This is elisp, produced by makeinfo version 4.0f from ./elisp.texi.

INFO-DIR-SECTION Editors
START-INFO-DIR-ENTRY
* Elisp: (elisp).	The Emacs Lisp Reference Manual.
END-INFO-DIR-ENTRY

   This Info file contains edition 2.8 of the GNU Emacs Lisp Reference
Manual, corresponding to Emacs version 21.2.

   Published by the Free Software Foundation 59 Temple Place, Suite 330
Boston, MA  02111-1307  USA

   Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1998, 1999,
2000, 2001, 2002 Free Software Foundation, Inc.

   Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.1 or
any later version published by the Free Software Foundation; with the
Invariant Sections being "Copying", with the Front-Cover texts being "A
GNU Manual", and with the Back-Cover Texts as in (a) below.  A copy of
the license is included in the section entitled "GNU Free Documentation
License".

   (a) The FSF's Back-Cover Text is: "You have freedom to copy and
modify this GNU Manual, like GNU software.  Copies published by the Free
Software Foundation raise funds for GNU development."


File: elisp,  Node: Buffer File Name,  Next: Buffer Modification,  Prev: Buffer Names,  Up: Buffers

Buffer File Name
================

   The "buffer file name" is the name of the file that is visited in
that buffer.  When a buffer is not visiting a file, its buffer file name
is `nil'.  Most of the time, the buffer name is the same as the
nondirectory part of the buffer file name, but the buffer file name and
the buffer name are distinct and can be set independently.  *Note
Visiting Files::.

 - Function: buffer-file-name &optional buffer
     This function returns the absolute file name of the file that
     BUFFER is visiting.  If BUFFER is not visiting any file,
     `buffer-file-name' returns `nil'.  If BUFFER is not supplied, it
     defaults to the current buffer.

          (buffer-file-name (other-buffer))
               => "/usr/user/lewis/manual/files.texi"

 - Variable: buffer-file-name
     This buffer-local variable contains the name of the file being
     visited in the current buffer, or `nil' if it is not visiting a
     file.  It is a permanent local variable, unaffected by
     `kill-all-local-variables'.

          buffer-file-name
               => "/usr/user/lewis/manual/buffers.texi"

     It is risky to change this variable's value without doing various
     other things.  Normally it is better to use
     `set-visited-file-name' (see below); some of the things done
     there, such as changing the buffer name, are not strictly
     necessary, but others are essential to avoid confusing Emacs.

 - Variable: buffer-file-truename
     This buffer-local variable holds the truename of the file visited
     in the current buffer, or `nil' if no file is visited.  It is a
     permanent local, unaffected by `kill-all-local-variables'.  *Note
     Truenames::.

 - Variable: buffer-file-number
     This buffer-local variable holds the file number and directory
     device number of the file visited in the current buffer, or `nil'
     if no file or a nonexistent file is visited.  It is a permanent
     local, unaffected by `kill-all-local-variables'.

     The value is normally a list of the form `(FILENUM DEVNUM)'.  This
     pair of numbers uniquely identifies the file among all files
     accessible on the system.  See the function `file-attributes', in
     *Note File Attributes::, for more information about them.

 - Function: get-file-buffer filename
     This function returns the buffer visiting file FILENAME.  If there
     is no such buffer, it returns `nil'.  The argument FILENAME, which
     must be a string, is expanded (*note File Name Expansion::), then
     compared against the visited file names of all live buffers.

          (get-file-buffer "buffers.texi")
              => #<buffer buffers.texi>

     In unusual circumstances, there can be more than one buffer
     visiting the same file name.  In such cases, this function returns
     the first such buffer in the buffer list.

 - Command: set-visited-file-name filename &optional no-query
          along-with-file
     If FILENAME is a non-empty string, this function changes the name
     of the file visited in the current buffer to FILENAME.  (If the
     buffer had no visited file, this gives it one.)  The _next time_
     the buffer is saved it will go in the newly-specified file.  This
     command marks the buffer as modified, since it does not (as far as
     Emacs knows) match the contents of FILENAME, even if it matched the
     former visited file.

     If FILENAME is `nil' or the empty string, that stands for "no
     visited file".  In this case, `set-visited-file-name' marks the
     buffer as having no visited file.

     Normally, this function asks the user for confirmation if the
     specified file already exists.  If NO-QUERY is non-`nil', that
     prevents asking this question.

     If ALONG-WITH-FILE is non-`nil', that means to assume that the
     former visited file has been renamed to FILENAME.

     When the function `set-visited-file-name' is called interactively,
     it prompts for FILENAME in the minibuffer.

 - Variable: list-buffers-directory
     This buffer-local variable specifies a string to display in a
     buffer listing where the visited file name would go, for buffers
     that don't have a visited file name.  Dired buffers use this
     variable.


File: elisp,  Node: Buffer Modification,  Next: Modification Time,  Prev: Buffer File Name,  Up: Buffers

Buffer Modification
===================

   Emacs keeps a flag called the "modified flag" for each buffer, to
record whether you have changed the text of the buffer.  This flag is
set to `t' whenever you alter the contents of the buffer, and cleared
to `nil' when you save it.  Thus, the flag shows whether there are
unsaved changes.  The flag value is normally shown in the mode line
(*note Mode Line Variables::), and controls saving (*note Saving
Buffers::) and auto-saving (*note Auto-Saving::).

   Some Lisp programs set the flag explicitly.  For example, the
function `set-visited-file-name' sets the flag to `t', because the text
does not match the newly-visited file, even if it is unchanged from the
file formerly visited.

   The functions that modify the contents of buffers are described in
*Note Text::.

 - Function: buffer-modified-p &optional buffer
     This function returns `t' if the buffer BUFFER has been modified
     since it was last read in from a file or saved, or `nil'
     otherwise.  If BUFFER is not supplied, the current buffer is
     tested.

 - Function: set-buffer-modified-p flag
     This function marks the current buffer as modified if FLAG is
     non-`nil', or as unmodified if the flag is `nil'.

     Another effect of calling this function is to cause unconditional
     redisplay of the mode line for the current buffer.  In fact, the
     function `force-mode-line-update' works by doing this:

          (set-buffer-modified-p (buffer-modified-p))

 - Command: not-modified
     This command marks the current buffer as unmodified, and not
     needing to be saved.  With prefix arg, it marks the buffer as
     modified, so that it will be saved at the next suitable occasion.

     Don't use this function in programs, since it prints a message in
     the echo area; use `set-buffer-modified-p' (above) instead.

 - Function: buffer-modified-tick &optional buffer
     This function returns BUFFER's modification-count.  This is a
     counter that increments every time the buffer is modified.  If
     BUFFER is `nil' (or omitted), the current buffer is used.


File: elisp,  Node: Modification Time,  Next: Read Only Buffers,  Prev: Buffer Modification,  Up: Buffers

Comparison of Modification Time
===============================

   Suppose that you visit a file and make changes in its buffer, and
meanwhile the file itself is changed on disk.  At this point, saving the
buffer would overwrite the changes in the file.  Occasionally this may
be what you want, but usually it would lose valuable information.  Emacs
therefore checks the file's modification time using the functions
described below before saving the file.

 - Function: verify-visited-file-modtime buffer
     This function compares what BUFFER has recorded for the
     modification time of its visited file against the actual
     modification time of the file as recorded by the operating system.
     The two should be the same unless some other process has written
     the file since Emacs visited or saved it.

     The function returns `t' if the last actual modification time and
     Emacs's recorded modification time are the same, `nil' otherwise.

 - Function: clear-visited-file-modtime
     This function clears out the record of the last modification time
     of the file being visited by the current buffer.  As a result, the
     next attempt to save this buffer will not complain of a
     discrepancy in file modification times.

     This function is called in `set-visited-file-name' and other
     exceptional places where the usual test to avoid overwriting a
     changed file should not be done.

 - Function: visited-file-modtime
     This function returns the buffer's recorded last file modification
     time, as a list of the form `(HIGH . LOW)'.  (This is the same
     format that `file-attributes' uses to return time values; see
     *Note File Attributes::.)

 - Function: set-visited-file-modtime &optional time
     This function updates the buffer's record of the last modification
     time of the visited file, to the value specified by TIME if TIME
     is not `nil', and otherwise to the last modification time of the
     visited file.

     If TIME is not `nil', it should have the form `(HIGH . LOW)' or
     `(HIGH LOW)', in either case containing two integers, each of
     which holds 16 bits of the time.

     This function is useful if the buffer was not read from the file
     normally, or if the file itself has been changed for some known
     benign reason.

 - Function: ask-user-about-supersession-threat filename
     This function is used to ask a user how to proceed after an
     attempt to modify an obsolete buffer visiting file FILENAME.  An
     "obsolete buffer" is an unmodified buffer for which the associated
     file on disk is newer than the last save-time of the buffer.  This
     means some other program has probably altered the file.

     Depending on the user's answer, the function may return normally,
     in which case the modification of the buffer proceeds, or it may
     signal a `file-supersession' error with data `(FILENAME)', in which
     case the proposed buffer modification is not allowed.

     This function is called automatically by Emacs on the proper
     occasions.  It exists so you can customize Emacs by redefining it.
     See the file `userlock.el' for the standard definition.

     See also the file locking mechanism in *Note File Locks::.


File: elisp,  Node: Read Only Buffers,  Next: The Buffer List,  Prev: Modification Time,  Up: Buffers

Read-Only Buffers
=================

   If a buffer is "read-only", then you cannot change its contents,
although you may change your view of the contents by scrolling and
narrowing.

   Read-only buffers are used in two kinds of situations:

   * A buffer visiting a write-protected file is normally read-only.

     Here, the purpose is to inform the user that editing the buffer
     with the aim of saving it in the file may be futile or
     undesirable.  The user who wants to change the buffer text despite
     this can do so after clearing the read-only flag with `C-x C-q'.

   * Modes such as Dired and Rmail make buffers read-only when altering
     the contents with the usual editing commands would probably be a
     mistake.

     The special commands of these modes bind `buffer-read-only' to
     `nil' (with `let') or bind `inhibit-read-only' to `t' around the
     places where they themselves change the text.

 - Variable: buffer-read-only
     This buffer-local variable specifies whether the buffer is
     read-only.  The buffer is read-only if this variable is non-`nil'.

 - Variable: inhibit-read-only
     If this variable is non-`nil', then read-only buffers and read-only
     characters may be modified.  Read-only characters in a buffer are
     those that have non-`nil' `read-only' properties (either text
     properties or overlay properties).  *Note Special Properties::,
     for more information about text properties.  *Note Overlays::, for
     more information about overlays and their properties.

     If `inhibit-read-only' is `t', all `read-only' character
     properties have no effect.  If `inhibit-read-only' is a list, then
     `read-only' character properties have no effect if they are members
     of the list (comparison is done with `eq').

 - Command: toggle-read-only
     This command changes whether the current buffer is read-only.  It
     is intended for interactive use; do not use it in programs.  At
     any given point in a program, you should know whether you want the
     read-only flag on or off; so you can set `buffer-read-only'
     explicitly to the proper value, `t' or `nil'.

 - Function: barf-if-buffer-read-only
     This function signals a `buffer-read-only' error if the current
     buffer is read-only.  *Note Interactive Call::, for another way to
     signal an error if the current buffer is read-only.


File: elisp,  Node: The Buffer List,  Next: Creating Buffers,  Prev: Read Only Buffers,  Up: Buffers

The Buffer List
===============

   The "buffer list" is a list of all live buffers.  Creating a buffer
adds it to this list, and killing a buffer excises it.  The order of
the buffers in the list is based primarily on how recently each buffer
has been displayed in the selected window.  Buffers move to the front
of the list when they are selected and to the end when they are buried
(see `bury-buffer', below).  Several functions, notably `other-buffer',
use this ordering.  A buffer list displayed for the user also follows
this order.

   In addition to the fundamental Emacs buffer list, each frame has its
own version of the buffer list, in which the buffers that have been
selected in that frame come first, starting with the buffers most
recently selected _in that frame_.  (This order is recorded in FRAME's
`buffer-list' frame parameter; see *Note Window Frame Parameters::.)
The buffers that were never selected in FRAME come afterward, ordered
according to the fundamental Emacs buffer list.

 - Function: buffer-list &optional frame
     This function returns the buffer list, including all buffers, even
     those whose names begin with a space.  The elements are actual
     buffers, not their names.

     If FRAME is a frame, this returns FRAME's buffer list.  If FRAME
     is `nil', the fundamental Emacs buffer list is used: all the
     buffers appear in order of most recent selection, regardless of
     which frames they were selected in.

          (buffer-list)
               => (#<buffer buffers.texi>
                   #<buffer  *Minibuf-1*> #<buffer buffer.c>
                   #<buffer *Help*> #<buffer TAGS>)
          
          ;; Note that the name of the minibuffer
          ;;   begins with a space!
          (mapcar (function buffer-name) (buffer-list))
              => ("buffers.texi" " *Minibuf-1*"
                  "buffer.c" "*Help*" "TAGS")

   The list that `buffer-list' returns is constructed specifically by
`buffer-list'; it is not an internal Emacs data structure, and
modifying it has no effect on the order of buffers.  If you want to
change the order of buffers in the frame-independent buffer list, here
is an easy way:

     (defun reorder-buffer-list (new-list)
       (while new-list
         (bury-buffer (car new-list))
         (setq new-list (cdr new-list))))

   With this method, you can specify any order for the list, but there
is no danger of losing a buffer or adding something that is not a valid
live buffer.

   To change the order or value of a frame's buffer list, set the
frame's `buffer-list' frame parameter with `modify-frame-parameters'
(*note Parameter Access::).

 - Function: other-buffer &optional buffer visible-ok frame
     This function returns the first buffer in the buffer list other
     than BUFFER.  Usually this is the buffer selected most recently (in
     frame FRAME or else the currently selected frame, *note Input
     Focus::), aside from BUFFER.  Buffers whose names start with a
     space are not considered at all.

     If BUFFER is not supplied (or if it is not a buffer), then
     `other-buffer' returns the first buffer in the selected frame's
     buffer list that is not now visible in any window in a visible
     frame.

     If FRAME has a non-`nil' `buffer-predicate' parameter, then
     `other-buffer' uses that predicate to decide which buffers to
     consider.  It calls the predicate once for each buffer, and if the
     value is `nil', that buffer is ignored.  *Note Window Frame
     Parameters::.

     If VISIBLE-OK is `nil', `other-buffer' avoids returning a buffer
     visible in any window on any visible frame, except as a last
     resort.   If VISIBLE-OK is non-`nil', then it does not matter
     whether a buffer is displayed somewhere or not.

     If no suitable buffer exists, the buffer `*scratch*' is returned
     (and created, if necessary).

 - Command: bury-buffer &optional buffer-or-name
     This function puts BUFFER-OR-NAME at the end of the buffer list,
     without changing the order of any of the other buffers on the list.
     This buffer therefore becomes the least desirable candidate for
     `other-buffer' to return.

     `bury-buffer' operates on each frame's `buffer-list' parameter as
     well as the frame-independent Emacs buffer list; therefore, the
     buffer that you bury will come last in the value of `(buffer-list
     FRAME)' and in the value of `(buffer-list nil)'.

     If BUFFER-OR-NAME is `nil' or omitted, this means to bury the
     current buffer.  In addition, if the buffer is displayed in the
     selected window, this switches to some other buffer (obtained using
     `other-buffer') in the selected window.  But if the buffer is
     displayed in some other window, it remains displayed there.

     To replace a buffer in all the windows that display it, use
     `replace-buffer-in-windows'.  *Note Buffers and Windows::.


File: elisp,  Node: Creating Buffers,  Next: Killing Buffers,  Prev: The Buffer List,  Up: Buffers

Creating Buffers
================

   This section describes the two primitives for creating buffers.
`get-buffer-create' creates a buffer if it finds no existing buffer
with the specified name; `generate-new-buffer' always creates a new
buffer and gives it a unique name.

   Other functions you can use to create buffers include
`with-output-to-temp-buffer' (*note Temporary Displays::) and
`create-file-buffer' (*note Visiting Files::).  Starting a subprocess
can also create a buffer (*note Processes::).

 - Function: get-buffer-create name
     This function returns a buffer named NAME.  It returns an existing
     buffer with that name, if one exists; otherwise, it creates a new
     buffer.  The buffer does not become the current buffer--this
     function does not change which buffer is current.

     An error is signaled if NAME is not a string.

          (get-buffer-create "foo")
               => #<buffer foo>

     The major mode for the new buffer is set to Fundamental mode.  The
     variable `default-major-mode' is handled at a higher level.  *Note
     Auto Major Mode::.

 - Function: generate-new-buffer name
     This function returns a newly created, empty buffer, but does not
     make it current.  If there is no buffer named NAME, then that is
     the name of the new buffer.  If that name is in use, this function
     adds suffixes of the form `<N>' to NAME, where N is an integer.
     It tries successive integers starting with 2 until it finds an
     available name.

     An error is signaled if NAME is not a string.

          (generate-new-buffer "bar")
               => #<buffer bar>
          (generate-new-buffer "bar")
               => #<buffer bar<2>>
          (generate-new-buffer "bar")
               => #<buffer bar<3>>

     The major mode for the new buffer is set to Fundamental mode.  The
     variable `default-major-mode' is handled at a higher level.  *Note
     Auto Major Mode::.

     See the related function `generate-new-buffer-name' in *Note
     Buffer Names::.


File: elisp,  Node: Killing Buffers,  Next: Indirect Buffers,  Prev: Creating Buffers,  Up: Buffers

Killing Buffers
===============

   "Killing a buffer" makes its name unknown to Emacs and makes its
text space available for other use.

   The buffer object for the buffer that has been killed remains in
existence as long as anything refers to it, but it is specially marked
so that you cannot make it current or display it.  Killed buffers retain
their identity, however; if you kill two distinct buffers, they remain
distinct according to `eq' although both are dead.

   If you kill a buffer that is current or displayed in a window, Emacs
automatically selects or displays some other buffer instead.  This means
that killing a buffer can in general change the current buffer.
Therefore, when you kill a buffer, you should also take the precautions
associated with changing the current buffer (unless you happen to know
that the buffer being killed isn't current).  *Note Current Buffer::.

   If you kill a buffer that is the base buffer of one or more indirect
buffers, the indirect buffers are automatically killed as well.

   The `buffer-name' of a killed buffer is `nil'.  You can use this
feature to test whether a buffer has been killed:

     (defun buffer-killed-p (buffer)
       "Return t if BUFFER is killed."
       (not (buffer-name buffer)))

 - Command: kill-buffer buffer-or-name
     This function kills the buffer BUFFER-OR-NAME, freeing all its
     memory for other uses or to be returned to the operating system.
     It returns `nil'.

     Any processes that have this buffer as the `process-buffer' are
     sent the `SIGHUP' signal, which normally causes them to terminate.
     (The basic meaning of `SIGHUP' is that a dialup line has been
     disconnected.)  *Note Deleting Processes::.

     If the buffer is visiting a file and contains unsaved changes,
     `kill-buffer' asks the user to confirm before the buffer is killed.
     It does this even if not called interactively.  To prevent the
     request for confirmation, clear the modified flag before calling
     `kill-buffer'.  *Note Buffer Modification::.

     Killing a buffer that is already dead has no effect.

          (kill-buffer "foo.unchanged")
               => nil
          (kill-buffer "foo.changed")
          
          ---------- Buffer: Minibuffer ----------
          Buffer foo.changed modified; kill anyway? (yes or no) yes
          ---------- Buffer: Minibuffer ----------
          
               => nil

 - Variable: kill-buffer-query-functions
     After confirming unsaved changes, `kill-buffer' calls the functions
     in the list `kill-buffer-query-functions', in order of appearance,
     with no arguments.  The buffer being killed is the current buffer
     when they are called.  The idea of this feature is that these
     functions will ask for confirmation from the user.  If any of them
     returns `nil', `kill-buffer' spares the buffer's life.

 - Variable: kill-buffer-hook
     This is a normal hook run by `kill-buffer' after asking all the
     questions it is going to ask, just before actually killing the
     buffer.  The buffer to be killed is current when the hook
     functions run.  *Note Hooks::.

 - Variable: buffer-offer-save
     This variable, if non-`nil' in a particular buffer, tells
     `save-buffers-kill-emacs' and `save-some-buffers' to offer to save
     that buffer, just as they offer to save file-visiting buffers.  The
     variable `buffer-offer-save' automatically becomes buffer-local
     when set for any reason.  *Note Buffer-Local Variables::.


File: elisp,  Node: Indirect Buffers,  Next: Buffer Gap,  Prev: Killing Buffers,  Up: Buffers

Indirect Buffers
================

   An "indirect buffer" shares the text of some other buffer, which is
called the "base buffer" of the indirect buffer.  In some ways it is
the analogue, for buffers, of a symbolic link among files.  The base
buffer may not itself be an indirect buffer.

   The text of the indirect buffer is always identical to the text of
its base buffer; changes made by editing either one are visible
immediately in the other.  This includes the text properties as well as
the characters themselves.

   In all other respects, the indirect buffer and its base buffer are
completely separate.  They have different names, different values of
point, different narrowing, different markers and overlays (though
inserting or deleting text in either buffer relocates the markers and
overlays for both), different major modes, and different buffer-local
variables.

   An indirect buffer cannot visit a file, but its base buffer can.  If
you try to save the indirect buffer, that actually saves the base
buffer.

   Killing an indirect buffer has no effect on its base buffer.  Killing
the base buffer effectively kills the indirect buffer in that it cannot
ever again be the current buffer.

 - Command: make-indirect-buffer base-buffer name
     This creates an indirect buffer named NAME whose base buffer is
     BASE-BUFFER.  The argument BASE-BUFFER may be a buffer or a string.

     If BASE-BUFFER is an indirect buffer, its base buffer is used as
     the base for the new buffer.

 - Function: buffer-base-buffer buffer
     This function returns the base buffer of BUFFER.  If BUFFER is not
     indirect, the value is `nil'.  Otherwise, the value is another
     buffer, which is never an indirect buffer.


File: elisp,  Node: Buffer Gap,  Prev: Indirect Buffers,  Up: Buffers

The Buffer Gap
==============

   Emacs buffers are implemented using an invisible "gap" to make
insertion and deletion faster.  Insertion works by filling in part of
the gap, and deletion adds to the gap.  Of course, this means that the
gap must first be moved to the locus of the insertion or deletion.
Emacs moves the gap only when you try to insert or delete.  This is why
your first editing command in one part of a large buffer, after
previously editing in another far-away part, sometimes involves a
noticeable delay.

   This mechanism works invisibly, and Lisp code should never be
affected by the gap's current location, but these functions are
available for getting information about the gap status.

 - Function: gap-position
     This function returns the current gap position in the current
     buffer.

 - Function: gap-size
     This function returns the current gap size of the current buffer.


File: elisp,  Node: Windows,  Next: Frames,  Prev: Buffers,  Up: Top

Windows
*******

   This chapter describes most of the functions and variables related to
Emacs windows.  See *Note Display::, for information on how text is
displayed in windows.

* Menu:

* Basic Windows::           Basic information on using windows.
* Splitting Windows::       Splitting one window into two windows.
* Deleting Windows::        Deleting a window gives its space to other windows.
* Selecting Windows::       The selected window is the one that you edit in.
* Cyclic Window Ordering::  Moving around the existing windows.
* Buffers and Windows::     Each window displays the contents of a buffer.
* Displaying Buffers::      Higher-lever functions for displaying a buffer
                              and choosing a window for it.
* Choosing Window::	    How to choose a window for displaying a buffer.
* Window Point::            Each window has its own location of point.
* Window Start::            The display-start position controls which text
                              is on-screen in the window.
* Textual Scrolling::       Moving text up and down through the window.
* Vertical Scrolling::      Moving the contents up and down on the window.
* Horizontal Scrolling::    Moving the contents sideways on the window.
* Size of Window::          Accessing the size of a window.
* Resizing Windows::        Changing the size of a window.
* Coordinates and Windows:: Converting coordinates to windows.
* Window Configurations::   Saving and restoring the state of the screen.
* Window Hooks::            Hooks for scrolling, window size changes,
                              redisplay going past a certain point,
                              or window configuration changes.


File: elisp,  Node: Basic Windows,  Next: Splitting Windows,  Up: Windows

Basic Concepts of Emacs Windows
===============================

   A "window" in Emacs is the physical area of the screen in which a
buffer is displayed.  The term is also used to refer to a Lisp object
that represents that screen area in Emacs Lisp.  It should be clear
from the context which is meant.

   Emacs groups windows into frames.  A frame represents an area of
screen available for Emacs to use.  Each frame always contains at least
one window, but you can subdivide it vertically or horizontally into
multiple nonoverlapping Emacs windows.

   In each frame, at any time, one and only one window is designated as
"selected within the frame".  The frame's cursor appears in that
window.  At any time, one frame is the selected frame; and the window
selected within that frame is "the selected window".  The selected
window's buffer is usually the current buffer (except when `set-buffer'
has been used).  *Note Current Buffer::.

   For practical purposes, a window exists only while it is displayed in
a frame.  Once removed from the frame, the window is effectively deleted
and should not be used, _even though there may still be references to
it_ from other Lisp objects.  Restoring a saved window configuration is
the only way for a window no longer on the screen to come back to life.
(*Note Deleting Windows::.)

   Each window has the following attributes:

   * containing frame

   * window height

   * window width

   * window edges with respect to the screen or frame

   * the buffer it displays

   * position within the buffer at the upper left of the window

   * amount of horizontal scrolling, in columns

   * point

   * the mark

   * how recently the window was selected

   Users create multiple windows so they can look at several buffers at
once.  Lisp libraries use multiple windows for a variety of reasons, but
most often to display related information.  In Rmail, for example, you
can move through a summary buffer in one window while the other window
shows messages one at a time as they are reached.

   The meaning of "window" in Emacs is similar to what it means in the
context of general-purpose window systems such as X, but not identical.
The X Window System places X windows on the screen; Emacs uses one or
more X windows as frames, and subdivides them into Emacs windows.  When
you use Emacs on a character-only terminal, Emacs treats the whole
terminal screen as one frame.

   Most window systems support arbitrarily located overlapping windows.
In contrast, Emacs windows are "tiled"; they never overlap, and
together they fill the whole screen or frame.  Because of the way in
which Emacs creates new windows and resizes them, not all conceivable
tilings of windows on an Emacs frame are actually possible.  *Note
Splitting Windows::, and *Note Size of Window::.

   *Note Display::, for information on how the contents of the window's
buffer are displayed in the window.

 - Function: windowp object
     This function returns `t' if OBJECT is a window.


File: elisp,  Node: Splitting Windows,  Next: Deleting Windows,  Prev: Basic Windows,  Up: Windows

Splitting Windows
=================

   The functions described here are the primitives used to split a
window into two windows.  Two higher level functions sometimes split a
window, but not always: `pop-to-buffer' and `display-buffer' (*note
Displaying Buffers::).

   The functions described here do not accept a buffer as an argument.
The two "halves" of the split window initially display the same buffer
previously visible in the window that was split.

 - Command: split-window &optional window size horizontal
     This function splits WINDOW into two windows.  The original window
     WINDOW remains the selected window, but occupies only part of its
     former screen area.  The rest is occupied by a newly created
     window which is returned as the value of this function.

     If HORIZONTAL is non-`nil', then WINDOW splits into two side by
     side windows.  The original window WINDOW keeps the leftmost SIZE
     columns, and gives the rest of the columns to the new window.
     Otherwise, it splits into windows one above the other, and WINDOW
     keeps the upper SIZE lines and gives the rest of the lines to the
     new window.  The original window is therefore the left-hand or
     upper of the two, and the new window is the right-hand or lower.

     If WINDOW is omitted or `nil', then the selected window is split.
     If SIZE is omitted or `nil', then WINDOW is divided evenly into
     two parts.  (If there is an odd line, it is allocated to the new
     window.)  When `split-window' is called interactively, all its
     arguments are `nil'.

     The following example starts with one window on a screen that is 50
     lines high by 80 columns wide; then the window is split.

          (setq w (selected-window))
               => #<window 8 on windows.texi>
          (window-edges)          ; Edges in order:
               => (0 0 80 50)     ;   left-top-right-bottom
          
          ;; Returns window created
          (setq w2 (split-window w 15))
               => #<window 28 on windows.texi>
          (window-edges w2)
               => (0 15 80 50)    ; Bottom window;
                                  ;   top is line 15
          (window-edges w)
               => (0 0 80 15)     ; Top window

     The screen looks like this:

                   __________
                  |          |  line 0
                  |    w     |
                  |__________|
                  |          |  line 15
                  |    w2    |
                  |__________|
                                line 50
           column 0   column 80

     Next, the top window is split horizontally:

          (setq w3 (split-window w 35 t))
               => #<window 32 on windows.texi>
          (window-edges w3)
               => (35 0 80 15)  ; Left edge at column 35
          (window-edges w)
               => (0 0 35 15)   ; Right edge at column 35
          (window-edges w2)
               => (0 15 80 50)  ; Bottom window unchanged

     Now, the screen looks like this:

               column 35
                   __________
                  |   |      |  line 0
                  | w |  w3  |
                  |___|______|
                  |          |  line 15
                  |    w2    |
                  |__________|
                                line 50
           column 0   column 80

     Normally, Emacs indicates the border between two side-by-side
     windows with a scroll bar (*note Scroll Bars: Window Frame
     Parameters.) or `|' characters.  The display table can specify
     alternative border characters; see *Note Display Tables::.

 - Command: split-window-vertically &optional size
     This function splits the selected window into two windows, one
     above the other, leaving the upper of the two windows selected,
     with SIZE lines.  (If SIZE is negative, then the lower of the two
     windows gets - SIZE lines and the upper window gets the rest, but
     the upper window is still the one selected.)

 - Command: split-window-horizontally &optional size
     This function splits the selected window into two windows
     side-by-side, leaving the selected window with SIZE columns.

     This function is basically an interface to `split-window'.  You
     could define a simplified version of the function like this:

          (defun split-window-horizontally (&optional arg)
            "Split selected window into two windows, side by side..."
            (interactive "P")
            (let ((size (and arg (prefix-numeric-value arg))))
              (and size (< size 0)
          	 (setq size (+ (window-width) size)))
              (split-window nil size t)))

 - Function: one-window-p &optional no-mini all-frames
     This function returns non-`nil' if there is only one window.  The
     argument NO-MINI, if non-`nil', means don't count the minibuffer
     even if it is active; otherwise, the minibuffer window is
     included, if active, in the total number of windows, which is
     compared against one.

     The argument ALL-FRAMES specifies which frames to consider.  Here
     are the possible values and their meanings:

    `nil'
          Count the windows in the selected frame, plus the minibuffer
          used by that frame even if it lies in some other frame.

    `t'
          Count all windows in all existing frames.

    `visible'
          Count all windows in all visible frames.

    0
          Count all windows in all visible or iconified frames.

    anything else
          Count precisely the windows in the selected frame, and no
          others.


File: elisp,  Node: Deleting Windows,  Next: Selecting Windows,  Prev: Splitting Windows,  Up: Windows

Deleting Windows
================

   A window remains visible on its frame unless you "delete" it by
calling certain functions that delete windows.  A deleted window cannot
appear on the screen, but continues to exist as a Lisp object until
there are no references to it.  There is no way to cancel the deletion
of a window aside from restoring a saved window configuration (*note
Window Configurations::).  Restoring a window configuration also
deletes any windows that aren't part of that configuration.

   When you delete a window, the space it took up is given to one
adjacent sibling.

 - Function: window-live-p window
     This function returns `nil' if WINDOW is deleted, and `t'
     otherwise.

     *Warning:* Erroneous information or fatal errors may result from
     using a deleted window as if it were live.

 - Command: delete-window &optional window
     This function removes WINDOW from display, and returns `nil'.  If
     WINDOW is omitted, then the selected window is deleted.  An error
     is signaled if there is only one window when `delete-window' is
     called.

 - Command: delete-other-windows &optional window
     This function makes WINDOW the only window on its frame, by
     deleting the other windows in that frame.  If WINDOW is omitted or
     `nil', then the selected window is used by default.

     The return value is `nil'.

 - Command: delete-windows-on buffer &optional frame
     This function deletes all windows showing BUFFER.  If there are no
     windows showing BUFFER, it does nothing.

     `delete-windows-on' operates frame by frame.  If a frame has
     several windows showing different buffers, then those showing
     BUFFER are removed, and the others expand to fill the space.  If
     all windows in some frame are showing BUFFER (including the case
     where there is only one window), then the frame reverts to having a
     single window showing another buffer chosen with `other-buffer'.
     *Note The Buffer List::.

     The argument FRAME controls which frames to operate on.  This
     function does not use it in quite the same way as the other
     functions which scan all windows; specifically, the values `t' and
     `nil' have the opposite of their meanings in other functions.
     Here are the full details:

        * If it is `nil', operate on all frames.

        * If it is `t', operate on the selected frame.

        * If it is `visible', operate on all visible frames.

        * If it is 0, operate on all visible or iconified frames.

        * If it is a frame, operate on that frame.

     This function always returns `nil'.


File: elisp,  Node: Selecting Windows,  Next: Cyclic Window Ordering,  Prev: Deleting Windows,  Up: Windows

Selecting Windows
=================

   When a window is selected, the buffer in the window becomes the
current buffer, and the cursor will appear in it.

 - Function: selected-window
     This function returns the selected window.  This is the window in
     which the cursor appears and to which many commands apply.

 - Function: select-window window
     This function makes WINDOW the selected window.  The cursor then
     appears in WINDOW (on redisplay).  The buffer being displayed in
     WINDOW is immediately designated the current buffer.

     The return value is WINDOW.

          (setq w (next-window))
          (select-window w)
               => #<window 65 on windows.texi>

 - Macro: save-selected-window forms...
     This macro records the selected window, executes FORMS in
     sequence, then restores the earlier selected window, unless it is
     no longer alive.

     This macro does not save or restore anything about the sizes,
     arrangement or contents of windows; therefore, if the FORMS change
     them, the change persists.

     Each frame, at any time, has a window selected within the frame.
     This macro saves only _the_ selected window; it does not save
     anything about other frames.  If the FORMS select some other frame
     and alter the window selected within it, the change persists.

   The following functions choose one of the windows on the screen,
offering various criteria for the choice.

 - Function: get-lru-window &optional frame
     This function returns the window least recently "used" (that is,
     selected).  The selected window is always the most recently used
     window.

     The selected window can be the least recently used window if it is
     the only window.  A newly created window becomes the least
     recently used window until it is selected.  A minibuffer window is
     never a candidate.

     The argument FRAME controls which windows are considered.

        * If it is `nil', consider windows on the selected frame.

        * If it is `t', consider windows on all frames.

        * If it is `visible', consider windows on all visible frames.

        * If it is 0, consider windows on all visible or iconified
          frames.

        * If it is a frame, consider windows on that frame.

 - Function: get-largest-window &optional frame
     This function returns the window with the largest area (height
     times width).  If there are no side-by-side windows, then this is
     the window with the most lines.  A minibuffer window is never a
     candidate.

     If there are two windows of the same size, then the function
     returns the window that is first in the cyclic ordering of windows
     (see following section), starting from the selected window.

     The argument FRAME controls which set of windows to consider.  See
     `get-lru-window', above.

 - Function: get-window-with-predicate predicate &optional minibuf
          all-frames default
     This function returns a window satisfying PREDICATE.  It cycles
     through all visible windows using `walk-windows' (*note Cyclic
     Window Ordering::), calling PREDICATE on each one one of them with
     that window as its argument.  The function returns the first
     window for which PREDICATE returns a non-`nil' value; if that
     never happens, it returns DEFAULT.

     The optional arguments MINIBUF and ALL-FRAMES specify the set of
     windows to include in the scan.  See the description of
     `next-window' in *Note Cyclic Window Ordering::, for details.


File: elisp,  Node: Cyclic Window Ordering,  Next: Buffers and Windows,  Prev: Selecting Windows,  Up: Windows

Cyclic Ordering of Windows
==========================

   When you use the command `C-x o' (`other-window') to select the next
window, it moves through all the windows on the screen in a specific
cyclic order.  For any given configuration of windows, this order never
varies.  It is called the "cyclic ordering of windows".

   This ordering generally goes from top to bottom, and from left to
right.  But it may go down first or go right first, depending on the
order in which the windows were split.

   If the first split was vertical (into windows one above each other),
and then the subwindows were split horizontally, then the ordering is
left to right in the top of the frame, and then left to right in the
next lower part of the frame, and so on.  If the first split was
horizontal, the ordering is top to bottom in the left part, and so on.
In general, within each set of siblings at any level in the window tree,
the order is left to right, or top to bottom.

 - Function: next-window &optional window minibuf all-frames
     This function returns the window following WINDOW in the cyclic
     ordering of windows.  This is the window that `C-x o' would select
     if typed when WINDOW is selected.  If WINDOW is the only window
     visible, then this function returns WINDOW.  If omitted, WINDOW
     defaults to the selected window.

     The value of the argument MINIBUF determines whether the
     minibuffer is included in the window order.  Normally, when
     MINIBUF is `nil', the minibuffer is included if it is currently
     active; this is the behavior of `C-x o'.  (The minibuffer window
     is active while the minibuffer is in use.  *Note Minibuffers::.)

     If MINIBUF is `t', then the cyclic ordering includes the
     minibuffer window even if it is not active.

     If MINIBUF is neither `t' nor `nil', then the minibuffer window is
     not included even if it is active.

     The argument ALL-FRAMES specifies which frames to consider.  Here
     are the possible values and their meanings:

    `nil'
          Consider all the windows in WINDOW's frame, plus the
          minibuffer used by that frame even if it lies in some other
          frame.

    `t'
          Consider all windows in all existing frames.

    `visible'
          Consider all windows in all visible frames.  (To get useful
          results, you must ensure WINDOW is in a visible frame.)

    0
          Consider all windows in all visible or iconified frames.

    anything else
          Consider precisely the windows in WINDOW's frame, and no
          others.

     This example assumes there are two windows, both displaying the
     buffer `windows.texi':

          (selected-window)
               => #<window 56 on windows.texi>
          (next-window (selected-window))
               => #<window 52 on windows.texi>
          (next-window (next-window (selected-window)))
               => #<window 56 on windows.texi>

 - Function: previous-window &optional window minibuf all-frames
     This function returns the window preceding WINDOW in the cyclic
     ordering of windows.  The other arguments specify which windows to
     include in the cycle, as in `next-window'.

 - Command: other-window count &optional all-frames
     This function selects the COUNTth following window in the cyclic
     order.  If count is negative, then it moves back -COUNT windows in
     the cycle, rather than forward.  It returns `nil'.

     The argument ALL-FRAMES has the same meaning as in `next-window',
     but the MINIBUF argument of `next-window' is always effectively
     `nil'.

     In an interactive call, COUNT is the numeric prefix argument.

 - Function: walk-windows proc &optional minibuf all-frames
     This function cycles through all windows, calling `proc' once for
     each window with the window as its sole argument.

     The optional arguments MINIBUF and ALL-FRAMES specify the set of
     windows to include in the scan.  See `next-window', above, for
     details.

 - Function: window-list &optional frame minibuf window
     This function returns a list of the windows on FRAME, starting
     with WINDOW.  If FRAME is `nil' or omitted, the selected frame is
     used instead; if WINDOW is `nil' or omitted, the selected window
     is used instead.

     The value of MINIBUF determines if the minibuffer window will be
     included in the result list.  If MINIBUF is `t', the minibuffer
     window will be included, even if it isn't active.  If MINIBUF is
     `nil' or omitted, the minibuffer window will only be included in
     the list if it is active.  If MINIBUF is neither `nil' nor `t',
     the minibuffer window is not included, whether or not it is active.