Merge from trunk after 24.3 was moved to emacs-24 branch.

[gnu-emacs] / doc / lispref / markers.texi

@c -*-texinfo-*-
@c This is part of the GNU Emacs Lisp Reference Manual.
@c Copyright (C) 1990-1995, 1998-1999, 2001-2012 Free Software Foundation, Inc.
@c See the file elisp.texi for copying conditions.
@node Markers
@chapter Markers
@cindex markers

  A @dfn{marker} is a Lisp object used to specify a position in a buffer
relative to the surrounding text.  A marker changes its offset from the
beginning of the buffer automatically whenever text is inserted or
deleted, so that it stays with the two characters on either side of it.

@menu
* Overview of Markers::      The components of a marker, and how it relocates.
* Predicates on Markers::    Testing whether an object is a marker.
* Creating Markers::         Making empty markers or markers at certain places.
* Information from Markers:: Finding the marker's buffer or character position.
* Marker Insertion Types::   Two ways a marker can relocate when you
                               insert where it points.
* Moving Markers::           Moving the marker to a new buffer or position.
* The Mark::                 How "the mark" is implemented with a marker.
* The Region::               How to access "the region".
@end menu

@node Overview of Markers
@section Overview of Markers

  A marker specifies a buffer and a position in that buffer.  A
marker can be used to represent a position in functions that
require one, just as an integer could be used.  In that case, the
marker's buffer is normally ignored.  Of course, a marker used in this
way usually points to a position in the buffer that the function
operates on, but that is entirely the programmer's responsibility.
@xref{Positions}, for a complete description of positions.

  A marker has three attributes: the marker position, the marker
buffer, and the insertion type.  The marker position is an integer
that is equivalent (at a given time) to the marker as a position in
that buffer.  But the marker's position value can change during
the life of the marker, and often does.  Insertion and deletion of
text in the buffer relocate the marker.  The idea is that a marker
positioned between two characters remains between those two characters
despite insertion and deletion elsewhere in the buffer.  Relocation
changes the integer equivalent of the marker.

@cindex marker relocation
  Deleting text around a marker's position leaves the marker between the
characters immediately before and after the deleted text.  Inserting
text at the position of a marker normally leaves the marker either in
front of or after the new text, depending on the marker's @dfn{insertion
type} (@pxref{Marker Insertion Types})---unless the insertion is done
with @code{insert-before-markers} (@pxref{Insertion}).

@cindex marker garbage collection
  Insertion and deletion in a buffer must check all the markers and
relocate them if necessary.  This slows processing in a buffer with a
large number of markers.  For this reason, it is a good idea to make a
marker point nowhere if you are sure you don't need it any more.
Markers that can no longer be accessed are eventually removed
(@pxref{Garbage Collection}).

@cindex markers as numbers
  Because it is common to perform arithmetic operations on a marker
position, most of these operations (including @code{+} and
@code{-}) accept markers as arguments.  In such cases, the marker
stands for its current position.

Here are examples of creating markers, setting markers, and moving point
to markers:

@example
@group
;; @r{Make a new marker that initially does not point anywhere:}
(setq m1 (make-marker))
     @result{} #<marker in no buffer>
@end group

@group
;; @r{Set @code{m1} to point between the 99th and 100th characters}
;;   @r{in the current buffer:}
(set-marker m1 100)
     @result{} #<marker at 100 in markers.texi>
@end group

@group
;; @r{Now insert one character at the beginning of the buffer:}
(goto-char (point-min))
     @result{} 1
(insert "Q")
     @result{} nil
@end group

@group
;; @r{@code{m1} is updated appropriately.}
m1
     @result{} #<marker at 101 in markers.texi>
@end group

@group
;; @r{Two markers that point to the same position}
;;   @r{are not @code{eq}, but they are @code{equal}.}
(setq m2 (copy-marker m1))
     @result{} #<marker at 101 in markers.texi>
(eq m1 m2)
     @result{} nil
(equal m1 m2)
     @result{} t
@end group

@group
;; @r{When you are finished using a marker, make it point nowhere.}
(set-marker m1 nil)
     @result{} #<marker in no buffer>
@end group
@end example

@node Predicates on Markers
@section Predicates on Markers

  You can test an object to see whether it is a marker, or whether it is
either an integer or a marker.  The latter test is useful in connection
with the arithmetic functions that work with both markers and integers.

@defun markerp object
This function returns @code{t} if @var{object} is a marker, @code{nil}
otherwise.  Note that integers are not markers, even though many
functions will accept either a marker or an integer.
@end defun

@defun integer-or-marker-p object
This function returns @code{t} if @var{object} is an integer or a marker,
@code{nil} otherwise.
@end defun

@defun number-or-marker-p object
This function returns @code{t} if @var{object} is a number (either
integer or floating point) or a marker, @code{nil} otherwise.
@end defun

@node Creating Markers
@section Functions that Create Markers

  When you create a new marker, you can make it point nowhere, or point
to the present position of point, or to the beginning or end of the
accessible portion of the buffer, or to the same place as another given
marker.

The next four functions all return markers with insertion type
@code{nil}.  @xref{Marker Insertion Types}.

@defun make-marker
This function returns a newly created marker that does not point
anywhere.

@example
@group
(make-marker)
     @result{} #<marker in no buffer>
@end group
@end example
@end defun

@defun point-marker
This function returns a new marker that points to the present position
of point in the current buffer.  @xref{Point}.  For an example, see
@code{copy-marker}, below.
@end defun

@defun point-min-marker
This function returns a new marker that points to the beginning of the
accessible portion of the buffer.  This will be the beginning of the
buffer unless narrowing is in effect.  @xref{Narrowing}.
@end defun

@defun point-max-marker
This function returns a new marker that points to the end of the
accessible portion of the buffer.  This will be the end of the buffer
unless narrowing is in effect.  @xref{Narrowing}.

Here are examples of this function and @code{point-min-marker}, shown in
a buffer containing a version of the source file for the text of this
chapter.

@example
@group
(point-min-marker)
     @result{} #<marker at 1 in markers.texi>
(point-max-marker)
     @result{} #<marker at 24080 in markers.texi>
@end group

@group
(narrow-to-region 100 200)
     @result{} nil
@end group
@group
(point-min-marker)
     @result{} #<marker at 100 in markers.texi>
@end group
@group
(point-max-marker)
     @result{} #<marker at 200 in markers.texi>
@end group
@end example
@end defun

@defun copy-marker &optional marker-or-integer insertion-type
If passed a marker as its argument, @code{copy-marker} returns a
new marker that points to the same place and the same buffer as does
@var{marker-or-integer}.  If passed an integer as its argument,
@code{copy-marker} returns a new marker that points to position
@var{marker-or-integer} in the current buffer.

The new marker's insertion type is specified by the argument
@var{insertion-type}.  @xref{Marker Insertion Types}.

If passed an integer argument less than 1, @code{copy-marker} returns a
new marker that points to the beginning of the current buffer.  If
passed an integer argument greater than the length of the buffer,
@code{copy-marker} returns a new marker that points to the end of the
buffer.

@example
@group
(copy-marker 0)
     @result{} #<marker at 1 in markers.texi>
@end group

@group
(copy-marker 90000)
     @result{} #<marker at 24080 in markers.texi>
@end group
@end example

An error is signaled if @var{marker} is neither a marker nor an
integer.
@end defun

  Two distinct markers are considered @code{equal} (even though not
@code{eq}) to each other if they have the same position and buffer, or
if they both point nowhere.

@example
@group
(setq p (point-marker))
     @result{} #<marker at 2139 in markers.texi>
@end group

@group
(setq q (copy-marker p))
     @result{} #<marker at 2139 in markers.texi>
@end group

@group
(eq p q)
     @result{} nil
@end group

@group
(equal p q)
     @result{} t
@end group
@end example

@node Information from Markers
@section Information from Markers

  This section describes the functions for accessing the components of a
marker object.

@defun marker-position marker
This function returns the position that @var{marker} points to, or
@code{nil} if it points nowhere.
@end defun

@defun marker-buffer marker
This function returns the buffer that @var{marker} points into, or
@code{nil} if it points nowhere.

@example
@group
(setq m (make-marker))
     @result{} #<marker in no buffer>
@end group
@group
(marker-position m)
     @result{} nil
@end group
@group
(marker-buffer m)
     @result{} nil
@end group

@group
(set-marker m 3770 (current-buffer))
     @result{} #<marker at 3770 in markers.texi>
@end group
@group
(marker-buffer m)
     @result{} #<buffer markers.texi>
@end group
@group
(marker-position m)
     @result{} 3770
@end group
@end example
@end defun

@node Marker Insertion Types
@section Marker Insertion Types

@cindex insertion type of a marker
  When you insert text directly at the place where a marker points,
there are two possible ways to relocate that marker: it can point before
the inserted text, or point after it.  You can specify which one a given
marker should do by setting its @dfn{insertion type}.  Note that use of
@code{insert-before-markers} ignores markers' insertion types, always
relocating a marker to point after the inserted text.

@defun set-marker-insertion-type marker type
This function sets the insertion type of marker @var{marker} to
@var{type}.  If @var{type} is @code{t}, @var{marker} will advance when
text is inserted at its position.  If @var{type} is @code{nil},
@var{marker} does not advance when text is inserted there.
@end defun

@defun marker-insertion-type marker
This function reports the current insertion type of @var{marker}.
@end defun

Most functions that create markers, without an argument allowing to
specify the insertion type, create them with insertion type
@code{nil}.  Also, the mark has, by default, insertion type
@code{nil}.

@node Moving Markers
@section Moving Marker Positions

  This section describes how to change the position of an existing
marker.  When you do this, be sure you know whether the marker is used
outside of your program, and, if so, what effects will result from
moving it---otherwise, confusing things may happen in other parts of
Emacs.

@defun set-marker marker position &optional buffer
This function moves @var{marker} to @var{position}
in @var{buffer}.  If @var{buffer} is not provided, it defaults to
the current buffer.

If @var{position} is less than 1, @code{set-marker} moves @var{marker}
to the beginning of the buffer.  If @var{position} is greater than the
size of the buffer (@pxref{Point}), @code{set-marker} moves marker to
the end of the buffer.  If @var{position} is @code{nil} or a marker
that points nowhere, then @var{marker} is set to point nowhere.

The value returned is @var{marker}.

@example
@group
(setq m (point-marker))
     @result{} #<marker at 4714 in markers.texi>
@end group
@group
(set-marker m 55)
     @result{} #<marker at 55 in markers.texi>
@end group
@group
(setq b (get-buffer "foo"))
     @result{} #<buffer foo>
@end group
@group
(set-marker m 0 b)
     @result{} #<marker at 1 in foo>
@end group
@end example
@end defun

@defun move-marker marker position &optional buffer
This is another name for @code{set-marker}.
@end defun

@node The Mark
@section The Mark
@cindex mark, the
@cindex mark ring

  Each buffer has a special marker, which is designated @dfn{the
mark}.  When a buffer is newly created, this marker exists but does
not point anywhere; this means that the mark ``doesn't exist'' in that
buffer yet.  Subsequent commands can set the mark.

  The mark specifies a position to bound a range of text for many
commands, such as @code{kill-region} and @code{indent-rigidly}.  These
commands typically act on the text between point and the mark, which
is called the @dfn{region}.  If you are writing a command that
operates on the region, don't examine the mark directly; instead, use
@code{interactive} with the @samp{r} specification.  This provides the
values of point and the mark as arguments to the command in an
interactive call, but permits other Lisp programs to specify arguments
explicitly.  @xref{Interactive Codes}.

  Some commands set the mark as a side-effect.  Commands should do
this only if it has a potential use to the user, and never for their
own internal purposes.  For example, the @code{replace-regexp} command
sets the mark to the value of point before doing any replacements,
because this enables the user to move back there conveniently after
the replace is finished.

  Once the mark ``exists'' in a buffer, it normally never ceases to
exist.  However, it may become @dfn{inactive}, if Transient Mark mode
is enabled.  The buffer-local variable @code{mark-active}, if
non-@code{nil}, means that the mark is active.  A command can call the
function @code{deactivate-mark} to deactivate the mark directly, or it
can request deactivation of the mark upon return to the editor command
loop by setting the variable @code{deactivate-mark} to a
non-@code{nil} value.

  If Transient Mark mode is enabled, certain editing commands that
normally apply to text near point, apply instead to the region when
the mark is active.  This is the main motivation for using Transient
Mark mode.  (Another is that this enables highlighting of the region
when the mark is active.  @xref{Display}.)

  In addition to the mark, each buffer has a @dfn{mark ring} which is a
list of markers containing previous values of the mark.  When editing
commands change the mark, they should normally save the old value of the
mark on the mark ring.  The variable @code{mark-ring-max} specifies the
maximum number of entries in the mark ring; once the list becomes this
long, adding a new element deletes the last element.

  There is also a separate global mark ring, but that is used only in a
few particular user-level commands, and is not relevant to Lisp
programming.  So we do not describe it here.

@defun mark &optional force
@cindex current buffer mark
This function returns the current buffer's mark position as an integer,
or @code{nil} if no mark has ever been set in this buffer.

If Transient Mark mode is enabled, and @code{mark-even-if-inactive} is
@code{nil}, @code{mark} signals an error if the mark is inactive.
However, if @var{force} is non-@code{nil}, then @code{mark} disregards
inactivity of the mark, and returns the mark position (or @code{nil})
anyway.
@end defun

@defun mark-marker
This function returns the marker that represents the current buffer's
mark.  It is not a copy, it is the marker used internally.  Therefore,
changing this marker's position will directly affect the buffer's
mark.  Don't do that unless that is the effect you want.

@example
@group
(setq m (mark-marker))
     @result{} #<marker at 3420 in markers.texi>
@end group
@group
(set-marker m 100)
     @result{} #<marker at 100 in markers.texi>
@end group
@group
(mark-marker)
     @result{} #<marker at 100 in markers.texi>
@end group
@end example

Like any marker, this marker can be set to point at any buffer you
like.  If you make it point at any buffer other than the one of which
it is the mark, it will yield perfectly consistent, but rather odd,
results.  We recommend that you not do it!
@end defun

@defun set-mark position
This function sets the mark to @var{position}, and activates the mark.
The old value of the mark is @emph{not} pushed onto the mark ring.

@strong{Please note:} Use this function only if you want the user to
see that the mark has moved, and you want the previous mark position to
be lost.  Normally, when a new mark is set, the old one should go on the
@code{mark-ring}.  For this reason, most applications should use
@code{push-mark} and @code{pop-mark}, not @code{set-mark}.

Novice Emacs Lisp programmers often try to use the mark for the wrong
purposes.  The mark saves a location for the user's convenience.  An
editing command should not alter the mark unless altering the mark is
part of the user-level functionality of the command.  (And, in that
case, this effect should be documented.)  To remember a location for
internal use in the Lisp program, store it in a Lisp variable.  For
example:

@example
@group
(let ((beg (point)))
  (forward-line 1)
  (delete-region beg (point))).
@end group
@end example
@end defun

@defun push-mark &optional position nomsg activate
This function sets the current buffer's mark to @var{position}, and
pushes a copy of the previous mark onto @code{mark-ring}.  If
@var{position} is @code{nil}, then the value of point is used.
@c Doesn't seem relevant.
@c @code{push-mark} returns @code{nil}.

The function @code{push-mark} normally @emph{does not} activate the
mark.  To do that, specify @code{t} for the argument @var{activate}.

A @samp{Mark set} message is displayed unless @var{nomsg} is
non-@code{nil}.
@end defun

@defun pop-mark
This function pops off the top element of @code{mark-ring} and makes
that mark become the buffer's actual mark.  This does not move point in
the buffer, and it does nothing if @code{mark-ring} is empty.  It
deactivates the mark.
@c
@c Seems even less relevant.
@c The return value is not meaningful.
@end defun

@defopt transient-mark-mode
This variable, if non-@code{nil}, enables Transient Mark mode.  In
Transient Mark mode, every buffer-modifying primitive sets
@code{deactivate-mark}.  As a consequence, most commands that modify
the buffer also deactivate the mark.

When Transient Mark mode is enabled and the mark is active, many
commands that normally apply to the text near point instead apply to
the region.  Such commands should use the function @code{use-region-p}
to test whether they should operate on the region.  @xref{The Region}.

Lisp programs can set @code{transient-mark-mode} to non-@code{nil},
non-@code{t} values to enable Transient Mark mode temporarily.  If the
value is @code{lambda}, Transient Mark mode is automatically turned
off after any action, such as buffer modification, that would normally
deactivate the mark.  If the value is @w{@code{(only . @var{oldval})}},
then @code{transient-mark-mode} is set to the value @var{oldval} after
any subsequent command that moves point and is not shift-translated
(@pxref{Key Sequence Input, shift-translation}), or after any other
action that would normally deactivate the mark.
@end defopt

@defopt mark-even-if-inactive
If this is non-@code{nil}, Lisp programs and the Emacs user can use the
mark even when it is inactive.  This option affects the behavior of
Transient Mark mode.  When the option is non-@code{nil}, deactivation of
the mark turns off region highlighting, but commands that use the mark
behave as if the mark were still active.
@end defopt

@defvar deactivate-mark
If an editor command sets this variable non-@code{nil}, then the editor
command loop deactivates the mark after the command returns (if
Transient Mark mode is enabled).  All the primitives that change the
buffer set @code{deactivate-mark}, to deactivate the mark when the
command is finished.

To write Lisp code that modifies the buffer without causing
deactivation of the mark at the end of the command, bind
@code{deactivate-mark} to @code{nil} around the code that does the
modification.  For example:

@example
(let (deactivate-mark)
  (insert " "))
@end example
@end defvar

@defun deactivate-mark &optional force
If Transient Mark mode is enabled or @var{force} is non-@code{nil},
this function deactivates the mark and runs the normal hook
@code{deactivate-mark-hook}.  Otherwise, it does nothing.
@end defun

@defvar mark-active
The mark is active when this variable is non-@code{nil}.  This
variable is always buffer-local in each buffer.  Do @emph{not} use the
value of this variable to decide whether a command that normally
operates on text near point should operate on the region instead.  Use
the function @code{use-region-p} for that (@pxref{The Region}).
@end defvar

@defvar activate-mark-hook
@defvarx deactivate-mark-hook
These normal hooks are run, respectively, when the mark becomes active
and when it becomes inactive.  The hook @code{activate-mark-hook} is
also run at the end of the command loop if the mark is active and it
is possible that the region may have changed.
@ignore
This piece of command_loop_1, run unless deactivating the mark:
  if (current_buffer != prev_buffer || MODIFF != prev_modiff)
    {
      Lisp_Object hook = intern ("activate-mark-hook");
      Frun_hooks (1, &hook);
    }
@end ignore
@end defvar

@defun handle-shift-selection
This function implements the ``shift-selection'' behavior of
point-motion commands.  @xref{Shift Selection,,, emacs, The GNU Emacs
Manual}.  It is called automatically by the Emacs command loop
whenever a command with a @samp{^} character in its @code{interactive}
spec is invoked, before the command itself is executed
(@pxref{Interactive Codes, ^}).

If @code{shift-select-mode} is non-@code{nil} and the current command
was invoked via shift translation (@pxref{Key Sequence Input,
shift-translation}), this function sets the mark and temporarily
activates the region, unless the region was already temporarily
activated in this way.  Otherwise, if the region has been activated
temporarily, it deactivates the mark and restores the variable
@code{transient-mark-mode} to its earlier value.
@end defun

@defvar mark-ring
The value of this buffer-local variable is the list of saved former
marks of the current buffer, most recent first.

@example
@group
mark-ring
@result{} (#<marker at 11050 in markers.texi>
    #<marker at 10832 in markers.texi>
    @dots{})
@end group
@end example
@end defvar

@defopt mark-ring-max
The value of this variable is the maximum size of @code{mark-ring}.  If
more marks than this are pushed onto the @code{mark-ring},
@code{push-mark} discards an old mark when it adds a new one.
@end defopt

@c There is also global-mark-ring-max, but this chapter explicitly
@c does not talk about the global mark.

@node The Region
@section The Region
@cindex region (between point and mark)

  The text between point and the mark is known as @dfn{the region}.
Various functions operate on text delimited by point and the mark, but
only those functions specifically related to the region itself are
described here.

The next two functions signal an error if the mark does not point
anywhere.  If Transient Mark mode is enabled and
@code{mark-even-if-inactive} is @code{nil}, they also signal an error
if the mark is inactive.

@defun region-beginning
This function returns the position of the beginning of the region (as
an integer).  This is the position of either point or the mark,
whichever is smaller.
@end defun

@defun region-end
This function returns the position of the end of the region (as an
integer).  This is the position of either point or the mark, whichever is
larger.
@end defun

  Instead of using @code{region-beginning} and @code{region-end}, a
command designed to operate on a region should normally use
@code{interactive} with the @samp{r} specification to find the
beginning and end of the region.  This lets other Lisp programs
specify the bounds explicitly as arguments.  @xref{Interactive Codes}.

@defun use-region-p
This function returns @code{t} if Transient Mark mode is enabled, the
mark is active, and there is a valid region in the buffer.  This
function is intended to be used by commands that operate on the
region, instead of on text near point, when the mark is active.

A region is valid if it has a non-zero size, or if the user option
@code{use-empty-active-region} is non-@code{nil} (by default, it is
@code{nil}).  The function @code{region-active-p} is similar to
@code{use-region-p}, but considers all regions as valid.  In most
cases, you should not use @code{region-active-p}, since if the region
is empty it is often more appropriate to operate on point.
@end defun