text when they
insert the text in the buffer.

  The way to do this is to specify the faces in the text with the
special text property @code{font-lock-face} (@pxref{Special
Properties}).  When Font Lock mode is enabled, this property controls
the display, just like the @code{face} property.  When Font Lock mode
is disabled, @code{font-lock-face} has no effect on the display.

  It is ok for a mode to use @code{font-lock-face} for some text and
also use the normal Font Lock machinery.  But if the mode does not use
the normal Font Lock machinery, it should not set the variable
@code{font-lock-defaults}.

@node Faces for Font Lock
@subsection Faces for Font Lock
@cindex faces for font lock
@cindex font lock faces

  Font Lock mode can highlight using any face, but Emacs defines several
faces specifically for syntactic highlighting.  These @dfn{Font Lock
faces} are listed below.  They can also be used by major modes for
syntactic highlighting outside of Font Lock mode (@pxref{Major Mode
Conventions}).

  Each of these symbols is both a face name, and a variable whose
default value is the symbol itself.  Thus, the default value of
@code{font-lock-comment-face} is @code{font-lock-comment-face}.

  The faces are listed with descriptions of their typical usage, and in
order of greater to lesser ``prominence''.  If a mode's syntactic
categories do not fit well with the usage descriptions, the faces can be
assigned using the ordering as a guide.

@table @code
@item font-lock-warning-face
@vindex font-lock-warning-face
for a construct that is peculiar, or that greatly changes the meaning of
other text, like @samp{;;;###autoload} in Emacs Lisp and @samp{#error}
in C.

@item font-lock-function-name-face
@vindex font-lock-function-name-face
for the name of a function being defined or declared.

@item font-lock-variable-name-face
@vindex font-lock-variable-name-face
for the name of a variable being defined or declared.

@item font-lock-keyword-face
@vindex font-lock-keyword-face
for a keyword with special syntactic significance, like @samp{for} and
@samp{if} in C.

@item font-lock-comment-face
@vindex font-lock-comment-face
for comments.

@item font-lock-comment-delimiter-face
@vindex font-lock-comment-delimiter-face
for comments delimiters, like @samp{/*} and @samp{*/} in C.  On most
terminals, this inherits from @code{font-lock-comment-face}.

@item font-lock-type-face
@vindex font-lock-type-face
for the names of user-defined data types.

@item font-lock-constant-face
@vindex font-lock-constant-face
for the names of constants, like @samp{NULL} in C.

@item font-lock-builtin-face
@vindex font-lock-builtin-face
for the names of built-in functions.

@item font-lock-preprocessor-face
@vindex font-lock-preprocessor-face
for preprocessor commands.  This inherits, by default, from
@code{font-lock-builtin-face}.

@item font-lock-string-face
@vindex font-lock-string-face
for string constants.

@item font-lock-doc-face
@vindex font-lock-doc-face
for documentation strings in the code.  This inherits, by default, from
@code{font-lock-string-face}.

@item font-lock-negation-char-face
@vindex font-lock-negation-char-face
for easily-overlooked negation characters.
@end table

@node Syntactic Font Lock
@subsection Syntactic Font Lock
@cindex syntactic font lock

Syntactic fontification uses the syntax table to find comments and
string constants (@pxref{Syntax Tables}).  It highlights them using
@code{font-lock-comment-face} and @code{font-lock-string-face}
(@pxref{Faces for Font Lock}), or whatever
@code{font-lock-syntactic-face-function} chooses.  There are several
variables that affect syntactic fontification; you should set them by
means of @code{font-lock-defaults} (@pxref{Font Lock Basics}).

@defvar font-lock-keywords-only
Non-@code{nil} means Font Lock should not do syntactic fontification;
it should only fontify based on @code{font-lock-keywords}.  The normal
way for a mode to set this variable to @code{t} is with
@var{keywords-only} in @code{font-lock-defaults}.
@end defvar

@defvar font-lock-syntax-table
This variable holds the syntax table to use for fontification of
comments and strings.  Specify it using @var{syntax-alist} in
@code{font-lock-defaults}.  If this is @code{nil}, fontification uses
the buffer's syntax table.
@end defvar

@defvar font-lock-beginning-of-syntax-function
If this variable is non-@code{nil}, it should be a function to move
point back to a position that is syntactically at ``top level'' and
outside of strings or comments.  Font Lock uses this when necessary
to get the right results for syntactic fontification.

This function is called with no arguments.  It should leave point at
the beginning of any enclosing syntactic block.  Typical values are
@code{beginning-of-line} (used when the start of the line is known to
be outside a syntactic block), or @code{beginning-of-defun} for
programming modes, or @code{backward-paragraph} for textual modes.

If the value is @code{nil}, Font Lock uses
@code{syntax-begin-function} to move back outside of any comment,
string, or sexp.  This variable is semi-obsolete; we recommend setting
@code{syntax-begin-function} instead.

Specify this variable using @var{syntax-begin} in
@code{font-lock-defaults}.
@end defvar

@defvar font-lock-syntactic-face-function
A function to determine which face to use for a given syntactic
element (a string or a comment).  The function is called with one
argument, the parse state at point returned by
@code{parse-partial-sexp}, and should return a face.  The default
value returns @code{font-lock-comment-face} for comments and
@code{font-lock-string-face} for strings.

This can be used to highlighting different kinds of strings or
comments differently.  It is also sometimes abused together with
@code{font-lock-syntactic-keywords} to highlight constructs that span
multiple lines, but this is too esoteric to document here.

Specify this variable using @var{other-vars} in
@code{font-lock-defaults}.
@end defvar

@node Setting Syntax Properties
@subsection Setting Syntax Properties

  Font Lock mode can be used to update @code{syntax-table} properties
automatically (@pxref{Syntax Properties}).  This is useful in
languages for which a single syntax table by itself is not sufficient.

@defvar font-lock-syntactic-keywords
This variable enables and controls updating @code{syntax-table}
properties by Font Lock.  Its value should be a list of elements of
this form:

@example
(@var{matcher} @var{subexp} @var{syntax} @var{override} @var{laxmatch})
@end example

The parts of this element have the same meanings as in the corresponding
sort of element of @code{font-lock-keywords},

@example
(@var{matcher} @var{subexp} @var{facespec} @var{override} @var{laxmatch})
@end example

However, instead of specifying the value @var{facespec} to use for the
@code{face} property, it specifies the value @var{syntax} to use for
the @code{syntax-table} property.  Here, @var{syntax} can be a string
(as taken by @code{modify-syntax-entry}), a syntax table, a cons cell
(as returned by @code{string-to-syntax}), or an expression whose value
is one of those two types.  @var{override} cannot be @code{prepend} or
@code{append}.

For example, an element of the form:

@example
("\\$\\(#\\)" 1 ".")
@end example

highlights syntactically a hash character when following a dollar
character, with a SYNTAX of @code{"."} (meaning punctuation syntax).
Assuming that the buffer syntax table specifies hash characters to
have comment start syntax, the element will only highlight hash
characters that do not follow dollar characters as comments
syntactically.

An element of the form:

@example
 ("\\('\\).\\('\\)"
  (1 "\"")
  (2 "\""))
@end example

highlights syntactically both single quotes which surround a single
character, with a SYNTAX of @code{"\""} (meaning string quote syntax).
Assuming that the buffer syntax table does not specify single quotes
to have quote syntax, the element will only highlight single quotes of
the form @samp{'@var{c}'} as strings syntactically.  Other forms, such
as @samp{foo'bar} or @samp{'fubar'}, will not be highlighted as
strings.

Major modes normally set this variable with @var{other-vars} in
@code{font-lock-defaults}.
@end defvar

@node Multiline Font Lock
@subsection Multiline Font Lock Constructs
@cindex multiline font lock

  Normally, elements of @code{font-lock-keywords} should not match
across multiple lines; that doesn't work reliably, because Font Lock
usually scans just part of the buffer, and it can miss a multi-line
construct that crosses the line boundary where the scan starts.  (The
scan normally starts at the beginning of a line.)

  Making elements that match multiline constructs work properly has
two aspects: correct @emph{identification} and correct
@emph{rehighlighting}.  The first means that Font Lock finds all
multiline constructs.  The second means that Font Lock will correctly
rehighlight all the relevant text when a multiline construct is
changed---for example, if some of the text that was previously part of
a multiline construct ceases to be part of it.  The two aspects are
closely related, and often getting one of them to work will appear to
make the other also work.  However, for reliable results you must
attend explicitly to both aspects.

  There are three ways to ensure correct identification of multiline
constructs:

@itemize
@item
Add a function to @code{font-lock-extend-region-functions} that does
the @emph{identification} and extends the scan so that the scanned
text never starts or ends in the middle of a multiline construct.
@item
Use the @code{font-lock-fontify-region-function} hook similarly to
extend the scan so that the scanned text never starts or ends in the
middle of a multiline construct.
@item
Somehow identify the multiline construct right when it gets inserted
into the buffer (or at any point after that but before font-lock
tries to highlight it), and mark it with a @code{font-lock-multiline}
which will instruct font-lock not to start or end the scan in the
middle of the construct.
@end itemize

  There are three ways to do rehighlighting of multiline constructs:

@itemize
@item
Place a @code{font-lock-multiline} property on the construct.  This
will rehighlight the whole construct if any part of it is changed.  In
some cases you can do this automatically by setting the
@code{font-lock-multiline} variable, which see.
@item
Make sure @code{jit-lock-contextually} is set and rely on it doing its
job.  This will only rehighlight the part of the construct that
follows the actual change, and will do it after a short delay.
This only works if the highlighting of the various parts of your
multiline construct never depends on text in subsequent lines.
Since @code{jit-lock-contextually} is activated by default, this can
be an attractive solution.
@item
Place a @code{jit-lock-defer-multiline} property on the construct.
This works only if @code{jit-lock-contextually} is used, and with the
same delay before rehighlighting, but like @code{font-lock-multiline},
it also handles the case where highlighting depends on
subsequent lines.
@end itemize

@menu
* Font Lock Multiline::         Marking multiline chunks with a text property.
* Region to Refontify::         Controlling which region gets refontified
                                  after a buffer change.
@end menu

@node Font Lock Multiline
@subsubsection Font Lock Multiline

  One way to ensure reliable rehighlighting of multiline Font Lock
constructs is to put on them the text property @code{font-lock-multiline}.
It should be present and non-@code{nil} for text that is part of a
multiline construct.

  When Font Lock is about to highlight a range of text, it first
extends the boundaries of the range as necessary so that they do not
fall within text marked with the @code{font-lock-multiline} property.
Then it removes any @code{font-lock-multiline} properties from the
range, and highlights it.  The highlighting specification (mostly
@code{font-lock-keywords}) must reinstall this property each time,
whenever it is appropriate.

  @strong{Warning:} don't use the @code{font-lock-multiline} property
on large ranges of text, because that will make rehighlighting slow.

@defvar font-lock-multiline
If the @code{font-lock-multiline} variable is set to @code{t}, Font
Lock will try to add the @code{font-lock-multiline} property
automatically on multiline constructs.  This is not a universal
solution, however, since it slows down Font Lock somewhat.  It can
miss some multiline constructs, or make the property larger or smaller
than necessary.

For elements whose @var{matcher} is a function, the function should
ensure that submatch 0 covers the whole relevant multiline construct,
even if only a small subpart will be highlighted.  It is often just as
easy to add the @code{font-lock-multiline} property by hand.
@end defvar

  The @code{font-lock-multiline} property is meant to ensure proper
refontification; it does not automatically identify new multiline
constructs.  Identifying the requires that Font-Lock operate on large
enough chunks at a time.  This will happen by accident on many cases,
which may give the impression that multiline constructs magically work.
If you set the @code{font-lock-multiline} variable non-@code{nil},
this impression will be even stronger, since the highlighting of those
constructs which are found will be properly updated from then on.
But that does not work reliably.

  To find multiline constructs reliably, you must either manually
place the @code{font-lock-multiline} property on the text before
Font-Lock looks at it, or use
@code{font-lock-fontify-region-function}.

@node Region to Refontify
@subsubsection Region to Fontify after a Buffer Change

  When a buffer is changed, the region that Font Lock refontifies is
by default the smallest sequence of whole lines that spans the change.
While this works well most of the time, sometimes it doesn't---for
example, when a change alters the syntactic meaning of text on an
earlier line.

  You can enlarge (or even reduce) the region to refontify by setting
the following variable:

@defvar font-lock-extend-after-change-region-function
This buffer-local variable is either @code{nil} or a function for
Font-Lock to call to determine the region to scan and fontify.

The function is given three parameters, the standard @var{beg},
@var{end}, and @var{old-len} from @code{after-change-functions}
(@pxref{Change Hooks}).  It should return either a cons of the
beginning and end buffer positions (in that order) of the region to
fontify, or @code{nil} (which means choose the region in the standard
way).  This function needs to preserve point, the match-data, and the
current restriction.  The region it returns may start or end in the
middle of a line.

Since this function is called after every buffer change, it should be
reasonably fast.
@end defvar

@node Auto-Indentation
@section Auto-indentation of code

For programming languages, an important feature of a major mode is to
provide automatic indentation.  This is controlled in Emacs by
@code{indent-line-function} (@pxref{Mode-Specific Indent}).
Writing a good indentation function can be difficult and to a large
extent it is still a black art.

Many major mode authors will start by writing a simple indentation
function that works for simple cases, for example by comparing with the
indentation of the previous text line.  For most programming languages
that are not really line-based, this tends to scale very poorly:
improving such a function to let it handle more diverse situations tends
to become more and more difficult, resulting in the end with a large,
complex, unmaintainable indentation function which nobody dares to touch.

A good indentation function will usually need to actually parse the
text, according to the syntax of the language.  Luckily, it is not
necessary to parse the text in as much detail as would be needed
for a compiler, but on the other hand, the parser embedded in the
indentation code will want to be somewhat friendly to syntactically
incorrect code.

Good maintainable indentation functions usually fall into 2 categories:
either parsing forward from some ``safe'' starting point until the
position of interest, or parsing backward from the position of interest.
Neither of the two is a clearly better choice than the other: parsing
backward is often more difficult than parsing forward because
programming languages are designed to be parsed forward, but for the
purpose of indentation it has the advantage of not needing to
guess a ``safe'' starting point, and it generally enjoys the property
that only a minimum of text will be analyzed to decide the indentation
of a line, so indentation will tend to be unaffected by syntax errors in
some earlier unrelated piece of code.  Parsing forward on the other hand
is usually easier and has the advantage of making it possible to
reindent efficiently a whole region at a time, with a single parse.

Rather than write your own indentation function from scratch, it is
often preferable to try and reuse some existing ones or to rely
on a generic indentation engine.  There are sadly few such
engines.  The CC-mode indentation code (used with C, C++, Java, Awk
and a few other such modes) has been made more generic over the years,
so if your language seems somewhat similar to one of those languages,
you might try to use that engine.  @c FIXME: documentation?
Another one is SMIE which takes an approach in the spirit
of Lisp sexps and adapts it to non-Lisp languages.

@menu
* SMIE::                        A simple minded indentation engine
@end menu

@node SMIE
@subsection Simple Minded Indentation Engine

SMIE is a package that provides a generic navigation and indentation
engine.  Based on a very simple parser using an ``operator precedence
grammar'', it lets major modes extend the sexp-based navigation of Lisp
to non-Lisp languages as well as provide a simple to use but reliable
auto-indentation.

Operator precedence grammar is a very primitive technology for parsing
compared to some of the more common techniques used in compilers.
It has the following characteristics: its parsing power is very limited,
and it is largely unable to detect syntax errors, but it has the
advantage of being algorithmically efficient and able to parse forward
just as well as backward.  In practice that means that SMIE can use it
for indentation based on backward parsing, that it can provide both
@code{forward-sexp} and @code{backward-sexp} functionality, and that it
will naturally work on syntactically incorrect code without any extra
effort.  The downside is that it also means that most programming
languages cannot be parsed correctly using SMIE, at least not without
resorting to some special tricks (@pxref{SMIE Tricks}).

@menu
* SMIE setup::                  SMIE setup and features
* Operator Precedence Grammars::  A very simple parsing technique
* SMIE Grammar::                Defining the grammar of a language
* SMIE Lexer::                  Defining tokens
* SMIE Tricks::                 Working around the parser's limitations
* SMIE Indentation::            Specifying indentation rules
* SMIE Indentation Helpers::    Helper functions for indentation rules
* SMIE Indentation Example::    Sample indentation rules
@end menu

@node SMIE setup
@subsubsection SMIE Setup and Features

SMIE is meant to be a one-stop shop for structural navigation and
various other features which rely on the syntactic structure of code, in
particular automatic indentation.  The main entry point is
@code{smie-setup} which is a function typically called while setting
up a major mode.

@defun smie-setup grammar rules-function &rest keywords
Setup SMIE navigation and indentation.
@var{grammar} is a grammar table generated by @code{smie-prec2->grammar}.
@var{rules-function} is a set of indentation rules for use on
@code{smie-rules-function}.
@var{keywords} are additional arguments, which can include the following
keywords:
@itemize
@item
@code{:forward-token} @var{fun}: Specify the forward lexer to use.
@item
@code{:backward-token} @var{fun}: Specify the backward lexer to use.
@end itemize
@end defun

Calling this function is sufficient to make commands such as
@code{forward-sexp}, @code{backward-sexp}, and @code{transpose-sexps} be
able to properly handle structural elements other than just the paired
parentheses already handled by syntax tables.  For example, if the
provided grammar is precise enough, @code{transpose-sexps} can correctly
transpose the two arguments of a @code{+} operator, taking into account
the precedence rules of the language.

Calling `smie-setup' is also sufficient to make TAB indentation work in
the expected way, extends @code{blink-matching-paren} to apply to
elements like @code{begin...end}, and provides some commands that you
can bind in the major mode keymap.

@deffn Command smie-close-block
This command closes the most recently opened (and not yet closed) block.
@end deffn

@deffn Command smie-down-list &optional arg
This command is like @code{down-list} but it also pays attention to
nesting of tokens other than parentheses, such as @code{begin...end}.
@end deffn

@node Operator Precedence Grammars
@subsubsection Operator Precedence Grammars

SMIE's precedence grammars simply give to each token a pair of
precedences: the left-precedence and the right-precedence.  We say
@code{T1 < T2} if the right-precedence of token @code{T1} is less than
the left-precedence of token @code{T2}.  A good way to read this
@code{<} is as a kind of parenthesis: if we find @code{... T1 something
T2 ...}  then that should be parsed as @code{... T1 (something T2 ...}
rather than as @code{... T1 something) T2 ...}.  The latter
interpretation would be the case if we had @code{T1 > T2}.  If we have
@code{T1 = T2}, it means that token T2 follows token T1 in the same
syntactic construction, so typically we have @code{"begin" = "end"}.
Such pairs of precedences are sufficient to express left-associativity
or right-associativity of infix operators, nesting of tokens like
parentheses and many other cases.

@c Let's leave this undocumented to leave it more open for change!
@c @defvar smie-grammar
@c The value of this variable is an alist specifying the left and right
@c precedence of each token.  It is meant to be initialized by using one of
@c the functions below.
@c @end defvar

@defun smie-prec2->grammar table
This function takes a @emph{prec2} grammar @var{table} and returns an
alist suitable for use in @code{smie-setup}.  The @emph{prec2}
@var{table} is itself meant to be built by one of the functions below.
@end defun

@defun smie-merge-prec2s &rest tables
This function takes several @emph{prec2} @var{tables} and merges them
into a new @emph{prec2} table.
@end defun

@defun smie-precs->prec2 precs
This function builds a @emph{prec2} table from a table of precedences
@var{precs}.  @var{precs} should be a list, sorted by precedence (for
example @code{"+"} will come before @code{"*"}), of elements of the form
@code{(@var{assoc} @var{op} ...)}, where each @var{op} is a token that
acts as an operator; @var{assoc} is their associativity, which can be
either @code{left}, @code{right}, @code{assoc}, or @code{nonassoc}.
All operators in a given element share the same precedence level
and associativity.
@end defun

@defun smie-bnf->prec2 bnf &rest resolvers
This function lets you specify the grammar using a BNF notation.
It accepts a @var{bnf} description of the grammar along with a set of
conflict resolution rules @var{resolvers}, and
returns a @emph{prec2} table.

@var{bnf} is a list of nonterminal definitions of the form
@code{(@var{nonterm} @var{rhs1} @var{rhs2} ...)} where each @var{rhs}
is a (non-empty) list of terminals (aka tokens) or non-terminals.

Not all grammars are accepted:
@itemize
@item
An @var{rhs} cannot be an empty list (an empty list is never needed,
since SMIE allows all non-terminals to match the empty string anyway).
@item
An @var{rhs} cannot have 2 consecutive non-terminals: each pair of
non-terminals needs to be separated by a terminal (aka token).
This is a fundamental limitation of operator precedence grammars.
@end itemize

Additionally, conflicts can occur:
@itemize
@item
The returned @emph{prec2} table holds constraints between pairs of tokens, and
for any given pair only one constraint can be present: T1 < T2,
T1 = T2, or T1 > T2.
@item
A token can be an @code{opener} (something similar to an open-paren),
a @code{closer} (like a close-paren), or @code{neither} of the two
(e.g. an infix operator, or an inner token like @code{"else"}).
@end itemize

Precedence conflicts can be resolved via @var{resolvers}, which
is a list of @emph{precs} tables (see @code{smie-precs->prec2}): for
each precedence conflict, if those @code{precs} tables
specify a particular constraint, then the conflict is resolved by using
this constraint instead, else a conflict is reported and one of the
conflicting constraints is picked arbitrarily and the others are
simply ignored.
@end defun

@node SMIE Grammar
@subsubsection Defining the Grammar of a Language

The usual way to define the SMIE grammar of a language is by
defining a new global variable that holds the precedence table by
giving a set of BNF rules.
For example, the grammar definition for a small Pascal-like language
could look like:
@example
@group
(require 'smie)
(defvar sample-smie-grammar
  (smie-prec2->grammar
   (smie-bnf->prec2
@end group
@group
    '((id)
      (inst ("begin" insts "end")
            ("if" exp "then" inst "else" inst)
            (id ":=" exp)
            (exp))
      (insts (insts ";" insts) (inst))
      (exp (exp "+" exp)
           (exp "*" exp)
           ("(" exps ")"))
      (exps (exps "," exps) (exp)))
@end group
@group
    '((assoc ";"))
    '((assoc ","))
    '((assoc "+") (assoc "*")))))
@end group
@end example

@noindent
A few things to note:

@itemize
@item
The above grammar does not explicitly mention the syntax of function
calls: SMIE will automatically allow any sequence of sexps, such as
identifiers, balanced parentheses, or @code{begin ... end} blocks
to appear anywhere anyway.
@item
The grammar category @code{id} has no right hand side: this does not
mean that it can match only the empty string, since as mentioned any
sequence of sexps can appear anywhere anyway.
@item
Because non terminals cannot appear consecutively in the BNF grammar, it
is difficult to correctly handle tokens that act as terminators, so the
above grammar treats @code{";"} as a statement @emph{separator} instead,
which SMIE can handle very well.
@item
Separators used in sequences (such as @code{","} and @code{";"} above)
are best defined with BNF rules such as @code{(foo (foo "separator" foo) ...)}
which generate precedence conflicts which are then resolved by giving
them an explicit @code{(assoc "separator")}.
@item
The @code{("(" exps ")")} rule was not needed to pair up parens, since
SMIE will pair up any characters that are marked as having paren syntax
in the syntax table.  What this rule does instead (together with the
definition of @code{exps}) is to make it clear that @code{","} should
not appear outside of parentheses.
@item
Rather than have a single @emph{precs} table to resolve conflicts, it is
preferable to have several tables, so as to let the BNF part of the
grammar specify relative precedences where possible.
@item
Unless there is a very good reason to prefer @code{left} or
@code{right}, it is usually preferable to mark operators as associative,
using @code{assoc}.  For that reason @code{"+"} and @code{"*"} are
defined above as @code{assoc}, although the language defines them
formally as left associative.
@end itemize

@node SMIE Lexer
@subsubsection Defining Tokens

SMIE comes with a predefined lexical analyzer which uses syntax tables
in the following way: any sequence of characters that have word or
symbol syntax is considered a token, and so is any sequence of
characters that have punctuation syntax.  This default lexer is
often a good starting point but is rarely actually correct for any given
language.  For example, it will consider @code{"2,+3"} to be composed
of 3 tokens: @code{"2"}, @code{",+"}, and @code{"3"}.

To describe the lexing rules of your language to SMIE, you need
2 functions, one to fetch the next token, and another to fetch the
previous token.  Those functions will usually first skip whitespace and
comments and then look at the next chunk of text to see if it
is a special token.  If so it should skip the token and
return a description of this token.  Usually this is simply the string
extracted from the buffer, but it can be anything you want.
For example:
@example
@group
(defvar sample-keywords-regexp
  (regexp-opt '("+" "*" "," ";" ">" ">=" "<" "<=" ":=" "=")))
@end group
@group
(defun sample-smie-forward-token ()
  (forward-comment (point-max))
  (cond
   ((looking-at sample-keywords-regexp)
    (goto-char (match-end 0))
    (match-string-no-properties 0))
   (t (buffer-substring-no-properties
       (point)
       (progn (skip-syntax-forward "w_")
              (point))))))
@end group
@group
(defun sample-smie-backward-token ()
  (forward-comment (- (point)))
  (cond
   ((looking-back sample-keywords-regexp (- (point) 2) t)
    (goto-char (match-beginning 0))
    (match-string-no-properties 0))
   (t (buffer-substring-no-properties
       (point)
       (progn (skip-syntax-backward "w_")
              (point))))))
@end group
@end example

Notice how those lexers return the empty string when in front of
parentheses.  This is because SMIE automatically takes care of the
parentheses defined in the syntax table.  More specifically if the lexer
returns nil or an empty string, SMIE tries to handle the corresponding
text as a sexp according to syntax tables.

@node SMIE Tricks
@subsubsection Living With a Weak Parser

The parsing technique used by SMIE does not allow tokens to behave
differently in different contexts.  For most programming languages, this
manifests itself by precedence conflicts when converting the
BNF grammar.

Sometimes, those conflicts can be worked around by expressing the
grammar slightly differently.  For example, for Modula-2 it might seem
natural to have a BNF grammar that looks like this:

@example
  ...
  (inst ("IF" exp "THEN" insts "ELSE" insts "END")
        ("CASE" exp "OF" cases "END")
        ...)
  (cases (cases "|" cases) (caselabel ":" insts) ("ELSE" insts))
  ...
@end example

But this will create conflicts for @code{"ELSE"}: on the one hand, the
IF rule implies (among many other things) that @code{"ELSE" = "END"};
but on the other hand, since @code{"ELSE"} appears within @code{cases},
which appears left of @code{"END"}, we also have @code{"ELSE" > "END"}.
We can solve the conflict either by using:
@example
  ...
  (inst ("IF" exp "THEN" insts "ELSE" insts "END")
        ("CASE" exp "OF" cases "END")
        ("CASE" exp "OF" cases "ELSE" insts "END")
        ...)
  (cases (cases "|" cases) (caselabel ":" insts))
  ...
@end example
or
@example
  ...
  (inst ("IF" exp "THEN" else "END")
        ("CASE" exp "OF" cases "END")
        ...)
  (else (insts "ELSE" insts))
  (cases (cases "|" cases) (caselabel ":" insts) (else))
  ...
@end example

Reworking the grammar to try and solve conflicts has its downsides, tho,
because SMIE assumes that the grammar reflects the logical structure of
the code, so it is preferable to keep the BNF closer to the intended
abstract syntax tree.

Other times, after careful consideration you may conclude that those
conflicts are not serious and simply resolve them via the
@var{resolvers} argument of @code{smie-bnf->prec2}.  Usually this is
because the grammar is simply ambiguous: the conflict does not affect
the set of programs described by the grammar, but only the way those
programs are parsed.  This is typically the case for separators and
associative infix operators, where you want to add a resolver like
@code{'((assoc "|"))}.  Another case where this can happen is for the
classic @emph{dangling else} problem, where you will use @code{'((assoc
"else" "then"))}.  It can also happen for cases where the conflict is
real and cannot really be resolved, but it is unlikely to pose a problem
in practice.

Finally, in many cases some conflicts will remain despite all efforts to
restructure the grammar.  Do not despair: while the parser cannot be
made more clever, you can make the lexer as smart as you want.  So, the
solution is then to look at the tokens involved in the conflict and to
split one of those tokens into 2 (or more) different tokens.  E.g. if
the grammar needs to distinguish between two incompatible uses of the
token @code{"begin"}, make the lexer return different tokens (say
@code{"begin-fun"} and @code{"begin-plain"}) depending on which kind of
@code{"begin"} it finds.  This pushes the work of distinguishing the
different cases to the lexer, which will thus have to look at the
surrounding text to find ad-hoc clues.

@node SMIE Indentation
@subsubsection Specifying Indentation Rules

Based on the provided grammar, SMIE will be able to provide automatic
indentation without any extra effort.  But in practice, this default
indentation style will probably not be good enough.  You will want to
tweak it in many different cases.

SMIE indentation is based on the idea that indentation rules should be
as local as possible.  To this end, it relies on the idea of
@emph{virtual} indentation, which is the indentation that a particular
program point would have if it were at the beginning of a line.
Of course, if that program point is indeed at the beginning of a line,
its virtual indentation is its current indentation.  But if not, then
SMIE uses the indentation algorithm to compute the virtual indentation
of that point.  Now in practice, the virtual indentation of a program
point does not have to be identical to the indentation it would have if
we inserted a newline before it.  To see how this works, the SMIE rule
for indentation after a @code{@{} in C does not care whether the
@code{@{} is standing on a line of its own or is at the end of the
preceding line.  Instead, these different cases are handled in the
indentation rule that decides how to indent before a @code{@{}.

Another important concept is the notion of @emph{parent}: The
@emph{parent} of a token, is the head token of the nearest enclosing
syntactic construct.  For example, the parent of an @code{else} is the
@code{if} to which it belongs, and the parent of an @code{if}, in turn,
is the lead token of the surrounding construct.  The command
@code{backward-sexp} jumps from a token to its parent, but there are
some caveats: for @emph{openers} (tokens which start a construct, like
@code{if}), you need to start with point before the token, while for
others you need to start with point after the token.
@code{backward-sexp} stops with point before the parent token if that is
the @emph{opener} of the token of interest, and otherwise it stops with
point after the parent token.

SMIE indentation rules are specified using a function that takes two
arguments @var{method} and @var{arg} where the meaning of @var{arg} and the
expected return value depend on @var{method}.

@var{method} can be:
@itemize
@item
@code{:after}, in which case @var{arg} is a token and the function
should return the @var{offset} to use for indentation after @var{arg}.
@item
@code{:before}, in which case @var{arg} is a token and the function
should return the @var{offset} to use to indent @var{arg} itself.
@item
@code{:elem}, in which case the function should return either the offset
to use to indent function arguments (if @var{arg} is the symbol
@code{arg}) or the basic indentation step (if @var{arg} is the symbol
@code{basic}).
@item
@code{:list-intro}, in which case @var{arg} is a token and the function
should return non-@code{nil} if the token is followed by a list of
expressions (not separated by any token) rather than an expression.
@end itemize

When @var{arg} is a token, the function is called with point just before
that token.  A return value of nil always means to fallback on the
default behavior, so the function should return nil for arguments it
does not expect.

@var{offset} can be:
@itemize
@item
@code{nil}: use the default indentation rule.
@item
@code{(column . @var{column})}: indent to column @var{column}.
@item
@var{number}: offset by @var{number}, relative to a base token which is
the current token for @code{:after} and its parent for @code{:before}.
@end itemize

@node SMIE Indentation Helpers
@subsubsection Helper Functions for Indentation Rules

SMIE provides various functions designed specifically for use in the
indentation rules function (several of those functions break if used in
another context).  These functions all start with the prefix
@code{smie-rule-}.

@defun smie-rule-bolp
Return non-@code{nil} if the current token is the first on the line.
@end defun

@defun smie-rule-hanging-p
Return non-@code{nil} if the current token is @emph{hanging}.
A token is @emph{hanging} if it is the last token on the line
and if it is preceded by other tokens: a lone token on a line is not
hanging.
@end defun

@defun smie-rule-next-p &rest tokens
Return non-@code{nil} if the next token is among @var{tokens}.
@end defun

@defun smie-rule-prev-p &rest tokens
Return non-@code{nil} if the previous token is among @var{tokens}.
@end defun

@defun smie-rule-parent-p &rest parents
Return non-@code{nil} if the current token's parent is among @var{parents}.
@end defun

@defun smie-rule-sibling-p
Return non-nil if the current token's parent is actually a sibling.
This is the case for example when the parent of a @code{","} is just the
previous @code{","}.
@end defun

@defun smie-rule-parent &optional offset
Return the proper offset to align the current token with the parent.
If non-@code{nil}, @var{offset} should be an integer giving an
additional offset to apply.
@end defun

@defun smie-rule-separator method
Indent current token as a @emph{separator}.

By @emph{separator}, we mean here a token whose sole purpose is to
separate various elements within some enclosing syntactic construct, and
which does not have any semantic significance in itself (i.e. it would
typically not exist as a node in an abstract syntax tree).

Such a token is expected to have an associative syntax and be closely
tied to its syntactic parent.  Typical examples are @code{","} in lists
of arguments (enclosed inside parentheses), or @code{";"} in sequences
of instructions (enclosed in a @code{@{...@}} or @code{begin...end}
block).

@var{method} should be the method name that was passed to
`smie-rules-function'.
@end defun

@node SMIE Indentation Example
@subsubsection Sample Indentation Rules

Here is an example of an indentation function:

@example
(defun sample-smie-rules (kind token)
  (pcase (cons kind token)
    (`(:elem . basic) sample-indent-basic)
    (`(,_ . ",") (smie-rule-separator kind))
    (`(:after . ":=") sample-indent-basic)
    (`(:before . ,(or `"begin" `"(" `"@{")))
     (if (smie-rule-hanging-p) (smie-rule-parent)))
    (`(:before . "if")
     (and (not (smie-rule-bolp)) (smie-rule-prev-p "else")
          (smie-rule-parent)))))
@end example

@noindent
A few things to note:

@itemize
@item
The first case indicates the basic indentation increment to use.
If @code{sample-indent-basic} is nil, then SMIE uses the global
setting @code{smie-indent-basic}.  The major mode could have set
@code{smie-indent-basic} buffer-locally instead, but that
is discouraged.

@item
The rule for the token @code{","} make SMIE try to be more clever when
the comma separator is placed at the beginning of lines.  It tries to
outdent the separator so as to align the code after the comma; for
example:

@example
x = longfunctionname (
        arg1
      , arg2
    );
@end example

@item
The rule for indentation after @code{":="} exists because otherwise
SMIE would treat @code{":="} as an infix operator and would align the
right argument with the left one.

@item
The rule for indentation before @code{"begin"} is an example of the use
of virtual indentation:  This rule is used only when @code{"begin"} is
hanging, which can happen only when @code{"begin"} is not at the
beginning of a line.  So this is not used when indenting
@code{"begin"} itself but only when indenting something relative to this
@code{"begin"}.  Concretely, this rule changes the indentation from:

@example
    if x > 0 then begin
            dosomething(x);
        end
@end example
to
@example
    if x > 0 then begin
        dosomething(x);
    end
@end example

@item
The rule for indentation before @code{"if"} is similar to the one for
@code{"begin"}, but where the purpose is to treat @code{"else if"}
as a single unit, so as to align a sequence of tests rather than indent
each test further to the right.  This function does this only in the
case where the @code{"if"} is not placed on a separate line, hence the
@code{smie-rule-bolp} test.

If we know that the @code{"else"} is always aligned with its @code{"if"}
and is always at the beginning of a line, we can use a more efficient
rule:
@example
((equal token "if")
 (and (not (smie-rule-bolp)) (smie-rule-prev-p "else")
      (save-excursion
        (sample-smie-backward-token)  ;Jump before the "else".
        (cons 'column (current-column)))))
@end example

The advantage of this formulation is that it reuses the indentation of
the previous @code{"else"}, rather than going all the way back to the
first @code{"if"} of the sequence.
@end itemize

@node Desktop Save Mode
@section Desktop Save Mode
@cindex desktop save mode

@dfn{Desktop Save Mode} is a feature to save the state of Emacs from
one session to another.  The user-level commands for using Desktop
Save Mode are described in the GNU Emacs Manual (@pxref{Saving Emacs
Sessions,,, emacs, the GNU Emacs Manual}).  Modes whose buffers visit
a file, don't have to do anything to use this feature.

For buffers not visiting a file to have their state saved, the major
mode must bind the buffer local variable @code{desktop-save-buffer} to
a non-@code{nil} value.

@defvar desktop-save-buffer
If this buffer-local variable is non-@code{nil}, the buffer will have
its state saved in the desktop file at desktop save.  If the value is
a function, it is called at desktop save with argument
@var{desktop-dirname}, and its value is saved in the desktop file along
with the state of the buffer for which it was called.  When file names
are returned as part of the auxiliary information, they should be
formatted using the call

@example
(desktop-file-name @var{file-name} @var{desktop-dirname})
@end example

@end defvar

For buffers not visiting a file to be restored, the major mode must
define a function to do the job, and that function must be listed in
the alist @code{desktop-buffer-mode-handlers}.

@defvar desktop-buffer-mode-handlers
Alist with elements

@example
(@var{major-mode} . @var{restore-buffer-function})
@end example

The function @var{restore-buffer-function} will be called with
argument list

@example
(@var{buffer-file-name} @var{buffer-name} @var{desktop-buffer-misc})
@end example

and it should return the restored buffer.
Here @var{desktop-buffer-misc} is the value returned by the function
optionally bound to @code{desktop-save-buffer}.
@end defvar

@ignore
   Local Variables:
   fill-column: 72
   End:
@end ignore