This file documents Ediff, a comprehensive visual interface to Unix diff and patch utilities.
Copyright © 1995–2023 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 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, 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 the freedom to copy and modify this GNU manual.”
Table of Contents
- 1 Introduction
- 2 Major Entry Points
- 3 Session Commands
- 4 Registry of Ediff Sessions
- 5 Session Groups
- 6 Remote and Compressed Files
- 7 Customization
- 7.1 Hooks
- 7.2 Quick Help Customization
- 7.3 Window and Frame Configuration
- 7.4 Selective Browsing
- 7.5 Highlighting Difference Regions
- 7.6 Narrowing
- 7.7 Refinement of Difference Regions
- 7.8 Patch and Diff Programs
- 7.9 Merging and diff3
- 7.10 Support for Version Control
- 7.11 Customizing the Mode Line
- 7.12 Miscellaneous
- 7.13 Notes on Heavy-duty Customization
- 8 Credits
- Appendix A GNU Free Documentation License
Ediff provides a convenient way for simultaneous browsing through the differences between a pair (or a triple) of files or buffers (which are called ‘variants’ for our purposes). The files being compared, file-A, file-B, and file-C (if applicable) are shown in separate windows (side by side, one above the another, or in separate frames), and the differences are highlighted as you step through them. You can also copy difference regions from one buffer to another (and recover old differences if you change your mind).
Another powerful feature is the ability to merge a pair of files into a third buffer. Merging with an ancestor file, (a.k.a. 3way merges) is also supported. Furthermore, Ediff is equipped with directory-level capabilities that allow the user to conveniently launch browsing or merging sessions on groups of files in two (or three) different directories.
In addition, Ediff can apply a patch to a file and then let you step through both files, the patched and the original one, simultaneously, difference-by-difference. You can even apply a patch right out of a mail buffer, i.e., patches received by mail don’t even have to be saved. Since Ediff lets you copy differences between variants, you can, in effect, apply patches selectively (i.e., you can copy a difference region from file.orig to file, thereby undoing any particular patch that you don’t like).
Ediff even understands multi-file patches and can apply them interactively!
(Ediff can recognize multi-file patches only if they are in the context
format or GNU unified format. All other patches are treated as 1-file
patches. Ediff is [hopefully] using the same algorithm as
determine which files need to be patched.)
Ediff is aware of version control, which lets you compare files with their older versions. Ediff also works with remote and compressed files, automatically ftp’ing them over and uncompressing them. See Remote and Compressed Files, for details.
This package builds upon ideas borrowed from Emerge, and several of Ediff’s functions are adaptations from Emerge. Although Ediff subsumes and greatly extends Emerge, much of the functionality in Ediff is influenced by Emerge. The architecture and the interface are, of course, drastically different.
2 Major Entry Points
When Ediff starts up, it displays a small control window, which accepts the Ediff commands, and two or three windows displaying the files to be compared or merged. The control window can be in its own small frame or it can be part of a bigger frame that displays other buffers. In any case, it is important that the control window be active (i.e., be the one receiving the keystrokes) when you use Ediff. You can switch to other Emacs buffers at will and even edit the files currently being compared with Ediff and then switch back to Ediff at any time by activating the appropriate Emacs windows.
Ediff can be invoked interactively using the following functions, which can be run either from the minibuffer or from the menu bar. In the menu bar, all Ediff’s entry points belong to three submenus of the Tools menu: Compare, Merge, and Apply Patch.
Compare two files.
Compare a file with its backup. If there are several numerical backups, use the latest. If the file is itself a backup, then compare it with its original.
Compare the buffer with its file on disk. This function can be used as a safe version of
Compare two buffers.
Compare three files.
Compare three buffers.
Compare files common to two directories.
Compare files common to three directories.
Compare versions of files in a given directory. Ediff selects only the files that are under version control.
Merge versions of files in a given directory. Ediff selects only the files that are under version control.
Merge versions of files in a given directory using other versions as ancestors. Ediff selects only the files that are under version control.
Compare text visible in 2 windows word-by-word.
Compare text visible in 2 windows line-by-line.
Compare regions word-by-word. The regions can come from the same buffer and they can even overlap. You will be asked to specify the buffers that contain the regions, which you want to compare. For each buffer, you will also be asked to mark the regions to be compared. Pay attention to the messages that appear in the minibuffer.
ediff-windows-linewise, but compares the regions line-by-line. See
ediff-windows-linewisefor more details.
Compare versions of the current buffer, if the buffer is visiting a file under version control.
Patch a file or multiple files, then compare. If the patch applies to just one file, Ediff will invoke a regular comparison session. If it is a multi-file patch, then a session group interface will be used and the user will be able to patch the files selectively. See Session Groups, for more details.
Since the patch might be in a buffer or a file, you will be asked which is the case. To avoid this extra prompt, you can invoke this command with a prefix argument. With an odd prefix argument, Ediff assumes the patch is in a file; with an even argument, a buffer is assumed.
ediff-patch-filewill actually use the
patchutility to change the original files on disk. This is not that dangerous, since you will always have the original contents of the file saved in another file that has the extension .orig. Furthermore, if the file is under version control, then you can always back out to one of the previous versions (see the section on Version Control in the Emacs manual).
ediff-patch-fileis careful about versions control: if the file to be patched is checked in, then Ediff will offer to check it out, because failing to do so may result in the loss of the changes when the file is checked out the next time.
If you don’t intend to modify the file via the patch and just want to see what the patch is all about (and decide later), then
ediff-patch-buffermight be a better choice.
Patch a buffer, then compare. The buffer being patched and the file visited by that buffer (if any) is not modified. The result of the patch appears in some other buffer that has the name ending with _patched.
This function would refuse to apply a multifile patch to a buffer. Use
ediff-patch-filefor that (and when you want the original file to be modified by the
Since the patch might be in a buffer or a file, you will be asked which is the case. To avoid this extra prompt, you can invoke this command with a prefix argument. With an odd prefix argument, Ediff assumes the patch is in a file; with an even argument, a buffer is assumed.
Merge two files.
ediff-merge, but with a third ancestor file.
Merge two buffers.
Same but with ancestor.
Merge files common to two directories.
Same but using files in a third directory as ancestors. If a pair of files doesn’t have an ancestor in the ancestor-directory, you will still be able to merge them without the ancestor.
Merge two versions of the file visited by the current buffer.
Same but with ancestor.
Brings up this manual.
Brings up Ediff session registry. This feature enables you to quickly find and restart active Ediff sessions.
When the above functions are invoked, the user is prompted for all the
necessary information—typically the files or buffers to compare, merge, or
patch. Ediff tries to be smart about these prompts. For instance, in
comparing/merging files, it will offer the visible buffers as defaults. In
prompting for files, if the user enters a directory, the previously input
file name will be appended to that directory. In addition, if the variable
ediff-use-last-dir is not
nil, Ediff will offer
previously entered directories as defaults (which will be maintained
separately for each type of file, A, B, or C).
All the above functions use the POSIX
to find differences between two files. They process the
and display it in a convenient form. At present, Ediff understands only
the plain output from diff. Options such as ‘-c’ are not supported,
nor is the format produced by incompatible file comparison programs.
ediff-buffers3 first display the coarse,
line-based difference regions, as reported by the
diff program. The
total number of difference regions and the current difference number are
always displayed in the mode line of the control window.
diff may report fairly large chunks of text as being different,
even though the difference may be localized to just a few words or even
to the white space or line breaks, Ediff further refines the
regions to indicate which exact words differ. If the only difference is
in the white space and line breaks, Ediff says so.
On a color display, fine differences are highlighted with color; on a monochrome display, they are underlined. See Highlighting Difference Regions, for information on how to customize this.
ediff-regions-linewise do comparison on parts of existing Emacs
buffers. The commands
ediff-regions-wordwise could be slow on very large buffers,
as they perform comparison on the basis of words rather than lines.
(Word-wise comparison of large chunks of text is relatively expensive.)
To compare very large regions, use
This command displays differences much like
ediff-patch-buffer apply a
patch to a file or a buffer and then run Ediff on the appropriate
files/buffers, displaying the difference regions.
The entry points
etc., provide a convenient interface for comparing and merging files in
different directories. The user is presented with Dired-like interface from
which one can run a group of related Ediff sessions.
For files under version control,
ediff-revision lets you compare
the file visited by the current buffer to one of its checked-in versions.
You can also compare two checked-in versions of the visited file.
Moreover, the functions
ediff-merge-directory-revisions, etc., let you run a group of
related Ediff sessions by taking a directory and comparing (or merging)
versions of files in that directory.
3 Session Commands
All Ediff commands are displayed in a Quick Help window, unless you type ? to shrink the window to just one line. You can redisplay the help window by typing ? again. The Quick Help commands are detailed below.
Many Ediff commands take numeric prefix arguments. For instance, if you
type a number, say 3, and then j (
Ediff moves to the third difference region. Typing 3 and then a
ediff-diff-to-diff) copies the 3rd difference region from variant A
to variant B. Likewise, 4 followed by ra restores the 4th difference
region in buffer A (if it was previously written over via the command
Some commands take negative prefix arguments as well. For instance, typing - and then j will make the last difference region current. Typing -2 then j makes the penultimate difference region current, etc.
Without the prefix argument, all commands operate on the currently selected difference region. You can make any difference region current using the various commands explained below.
For some commands, the actual value of the prefix argument is immaterial. However, if supplied, the prefix argument may modify the command (see ga, gb, and gc).
3.1 Quick Help Commands
- ? ¶
Toggles the Ediff Quick Help window ON and OFF.
- G ¶
Prepares a mail buffer for sending a praise or a curse to the Ediff maintainer.
- E ¶
Brings up the top node of this manual, where you can find further information on the various Ediff functions and advanced issues, such as customization, session groups, etc.
- v ¶
Scrolls up buffers A and B (and buffer C where appropriate) in a coordinated fashion.
- V ¶
Scrolls the buffers down.
- < ¶
Scrolls the buffers to the left simultaneously.
- > ¶
Scrolls buffers to the right.
- wd ¶
Saves the output from the diff utility, for further reference.
With prefix argument, saves the plain output from
ediff-diff-options). Without the argument, it saves customized
ediff-custom-diff-options), if it is available.
- wa ¶
Saves buffer A, if it was modified.
- wb ¶
Saves buffer B, if it was modified.
- wc ¶
Saves buffer C, if it was modified (if you are in a session that compares three files simultaneously).
- a ¶
In comparison sessions: Copies the current difference region (or the region specified as the prefix to this command) from buffer A to buffer B. Ediff saves the old contents of buffer B’s region; it can be restored via the command rb, which see.
In merge sessions: Copies the current difference region (or the region specified as the prefix to this command) from buffer A to the merge buffer. The old contents of this region in buffer C can be restored via the command r.
- b ¶
Works similarly, but copies the current difference region from buffer B to buffer A (in comparison sessions) or the merge buffer (in merge sessions).
Ediff saves the old contents of the difference region copied over; it can be reinstated via the command ra in comparison sessions and r in merge sessions.
- ab ¶
Copies the current difference region (or the region specified as the prefix to this command) from buffer A to buffer B. This (and the next five) command is enabled only in sessions that compare three files simultaneously. The old region in buffer B is saved and can be restored via the command rb.
- ac ¶
Copies the difference region from buffer A to buffer C. The old region in buffer C is saved and can be restored via the command rc.
- ba ¶
Copies the difference region from buffer B to buffer A. The old region in buffer A is saved and can be restored via the command ra.
- bc ¶
Copies the difference region from buffer B to buffer C. The command rc undoes this.
- ca ¶
Copies the difference region from buffer C to buffer A. The command ra undoes this.
- cb ¶
Copies the difference region from buffer C to buffer B. The command rb undoes this.
- p ¶
Makes the previous difference region current.
- n ¶
Makes the next difference region current.
- j ¶
Makes the very first difference region current.
-j makes the last region current. Typing a number, N, and then j makes the difference region N current. Typing −N (a negative number) then j makes current the region Last − N.
- ga ¶
Makes current the difference region closest to the position of the point in buffer A.
However, with a prefix argument, Ediff would position all variants around the area indicated by the current point in buffer A: if the point is inside a difference region, then the variants will be positioned at this difference region. If the point is not in any difference region, then it is in an area where all variants agree with each other. In this case, the variants will be positioned so that each would display this area (of agreement).
- gb ¶
Makes current the difference region closest to the position of the point in buffer B.
With a prefix argument, behaves like ga, but with respect to buffer B.
- gc ¶
In merge sessions: makes current the difference region closest to the point in the merge buffer.
In 3-file comparison sessions: makes current the region closest to the point in buffer C.
With a prefix argument, behaves like ga, but with respect to buffer C.
- ! ¶
Recomputes the difference regions, bringing them up to date. This is often needed because it is common to do all sorts of editing during Ediff sessions, so after a while, the highlighted difference regions may no longer reflect the actual differences among the buffers.
- * ¶
Forces refinement of the current difference region, which highlights the exact words of disagreement among the buffers. With a negative prefix argument, unhighlights the current region.
Forceful refinement may be needed if Ediff encounters a difference region that is larger than
ediff-auto-refine-limit. In this situation, Ediff doesn’t do automatic refinement in order to improve response time. (Ediff doesn’t auto-refine on dumb terminals as well, but * still works there. However, the only useful piece of information it can tell you is whether or not the difference regions disagree only in the amount of white space.)
This command is also useful when the highlighted fine differences are no longer current, due to user editing.
- m ¶
Displays the current Ediff session in a frame as wide as the physical display. This is useful when comparing files side-by-side. Typing m again restores the original size of the frame.
- | ¶
Toggles the horizontal/vertical split of the Ediff display. Horizontal split is convenient when it is possible to compare files side-by-side. If the frame in which files are displayed is too narrow and lines are cut off, typing m may help some.
- @ ¶
Toggles auto-refinement of difference regions (i.e., automatic highlighting of the exact words that differ among the variants). Auto-refinement is turned off on devices where Emacs doesn’t support highlighting.
On slow machines, it may be advantageous to turn auto-refinement off. The user can always forcefully refine specific difference regions by typing *.
- h ¶
Cycles between full highlighting, the mode where fine differences are not highlighted (but computed), and the mode where highlighting is done with ASCII strings. The latter is not really recommended, unless on a dumb TTY.
- r ¶
Restores the old contents of the region in the merge buffer. (If you copied a difference region from buffer A or B into the merge buffer using the commands a or b, Ediff saves the old contents of the region in case you change your mind.)
This command is enabled in merge sessions only.
- ra ¶
Restores the old contents of the current difference region in buffer A, which was previously saved when the user invoked one of these commands: b, ba, ca, which see. This command is enabled in comparison sessions only.
- rb ¶
Restores the old contents of the current difference region in buffer B, which was previously saved when the user invoked one of these commands: a, ab, cb, which see. This command is enabled in comparison sessions only.
- rc ¶
Restores the old contents of the current difference region in buffer C, which was previously saved when the user invoked one of these commands: ac, bc, which see. This command is enabled in 3-file comparison sessions only.
- ## ¶
Tell Ediff to skip over regions that disagree among themselves only in the amount of white space and line breaks.
Even though such regions will be skipped over, you can still jump to any one of them by typing the region number and then j. Typing ## again puts Ediff back in the original state.
- #c ¶
Toggle case sensitivity in the diff program. All diffs are recomputed. Case sensitivity is controlled by the variables
ediff-ignore-case, which are explained elsewhere.
- #h ¶
Ediff works hard to ameliorate the effects of boredom in the workplace...
Quite often differences are due to identical replacements (e.g., the word “foo” is replaced with the word “bar” everywhere). If the number of regions with such boring differences exceeds your tolerance threshold, you may be tempted to tell Ediff to skip these regions altogether (you will still be able to jump to them via the command j). The above commands, #h and #f, may well save your day!
#h prompts you to specify regular expressions for each variant. Difference regions where each variant’s region matches the corresponding regular expression will be skipped from then on. (You can also tell Ediff to skip regions where at least one variant matches its regular expression.)
#f does dual job: it focuses on regions that match the corresponding regular expressions. All other regions will be skipped over. See Selective Browsing, for more.
- A ¶
Toggles the read-only property in buffer A. If file A is under version control and is checked in, it is checked out (with your permission).
- B ¶
Toggles the read-only property in buffer B. If file B is under version control and is checked in, it is checked out.
- C ¶
Toggles the read-only property in buffer C (in 3-file comparison sessions). If file C is under version control and is checked in, it is checked out.
- ~ ¶
Swaps the windows where buffers A and B are displayed. If you are comparing three buffers at once, then this command would rotate the windows among buffers A, B, and C.
- i ¶
Displays all kinds of useful data about the current Ediff session.
- D ¶
ediff-custom-diff-programon the variants and displays the buffer containing the output. This is useful when you must send the output to your Mom.
With a prefix argument, displays the plain
diffoutput. See Patch and Diff Programs, for details.
- R ¶
Displays a list of currently active Ediff sessions—the Ediff Registry. You can then restart any of these sessions by either clicking on a session record or by putting the cursor over it and then typing the return key.
(Some poor souls leave so many active Ediff sessions around that they lose track of them completely... The R command is designed to save these people from the recently discovered Ediff Proficiency Syndrome.)
Typing R brings up Ediff Registry only if it is typed into an Ediff Control Panel. If you don’t have a control panel handy, type this in the minibuffer: M-x eregistry. See Registry of Ediff Sessions.
- M ¶
Shows the session group buffer that invoked the current Ediff session. See Session Groups, for more information on session groups.
- z ¶
Suspends the current Ediff session. (If you develop a condition known as Repetitive Ediff Injury—a serious but curable illness—you must change your current activity. This command tries hard to hide all Ediff-related buffers.)
The easiest way to resume a suspended Ediff session is through the registry of active sessions. See Registry of Ediff Sessions, for details.
- q ¶
Terminates this Ediff session. With a prefix argument (e.g.,1q), asks if you also want to delete the buffers of the variants. Modified files and the results of merges are never deleted.
- % ¶
Toggles narrowing in Ediff buffers. Ediff buffers may be narrowed if you are comparing only parts of these buffers via the commands
ediff-regions-*, which see.
- C-l ¶
Restores the usual Ediff window setup. This is the quickest way to resume an Ediff session, but it works only if the control panel of that session is visible.
- $$ ¶
While merging with an ancestor file, Ediff is determined to reduce user’s wear and tear by saving him and her much of unproductive, repetitive typing. If it notices that, say, file A’s difference region is identical to the same difference region in the ancestor file, then the merge buffer will automatically get the difference region taken from buffer B. The rationale is that this difference region in buffer A is as old as that in the ancestor buffer, so the contents of that region in buffer B represents real change.
You may want to ignore such “obvious” merges and concentrate on difference regions where both files “clash” with the ancestor, since this means that two different people have been changing this region independently and they had different ideas on how to do this.
The above command does this for you by skipping the regions where only one of the variants clashes with the ancestor but the other variant agrees with it. Typing $$ again undoes this setting.
- $* ¶
When merging files with large number of differences, it is sometimes convenient to be able to skip the difference regions for which you already decided which variant is most appropriate. Typing $* will accomplish precisely this.
To be more precise, this toggles the check for whether the current merge is identical to its default setting, as originally decided by Ediff. For instance, if Ediff is merging according to the “combined” policy, then the merge region is skipped over if it is different from the combination of the regions in buffers A and B. (Warning: swapping buffers A and B will confuse things in this respect.) If the merge region is marked as “prefer-A” then this region will be skipped if it differs from the current difference region in buffer A, etc.
- / ¶
Toggle to display the ancestor file in 3way merges. You can enable permanently this setting customizing the variable
- & ¶
In some situations, such as when one of the files agrees with the ancestor file on a difference region and the other doesn’t, Ediff knows what to do: it copies the current difference region from the second buffer into the merge buffer.
In other cases, the right course of action is not that clearcut, and Ediff would use a default action. The above command changes the default action. The default action can be ‘default-A’ (choose the region from buffer A), ‘default-B’ (choose the region from buffer B), or ‘combined’ (combine the regions from the two buffers). See Merging and diff3, for further details.
The command & also affects the regions in the merge buffers that have ‘default-A’, ‘default-B’, or ‘combined’ status, provided they weren’t changed with respect to the original. For instance, if such a region has the status ‘default-A’ then changing the default action to ‘default-B’ will also replace this merge-buffer’s region with the corresponding region from buffer B.
- s ¶
Causes the merge window shrink to its minimum size, thereby exposing as much of the variant buffers as possible. Typing s again restores the original size of that window.
With a positive prefix argument, this command enlarges the merge window. E.g., 4s increases the size of the window by about 4 lines, if possible. With a negative numeric argument, the size of the merge window shrinks by that many lines, if possible. Thus, -s shrinks the window by about 1 line and -3s by about 3 lines.
This command is intended only for temporary viewing; therefore, Ediff restores window C to its original size whenever it makes any other change in the window configuration. However, redisplaying (C-l) or jumping to another difference does not affect window C’s size.
The split between the merge window and the variant windows is controlled by the variable
ediff-merge-window-share, which see.
- + ¶
Combines the difference regions from buffers A and B and copies the result into the merge buffer. See Merging and diff3, and the variables
- = ¶
You may run into situations when a large chunk of text in one file has been edited and then moved to a different place in another file. In such a case, these two chunks of text are unlikely to belong to the same difference region, so the refinement feature of Ediff will not be able to tell you what exactly differs inside these chunks. Since eyeballing large pieces of text is contrary to human nature, Ediff has a special command to help reduce the risk of developing a cataract.
In other situations, the currently highlighted region might be big and you might want to reconcile of them interactively.
All of this can be done with the above command, =, which compares regions within Ediff buffers. Typing = creates a child Ediff session for comparing regions in buffers A, B, or C as follows.
First, you will be asked whether you want to compare the fine differences between the currently highlighted buffers on a word-by-word basis. If you accept, a child Ediff session will start using the currently highlighted regions. Ediff will let you step over the differences word-wise.
If you reject the offer, you will be asked to select regions of your choice.
If you are comparing 2 files or buffers: Ediff will ask you to select regions in buffers A and B.
If you are comparing 3 files or buffers simultaneously: Ediff will ask you to choose buffers and then select regions inside those buffers.
If you are merging files or buffers (with or without ancestor): Ediff will ask you to choose which buffer (A or B) to compare with the merge buffer and then select regions in those buffers.
3.2 Other Session Commands
The following commands can be invoked from within any Ediff session, although some of them are not bound to a key.
This command brings up the registry of active Ediff sessions. Ediff registry is a device that can be used to resume any active Ediff session (which may have been postponed because the user switched to some other activity). This command is also useful for switching between multiple active Ediff sessions that are run at the same time. The function
eregistryis an alias for
ediff-show-registry. See Registry of Ediff Sessions, for more information on this registry.
Changes the display from the multi-frame mode (where the quick help window is in a separate frame) to the single-frame mode (where all Ediff buffers share the same frame), and vice versa. See
ediff-window-setup-functionfor details on how to make either of these modes the default one.
This function can also be invoked from the Menubar. However, in some cases, the change will take place only after you execute one of the Ediff commands, such as going to the next difference or redisplaying.
This command reverts the buffers you are comparing and recomputes their differences. It is useful when, after making changes, you decided to make a fresh start, or if at some point you changed the files being compared but want to discard any changes to comparison buffers that were done since then.
This command normally asks for confirmation before reverting files. With a prefix argument, it reverts files without asking.
Ediff has an admittedly primitive (but useful) facility for profiling Ediff’s commands. It is meant for Ediff maintenance—specifically, for making it run faster. The function
ediff-profiletoggles profiling of ediff commands.
4 Registry of Ediff Sessions
Ediff maintains a registry of all its invocations that are still active. This feature is very convenient for switching among active Ediff sessions or for quickly restarting a suspended Ediff session.
The focal point of this activity is a buffer called *Ediff Registry*. You can display this buffer by typing R in any Ediff Control Buffer or Session Group Buffer (see Session Groups), or by typing M-x eregistry into the Minibuffer. The latter would be the fastest way to bring up the registry buffer if no control or group buffer is displayed in any of the visible Emacs windows. If you are in a habit of running multiple long Ediff sessions and often need to suspend, resume, or switch between them, it may be a good idea to have the registry buffer permanently displayed in a separate, dedicated window.
The registry buffer has several convenient key bindings. For instance, clicking mouse button 2 or typing RET or v over any session record resumes that session. Session records in the registry buffer provide a fairly complete description of each session, so it is usually easy to identify the right session to resume.
Other useful commands are bound to SPC (next registry record) and DEL (previous registry record). There are other commands as well, but you don’t need to memorize them, since they are listed at the top of the registry buffer.
5 Session Groups
Several major entries of Ediff perform comparison and merging on
directories. On entering
the user is presented with a
Dired-like buffer that lists files common to the directories involved along
with their sizes. (The list of common files can be further filtered through
a regular expression, which the user is prompted for.) We call this buffer
Session Group Panel because all Ediff sessions associated with the
listed files will have this buffer as a common focal point.
Clicking button 2 or typing RET or v over a record describing files invokes Ediff in the appropriate mode on these files. You can come back to the session group buffer associated with a particular invocation of Ediff by typing M in Ediff control buffer of that invocation.
Many commands are available in the session group buffer; some are applicable only to certain types of work. The relevant commands are always listed at the top of each session group buffer, so there is no need to memorize them.
In directory comparison or merging, a session group panel displays only the files common to all directories involved. The differences are kept in a separate directory difference buffer and are conveniently displayed by typing D to the corresponding session group panel. Thus, as an added benefit, Ediff can be used to compare the contents of up to three directories.
Sometimes it is desirable to copy some files from one directory to another without exiting Ediff. The directory difference buffer, which is displayed by typing D as discussed above, can be used for this purpose. If a file is, say, in Ediff’s Directory A, but is missing in Ediff’s Directory B (Ediff will refuse to override existing files), then typing C or clicking mouse button 2 over that file (which must be displayed in directory difference buffer) will copy that file from Directory A to Directory B.
Session records in session group panels are also marked with +, for active sessions, and with -, for finished sessions.
Sometimes, it is convenient to exclude certain sessions from a group. Usually this happens when the user doesn’t intend to run Ediff of certain files in the group, and the corresponding session records just add clutter to the session group buffer. To help alleviate this problem, the user can type h to mark a session as a candidate for exclusion and x to actually hide the marked sessions. There actions are reversible: with a prefix argument, h unmarks the session under the cursor, and x brings the hidden sessions into the view (x doesn’t unmark them, though, so the user has to explicitly unmark the sessions of interest).
Group sessions also understand the command m, which marks sessions for future operations (other than hiding) on a group of sessions. At present, the only such group-level operation is the creation of a multi-file patch.
For group sessions created to merge files, Ediff can store all merges
automatically in a directory. The user is asked to specify such directory
if the value of
ediff-autostore-merges is non-
nil. If the value is
nil, nothing is done to the merge buffers—it will be the user’s
responsibility to save them. If the value is
t, the user will be
asked where to save the merge buffers in all merge jobs, even those that do
not originate from a session group. It the value is neither
t, the merge buffer is saved only if this merge session was
invoked from a session group. This behavior is implemented in the function
ediff-maybe-save-and-delete-merge, which is a hook in
ediff-quit-merge-hook. The user can supply a different hook, if
ediff-autostore-merges is buffer-local, so it can be
set on a per-buffer basis. Therefore, use
setq-default to change
this variable globally.
A multi-file patch is a concatenated output of several runs of the Unix
diff command (some versions of
diff let you create a
multi-file patch in just one run). Ediff facilitates creation of
multi-file patches as follows. If you are in a session group buffer
created in response to
ediff-directory-revisions, you can mark (by typing m) the
desired Ediff sessions and then type P to create a
multi-file patch of those marked sessions.
Ediff will then display a buffer containing the patch.
The patch is generated by invoking
diff on all marked individual
sessions (represented by files) and session groups (represented by
directories). Ediff will also recursively descend into any unmarked
session group and will search for marked sessions there. In this way, you
can create multi-file patches that span file subtrees that grow out of
any given directory.
ediff-directories session, it is enough to just mark the
requisite sessions. In
ediff-directory-revisions revisions, the
marked sessions must also be active, or else Ediff will refuse to produce a
multi-file patch. This is because, in the latter-style sessions, there are
many ways to create diff output, and it is easier to handle by running
Ediff on the inactive sessions.
Last, but not least, by typing ==, you can quickly find out which sessions have identical entries, so you won’t have to run Ediff on those sessions. This, however, works only on local, uncompressed files. For compressed or remote files, this command won’t report anything. Likewise, you can use =h to mark sessions with identical entries for hiding or, with =m, for further operations.
The comparison operations ==, =h, and =m can recurse into subdirectories to see if they have identical contents (so the user will not need to descend into those subdirectories manually). These commands ask the user whether or not to do a recursive descent.
6 Remote and Compressed Files
Ediff works with remote, compressed, and encrypted files. Ediff supports ange-ftp.el, jka-compr.el, uncompress.el and crypt++.el, but it may work with other similar packages as well. This means that you can compare files residing on another machine, or you can apply a patch to a file on another machine. Even the patch itself can be a remote file!
When patching compressed or remote files, Ediff does not rename the source
file (unlike what the
patch utility would usually do). Instead, the
source file retains its name and the result of applying the patch is placed
in a temporary file that has the suffix _patched attached.
Generally, this applies to files that are handled using black magic, such
as special file name handlers (ange-ftp and some compression and encryption
packages also use this method).
Regular files are treated by the
patch utility in the usual manner,
i.e., the original is renamed into source-name.orig and the result
of the patch is placed into the file source-name (_orig is used
on systems like DOS, etc.).
Ediff has a rather self-explanatory interface, and in most cases you won’t need to change anything. However, should the need arise, there are extensive facilities for changing the default behavior.
Most of the customization can be done by setting various variables in the .emacs file. Some customization (mostly window-related customization and faces) can be done by putting appropriate lines in .Xdefaults, .xrdb, or whatever X resource file is in use.
With respect to the latter, please note that the X resource for Ediff customization is “Ediff”, not “emacs”. See Window and Frame Configuration, See Highlighting Difference Regions, for further details. Please also refer to Emacs manual for the information on how to set Emacs X resources.
- Quick Help Customization
- Window and Frame Configuration
- Selective Browsing
- Highlighting Difference Regions
- Refinement of Difference Regions
- Patch and Diff Programs
- Merging and diff3
- Support for Version Control
- Customizing the Mode Line
- Notes on Heavy-duty Customization
The bulk of customization can be done via the following hooks:
Hook that is run just before Ediff rearranges windows to its liking. Can be used to save windows configuration.
This hook can be used to alter bindings in Ediff’s keymap,
ediff-mode-map. These hooks are run right after the default bindings are set. The regular user needs not be concerned with this hook—it is provided for implementers of other Emacs packages built on top of Ediff.
These two hooks are called before and after Ediff sets up its window configuration. These hooks are run each time Ediff rearranges windows to its liking. This happens whenever it detects that the user changed the windows setup.
These two hooks are run when you suspend or quit Ediff. They can be used to set desired window configurations, delete files Ediff didn’t want to clean up after exiting, etc.
ediff-quit-hookholds one hook function,
ediff-cleanup-mess, which cleans after Ediff, as appropriate in most cases. You probably won’t want to change it, but you might want to add other hook functions.
Keep in mind that hooks executing before
ediff-control-buffer;they should also leave
ediff-control-bufferas the current buffer when they finish. Hooks that are executed after
ediff-cleanup-messshould expect the current buffer be either buffer A or buffer B.
ediff-cleanup-messdoesn’t kill the buffers being compared or merged (see
This hook is run just before
ediff-quit-hook. This is a good place to do various cleanups, such as deleting the variant buffers. Ediff provides a helper function,
ediff-janitor, that you can invoke from a private hook function. For example:
(defun my-ediff-janitor () (ediff-janitor nil nil)) (add-hook 'ediff-cleanup-hook #'my-ediff-janitor)
This function kills buffers A, B, and, possibly, C, if these buffers aren’t modified. In merge jobs, buffer C is never deleted. However, the side effect of using this function is that you may not be able to compare the same buffer in two separate Ediff sessions: quitting one of them will delete this buffer in another session as well.
This hook is called when Ediff quits a merge job. By default, the value is
ediff-maybe-save-and-delete-merge, which is a function that attempts to save the merge buffer according to the value of
ediff-autostore-merges, as described later.
These two hooks run before and after Ediff sets up the control frame. They can be used to relocate Ediff control frame when Ediff runs in a multiframe mode (i.e., when the control buffer is in its own dedicated frame). Be aware that many variables that drive Ediff are local to Ediff Control Panel (
ediff-control-buffer), which requires special care in writing these hooks. Take a look at
ediff-default-quit-hookto see what’s involved.
This hook is run at the end of Ediff startup.
This hook is run after Ediff selects the next difference region.
This hook is run after Ediff unselects the current difference region.
This hook is run for each Ediff buffer (A, B, C) right after the buffer is arranged.
Ediff runs this hook each time after setting up the help message. It can be used to alter the help message for custom packages that run on top of Ediff.
This hook is run just after Ediff mode is set up in the control buffer. This is done before any Ediff window is created. You can use it to set local variables that alter the look of the display.
Hooks run after setting up the registry for all active Ediff session. See Session Groups, for details.
Hooks run before setting up a control panel for a group of related Ediff sessions. Can be used, for example, to save window configuration to restore later.
Hooks run after setting up a control panel for a group of related Ediff sessions. See Session Groups, for details.
Hooks run just before exiting a session group.
Hooks run just after setting up the
ediff-meta-buffer-map, the map that controls key bindings in the meta buffer. Since
ediff-meta-buffer-mapis a local variable, you can set different bindings for different kinds of meta buffers.
7.2 Quick Help Customization
Ediff provides quick help using its control panel window. Since this window takes a fair share of the screen real estate, you can toggle it off by typing ?. The control window will then shrink to just one line and a mode line, displaying a short help message.
ediff-use-long-help-message tells Ediff whether
you use the short message or the long one. By default, it
is set to
nil, meaning that the short message is used.
Set this to
t, if you want Ediff to use the long
message by default. This property can always be changed interactively, by
typing ? into Ediff Control Buffer.
If you want to change the appearance of the help message on a per-buffer
basis, you must use
ediff-startup-hook to change the value of
ediff-help-message, which is local to
7.3 Window and Frame Configuration
On a non-windowing display, Ediff sets things up in one frame, splitting it between a small control window and the windows for buffers A, B, and C. The split between these windows can be horizontal or vertical, which can be changed interactively by typing | while the cursor is in the control window.
On a window display, Ediff sets up a dedicated frame for Ediff Control
Panel and then it chooses windows as follows: If one of the buffers
is invisible, it is displayed in the currently selected frame. If
a buffer is visible, it is displayed in the frame where it is visible.
If, according to the above criteria, the two buffers fall into the same
frame, then so be it—the frame will be shared by the two. The same
algorithm works when you type C-l (
The above behavior also depends on whether the current frame is splittable, dedicated, etc. Unfortunately, the margin of this book is too narrow to present the details of this remarkable algorithm.
The upshot of all this is that you can compare buffers in one frame or in different frames. The former is done by default, while the latter can be achieved by arranging buffers A, B (and C, if applicable) to be seen in different frames. Ediff respects these arrangements, automatically adapting itself to the multi-frame mode.
Ediff uses the following variables to set up its control panel (a.k.a. “control buffer”, a.k.a. “quick help window”):
You can change or augment this variable including the font, color, etc. The X resource name of Ediff Control Panel frames is ‘Ediff’. Under X-windows, you can use this name to set up preferences in your ~/.Xdefaults, ~/.xrdb, or whatever X resource file is in use. Usually this is preferable to changing
ediff-control-frame-parametersdirectly. For instance, you can specify in ~/.Xdefaults the color of the control frame using the resource ‘Ediff*background’.
In general, any X resource pertaining the control frame can be reached via the prefix
The preferred way of specifying the position of the control frame is by setting the variable
ediff-control-frame-position-functionto an appropriate function. The default value of this variable is
ediff-make-frame-position. This function places the control frame in the vicinity of the North-East corner of the frame displaying buffer A.
The following variables can be used to adjust the location produced by
ediff-make-frame-position and for related customization.
Specifies the number of characters for shifting the control frame from the rightmost edge of frame A when the control frame is displayed as a small window.
Specifies the rightward shift of the control frame from the left edge of frame A when the control frame shows the full menu of options.
Specifies the number of pixels for the upward shift of the control frame.
If this variable is
t, the control frame becomes iconified automatically when you toggle the quick help message off. This saves valuable real estate on the screen. Toggling help back will deiconify the control frame.
To start Ediff with an iconified Control Panel, you should set this variable to
nil(see Quick Help Customization). This behavior is useful only if icons are allowed to accept keyboard input (which depends on the window manager and other factors).
To make more creative changes in the way Ediff sets up windows, you can
rewrite the function
ediff-setup-windows. However, we believe
that detaching Ediff Control Panel from the rest and making it into a
separate frame offers an important opportunity by allowing you to
iconify that frame. The icon will usually accept all of the Ediff
commands, but will free up valuable real estate on your screen (this may
depend on your window manager, though).
The following variable controls how windows are set up:
The multiframe setup is done by the
ediff-setup-windows-multiframefunction, which is the default on windowing displays. The plain setup, one where all windows are always in one frame, is done by
ediff-setup-windows-plain, which is the default on a non-windowing display (or in an xterm window). In fact, under Emacs, you can switch freely between these two setups by executing the command
ediff-toggle-multiframeusing the Minibuffer of the Menubar.
If you don’t like any of these setups, write your own function. See the documentation for
ediff-window-setup-functionfor the basic guidelines. However, writing window setups is not easy, so you should first take a close look at
You can run multiple Ediff sessions at once, by invoking Ediff several times without exiting previous Ediff sessions. Different sessions may even operate on the same pair of files.
Each session has its own Ediff Control Panel and all the regarding a particular session is local to the associated control panel buffer. You can switch between sessions by suspending one session and then switching to another control panel. (Different control panel buffers are distinguished by a numerical suffix, e.g., ‘Ediff Control Panel<3>’.)
7.4 Selective Browsing
Sometimes it is convenient to be able to step through only some difference regions, those that match certain regular expressions, and to ignore all others. On other occasions, you may want to ignore difference regions that match some regular expressions, and to look only at the rest.
The commands #f and #h let you do precisely this.
Typing #f lets you specify regular expressions that match difference regions you want to focus on. We shall call these regular expressions regexp-A, regexp-B and regexp-C. Ediff will then start stepping through only those difference regions where the region in buffer A matches regexp-A and/or the region in buffer B matches regexp-B, etc. Whether “and” or “or” will be used depends on how you respond to a question.
When scanning difference regions for the aforesaid regular expressions, Ediff narrows the buffers to those regions. This means that you can use the expressions \` and \' to tie search to the beginning or end of the difference regions.
On the other hand, typing #h lets you specify (hide) uninteresting
regions. That is, if a difference region in buffer A matches
regexp-A, the corresponding region in buffer B matches regexp-B
and (if applicable) buffer C’s region matches regexp-C, then the
region will be ignored by the commands n/SPC
ediff-next-difference) and p/DEL
Typing #f and #h toggles selective browsing on and off.
Note that selective browsing affects only
ediff-previous-difference, i.e., the commands
n/SPC and p/DEL. #f and #h do not
change the position of the point in the buffers. And you can still jump
directly (using j) to any numbered
Users can supply their own functions to specify how Ediff should do selective browsing. To change the default Ediff function, use something like the following:
(with-eval-after-load 'ediff (setq ediff-hide-regexp-matches-function 'your-hide-function) (setq ediff-focus-on-regexp-matches-function 'your-focus-function))
Useful hint: To specify a regexp that matches everything, don’t simply type RET in response to a prompt. Typing RET tells Ediff to accept the default value, which may not be what you want. Instead, you should enter something like ^ or $. These match every line.
You can use the status command, i, to find out whether selective browsing is currently in effect.
The regular expressions you specified are kept in the local variables
ediff-regexp-hide-C. Their default value
is the empty string (i.e., nothing is hidden or focused on). To change the
default, set these variables in .emacs using
In addition to the ability to ignore regions that match regular expressions, Ediff can be ordered to start skipping over certain “uninteresting” difference regions. This is controlled by the following variable:
t, causes Ediff to skip over "uninteresting" difference regions, which are the regions where the variants differ only in the amount of the white space and newlines. This feature can be toggled on/off interactively, via the command ##.
Please note: in order for this feature to work, auto-refining of difference regions must be on, since otherwise Ediff won’t know if there are fine differences between regions. On devices where Emacs can display faces, auto-refining is a default, but it is not turned on by default on text-only terminals. In that case, you must explicitly turn auto-refining on (such as, by typing @).
Reassurance: If many such uninteresting regions appear in a row, Ediff may take a long time to skip over them because it has to compute fine differences of all intermediate regions. This delay does not indicate any problem.
Finally, Ediff can be told to ignore the case of the letters. This behavior
can be toggled with #c and it is controlled with three variables:
ediff-ignore-case-option specifies the option to pass
to the diff program for comparing two files or buffers. For GNU
diff, this option is
"-i". The variable
ediff-ignore-case-option3 specifies the option to pass to the
diff3 program in order to make it case-insensitive. GNU
does not have such an option, so when merging or comparing three files with
this program, ignoring the letter case is not supported.
ediff-ignore-case controls whether Ediff starts out by
ignoring letter case or not. It can be set in .emacs using
When case sensitivity is toggled, all difference regions are recomputed.
7.5 Highlighting Difference Regions
The following variables control the way Ediff highlights difference regions:
These variables hold strings that Ediff uses to mark the beginning and the end of the differences found in files A, B, and C on devices where Emacs cannot display faces. Ediff uses different flags to highlight regions that begin/end at the beginning/end of a line or in a middle of a line.
Ediff uses these faces to highlight current differences on devices where Emacs can display faces. These and subsequently described faces can be set either in .emacs or in .Xdefaults. The X resource for Ediff is ‘Ediff’, not ‘emacs’. Please refer to Emacs manual for the information on how to set X resources.
Ediff uses these faces to show the fine differences between the current differences regions in buffers A, B, and C, respectively.
Non-current difference regions are displayed using these alternating faces. The odd and the even faces are actually identical on monochrome displays, because without colors options are limited. So, Ediff uses italics to highlight non-current differences.
Ediff generally can detect when Emacs is running on a device where it can use highlighting with faces. However, if it fails to determine that faces can be used, the user can set this variable to
tto make sure that Ediff uses faces to highlight differences.
Indicates whether—on a windowing display—Ediff should highlight differences using inserted strings (as on text-only terminals) or using colors and highlighting. Normally, Ediff highlights all differences, but the selected difference is highlighted more visibly. One can cycle through various modes of highlighting by typing h. By default, Ediff starts in the mode where all difference regions are highlighted. If you prefer to start in the mode where unselected differences are not highlighted, you should set
nil. Type h to restore highlighting for all differences.
Ediff lets you switch between the two modes of highlighting. That is, you can switch interactively from highlighting using faces to highlighting using string flags, and back. Of course, switching has effect only under a windowing system. On a text-only terminal or in an xterm window, the only available option is highlighting with strings.
If you want to change the default settings for
ediff-highlight-all-diffs, you must do it before Ediff is
You can also change the defaults for the faces used to highlight the difference regions. There are two ways to do this. The simplest and the preferred way is to use the customization widget accessible from the menubar. Ediff’s customization group is located under "Tools", which in turn is under "Programming". The faces that are used to highlight difference regions are located in the "Highlighting" subgroup of the Ediff customization group.
The second, much more arcane, method to change default faces is to include some Lisp code in ~/.emacs. For instance,
(setq ediff-current-diff-face-A (copy-face 'bold-italic 'ediff-current-diff-face-A))
would use the pre-defined face
bold-italic to highlight the current
difference region in buffer A (this face is not a good choice, by the way).
If you are unhappy with just some of the aspects of the default
faces, you can modify them when Ediff is being loaded using
with-eval-after-load. For instance:
(with-eval-after-load 'ediff (set-face-foreground ediff-current-diff-face-B "blue") (set-face-background ediff-current-diff-face-B "red") (make-face-italic ediff-current-diff-face-B))
If buffers being compared are narrowed at the time of invocation of
ediff-buffers will preserve the narrowing range. However,
ediff-files is invoked on the files visited by these buffers,
that would widen the buffers, since this command is defined to compare the
the corresponding ‘-wordwise’ commands, narrows the variants to the
particular regions being compared. The original accessible ranges are
restored when you quit Ediff. During the command, you can toggle this
narrowing on and off with the % command.
These two variables control this narrowing behavior:
t, Ediff narrows the display to the appropriate range when it is invoked with an ‘ediff-regions…’ or ‘ediff-windows…’ command. If
nil, these commands do not automatically narrow, but you can still toggle narrowing on and off by typing %.
Controls whether on quitting Ediff should restore the accessible range that existed before the current invocation.
7.7 Refinement of Difference Regions
Ediff has variables to control the way fine differences are highlighted. This feature gives you control over the process of refinement. Note that refinement ignores spaces, tabs, and newlines.
This variable controls whether fine differences within regions are highlighted automatically (“auto-refining”). The default is yes (‘on’).
On a slow machine, automatic refinement may be painful. In that case, you can turn auto-refining on or off interactively by typing @. You can also turn off display of refining that has already been done.
When auto-refining is off, fine differences are shown only for regions for which these differences have been computed and saved before. If auto-refining and display of refining are both turned off, fine differences are not shown at all.
Typing * computes and displays fine differences for the current difference region, regardless of whether auto-refining is turned on.
If auto-refining is on, this variable limits the size of the regions to be auto-refined. This guards against the possible slowdown that may be caused by extraordinary large difference regions.
You can always refine the current region by typing *.
This variable controls how fine differences are computed. The value must be a Lisp function that determines how the current difference region should be split into words.
Fine differences are computed by first splitting the current difference region into words and then passing the result to
ediff-diff-program. For the default forward word function (which is
ediff-forward-word), a word is a string consisting of letters, ‘-’, or ‘_’; a string of punctuation symbols; a string of digits, or a string consisting of symbols that are neither space, nor a letter.
This default behavior is controlled by four variables:
ediff-word-4. See the on-line documentation for these variables and for the function
ediff-forward-wordfor an explanation of how to modify these variables.
Sometimes, when a region has too many differences between the variants, highlighting of fine differences is inconvenient, especially on color displays. If that is the case, type * with a negative prefix argument. This unhighlights fine differences for the current region.
To unhighlight fine differences in all difference regions, use the command @. Repeated typing of this key cycles through three different states: auto-refining, no-auto-refining, and no-highlighting of fine differences.
7.8 Patch and Diff Programs
This section describes variables that specify the programs to be used for applying patches and for computing the main difference regions (not the fine difference regions):
These variables specify the programs to use to produce differences and do patching.
These variables specify the options to pass to the above utilities.
ediff-diff-options, it may be useful to specify options such as ‘-w’ that ignore certain kinds of changes. However, Ediff does not let you use the option ‘-c’, as it doesn’t recognize this format yet. (If you alter this variable, it should be done via the Customize interface instead of using
This variable specifies the coding system to use when reading the output that the programs
diffsend to Emacs. The default is
raw-text, and this should work fine in Unix and in most cases under Windows NT/95/98/2000. There are
diffprograms for which the default option doesn’t work under Windows. In such cases,
raw-text-dosmight work. If not, you will have to experiment with other coding systems or use GNU diff.
The program to use to apply patches. Since there are certain incompatibilities between the different versions of the patch program, the best way to stay out of trouble is to use a GNU-compatible version. Otherwise, you may have to tune the values of the variables
ediff-backup-extensionas described below.
Options to pass to
Note: the -b and -z options should be specified in
ediff-backup-specs, not in
It is recommended to pass the -f option to the patch program, so it won’t ask questions. However, some implementations don’t accept this option, in which case the default value of this variable should be changed.
Backup extension used by the patch program. Must be specified, even if
Backup directives to pass to the patch program. Ediff requires that the old version of the file (before applying the patch) is saved in a file named the-patch-file.extension. Usually extension is .orig, but this can be changed by the user, and may also be system-dependent. Therefore, Ediff needs to know the backup extension used by the patch program.
Some versions of the patch program let the user specify -b extension to specify a backup file name extension. Other versions only permit -b, which (usually) assumes the extension .orig. Yet others force you to use -zextension.
ediff-backup-specsmust be properly set. If your patch program takes the option -b, but not -b extension, the variable
ediff-backup-extensionmust still be set so Ediff will know which extension to use.
Because Ediff limits the options you may want to pass to the
diffprogram, it partially makes up for this drawback by letting you save the output from
diffin your preferred format, which is specified via the above two variables.
The output generated by
ediff-custom-diff-program(which doesn’t even have to be a standard-style
diff!) is not used by Ediff. It is provided exclusively so that you can refer to it later, send it over email, etc. For instance, after reviewing the differences, you may want to send context differences to a colleague. Since Ediff ignores the ‘-c’ option in
ediff-diff-program, you would have to run
diff -cseparately just to produce the list of differences. Fortunately,
ediff-custom-diff-optionseliminate this nuisance by keeping a copy of a difference list in the desired format in a buffer that can be displayed via the command D.
Specifies the default directory to look for patches.
7.9 Merging and diff3
Ediff supports three-way comparison via the functions
ediff-buffers3. The interface is the same as for two-way comparison.
In three-way comparison and merging, Ediff reports if any two difference
regions are identical. For instance, if the current region in buffer A
is the same as the region in buffer C, then the mode line of buffer A will
display ‘[=diff(C)]’ and the mode line of buffer C will display
Merging is done according to the following algorithm.
If a difference region in one of the buffers, say B, differs from the ancestor file while the region in the other buffer, A, doesn’t, then the merge buffer, C, gets B’s region. Similarly when buffer A’s region differs from the ancestor and B’s doesn’t, A’s region is used.
If both regions in buffers A and B differ from the ancestor file, Ediff
chooses the region according to the value of the variable
ediff-default-variant. If its value is
default-A then A’s
region is chosen. If it is
default-B then B’s region is chosen.
If it is
combined then the region in buffer C will look like
<<<<<<< variant A the difference region from buffer A >>>>>>> variant B the difference region from buffer B ####### Ancestor the difference region from the ancestor buffer, if available ======= end
The above is the default template for the combined region. The user can
customize this template using the variable
ediff-combination-pattern specifies the template that
determines how the combined merged region looks like. The template is
represented as a list of the form
(STRING1 Symbol1 STRING2 Symbol2
STRING3 Symbol3 STRING4). The symbols here must be atoms of the form
Ancestor. They determine the order in which
the corresponding difference regions (from buffers A, B, and the ancestor
buffer) are displayed in the merged region of buffer C. The strings in the
template determine the text that separates the aforesaid regions. The
default template is
("<<<<<<< variant A" A ">>>>>>> variant B" B "####### Ancestor" Ancestor "======= end")
(this is one long line) and the corresponding combined region is shown above. The order in which the regions are shown (and the separator strings) can be changed by changing the above template. It is even possible to add or delete region specifiers in this template (although the only possibly useful such modification seems to be the deletion of the ancestor).
In addition to the state of the difference, Ediff displays the state of the
merge for each region. If a difference came from buffer A by default
(because both regions A and B were different from the ancestor and
ediff-default-variant was set to
‘[=diff(A) default-A]’ is displayed in the mode line. If the
difference in buffer C came, say, from buffer B because the difference
region in that buffer differs from the ancestor, but the region in buffer A
does not (if merging with an ancestor) then ‘[=diff(B) prefer-B]’ is
displayed. The indicators default-A/B and prefer-A/B are inspired by
Emerge and have the same meaning.
Another indicator of the state of merge is ‘combined’. It appears with any difference region in buffer C that was obtained by combining the difference regions in buffers A and B as explained above.
In addition to the state of merge and state of difference indicators, while merging with an ancestor file or buffer, Ediff informs the user when the current difference region in the (normally invisible) ancestor buffer is empty via the AncestorEmpty indicator. This helps determine if the changes made to the original in variants A and B represent pure insertion or deletion of text: if the mode line shows AncestorEmpty and the corresponding region in buffers A or B is not empty, this means that new text was inserted. If this indicator is not present and the difference regions in buffers A or B are non-empty, this means that text was modified. Otherwise, the original text was deleted.
Although the ancestor buffer is normally invisible, Ediff maintains difference regions there and advances the current difference region accordingly. All highlighting of difference regions is provided in the ancestor buffer, except for the fine differences. Therefore, if desired, the user can put the ancestor buffer in a separate frame and watch it there. However, on a TTY, only one frame can be visible at any given time, and Ediff doesn’t support any single-frame window configuration where all buffers, including the ancestor buffer, would be visible. However, the ancestor buffer can be displayed by typing / to the control window. (Type C-l to hide it again.)
Note that the state-of-difference indicators ‘=diff(A)’ and ‘=diff(B)’ above are not redundant, even in the presence of a state-of-merge indicator. In fact, the two serve different purposes.
For instance, if the mode line displays ‘=diff(B) prefer(B)’ and you copy a difference region from buffer A to buffer C then ‘=diff(B)’ will change to ‘diff-A’ and the mode line will display ‘=diff(A) prefer-B’. This indicates that the difference region in buffer C is identical to that in buffer A, but originally buffer C’s region came from buffer B. This is useful to know because you can recover the original difference region in buffer C by typing r.
Ediff never changes the state-of-merge indicator, except in response to the ! command (see below), in which case the indicator is lost. On the other hand, the state-of-difference indicator is changed automatically by the copying/recovery commands, a, b, r, +.
The ! command loses the information about origins of the regions
in the merge buffer (default-A, prefer-B, or combined). This is because
recomputing differences in this case means running
buffers A, B, and the merge buffer, not on the ancestor buffer. (It
makes no sense to recompute differences using the ancestor file, since
in the merging mode Ediff assumes that you have not edited buffers A and
B, but that you may have edited buffer C, and these changes are to be
preserved.) Since some difference regions may disappear as a result of
editing buffer C and others may arise, there is generally no simple way
to tell where the various regions in the merge buffer came from.
In three-way comparison, Ediff tries to disregard regions that consist entirely of white space. For instance, if, say, the current region in buffer A consists of the white space only (or if it is empty), Ediff will not take it into account for the purpose of computing fine differences. The result is that Ediff can provide a better visual information regarding the actual fine differences in the non-white regions in buffers B and C. Moreover, if the regions in buffers B and C differ in the white space only, then a message to this effect will be displayed.
In the merge mode, the share of the split between window C (the window
displaying the merge-buffer) and the windows displaying buffers A and B
is controlled by the variable
default value is 0.5. To make the merge-buffer window smaller, reduce
We don’t recommend increasing the size of the merge-window to more than
half the frame (i.e., to increase the value of
ediff-merge-window-share) to more than 0.5, since it would be
hard to see the contents of buffers A and B.
You can temporarily shrink the merge window to just one line by typing s. This change is temporary, until Ediff finds a reason to redraw the screen. Typing s again restores the original window size.
With a positive prefix argument, the s command will make the merge window slightly taller. This change is persistent. With “-” or with a negative prefix argument, the command s makes the merge window slightly shorter. This change also persistent.
Ediff lets you automatically ignore the regions where only one of the
buffers A and B disagrees with the ancestor. To do this, set the
ediff-show-clashes-only to non-
You can toggle this feature interactively by typing $$.
Note that this variable affects only the show next/previous difference commands. You can still jump directly to any difference region directly using the command j (with a prefix argument specifying the difference number).
ediff-autostore-merges controls what happens to the
merge buffer when Ediff quits. If the value is
nil, nothing is done
to the merge buffer—it will be the user’s responsibility to save it.
If the value is
t, the user will be asked where to save the buffer
and whether to delete it afterwards. It the value is neither
t, the merge buffer is saved only if this merge session was
invoked from a group of related Ediff session, such as those that result
See Session Groups. This behavior is implemented in the function
ediff-maybe-save-and-delete-merge, which is a hook in
ediff-quit-merge-hook. The user can supply a different hook, if
ediff-autostore-merges is buffer-local, so it can be
set in a per-buffer manner. Therefore, use
setq-default to globally
change this variable.
When merge buffers are saved automatically as directed by
ediff-autostore-merges, Ediff attaches a prefix to each file, as
specified by the variable
ediff-merge-filename-prefix. The default
merge_, but this can be changed by the user.
7.10 Support for Version Control
Ediff supports version control and lets you compare versions of files
visited by Emacs buffers via the function
feature is controlled by the following variables:
A symbol. The default is ‘vc’.
If you are like most Emacs users, Ediff will use VC as the version control package. This is the standard Emacs interface to RCS, CVS, and SCCS.
However, if your needs are better served by other interfaces, you will have to tell Ediff which version control package you are using, e.g.,
(setq ediff-version-control-package 'rcs)
Apart from the standard vc.el, Ediff supports three other interfaces to version control: rcs.el, pcl-cvs.el (recently renamed pcvs.el), and generic-sc.el. The package rcs.el is written by Sebastian Kremer <sk@thp.Uni-Koeln.DE> and is available as
Ediff’s interface to the above packages allows the user to compare the versions of the current buffer or to merge them (with or without an ancestor-version). These operations can also be performed on directories containing files under version control.
In case of pcl-cvs.el, Ediff can also be invoked via the function
run-ediff-from-cvs-buffer—see the documentation string for this
7.11 Customizing the Mode Line
When Ediff is running, the mode line of ‘Ediff Control Panel’ buffer shows the current difference number and the total number of difference regions in the two files.
The mode line of the buffers being compared displays the type of the
buffer (‘A:’, ‘B:’, or ‘C:’) and (usually) the file name.
Ediff tries to be intelligent in choosing the mode line buffer
identification. In particular, it works well with the
uniquify.el and mode-line.el packages (which improve on
the default way in which Emacs displays buffer identification). If you
don’t like the way Ediff changes the mode line, you can use
ediff-prepare-buffer-hook to modify the mode line.
Here are a few other variables for customizing Ediff:
Controls the way you want the window be split between file-A and file-B (and file-C, if applicable). It defaults to the vertical split (
split-window-vertically, but you can set it to
split-window-horizontally, if you so wish. Ediff also lets you switch from vertical to horizontal split and back interactively.
Note that if Ediff detects that all the buffers it compares are displayed in separate frames, it assumes that the user wants them to be so displayed and stops splitting windows. Instead, it arranges for each buffer to be displayed in a separate frame. You can switch to the one-frame mode by hiding one of the buffers A/B/C.
You can also swap the windows where buffers are displayed by typing ~.
Controls how windows are split between buffers A and B in the merge mode. This variable is like
ediff-split-window-function, but it defaults to
The value is a function to be called to widen the frame for displaying the Ediff buffers. See the on-line documentation for
ediff-make-wide-display-functionfor details. It is also recommended to look into the source of the default function
You can toggle wide/regular display by typing m. In the wide display mode, buffers A, B (and C, when applicable) are displayed in a single frame that is as wide as the entire workstation screen. This is useful when files are compared side-by-side. By default, the display is widened without changing its height.
Controls the way Ediff presents the default directory when it prompts the user for files to compare. If
nil, Ediff uses the default directory of the current buffer when it prompts the user for file names. Otherwise, it will use the directories it had previously used for files A, B, or C, respectively.
t, makes C-h behave like the DEL key, i.e., it will move you back to the previous difference rather than invoking help. This is useful when, in an xterm window or a text-only terminal, the Backspace key is bound to C-h and is positioned more conveniently than the DEL key.
This variable’s value is a function that Ediff uses to toggle the read-only property in its buffers.
The default function that Ediff uses simply toggles the read-only property, unless the file is under version control. For a checked-in file under version control, Ediff first tries to check the file out.
t, all variant buffers are made read-only at Ediff startup.
The default is
t, meaning that the buffers being compared or merged will be preserved when Ediff quits. Setting this to
nilcauses Ediff to offer the user a chance to delete these buffers (if they are not modified). Supplying a prefix argument to the quit command (
q) temporarily reverses the meaning of this variable. This is convenient when the user prefers one of the behaviors most of the time, but occasionally needs the other behavior.
However, Ediff temporarily resets this variable to
tif it is invoked via one of the "buffer" jobs, such as
ediff-buffers. This is because it is all too easy to lose a day’s work otherwise. Besides, in a "buffer" job, the variant buffers have already been loaded prior to starting Ediff, so Ediff just preserves status quo here.
ediff-cleanup-hook, one can make Ediff delete the variants unconditionally (e.g., by making
ediff-janitorinto one of these hooks).
t, the versions of the files being compared or merged using operations such as
ediff-merge-revisionsare not deleted on exit. The normal action is to clean up and delete these version files.
t. Normally, Ediff grabs mouse and puts it in its control frame. This is useful since the user can be sure that when he needs to type an Ediff command the focus will be in an appropriate Ediff’s frame. However, some users prefer to move the mouse by themselves. The above variable, if set to
maybe, will prevent Ediff from grabbing the mouse in many situations, usually after commands that may take more time than usual. In other situation, Ediff will continue grabbing the mouse and putting it where it believes is appropriate. If the value is
nil, then mouse is entirely user’s responsibility. Try different settings and see which one is for you.
7.13 Notes on Heavy-duty Customization
Some users need to customize Ediff in rather sophisticated ways, which
requires different defaults for different kinds of files (e.g., SGML,
etc.). Ediff supports this kind of customization in several ways. First,
most customization variables are buffer-local. Those that aren’t are
usually accessible from within Ediff Control Panel, so one can make them
local to the panel by calling make-local-variable from within
Second, the function
ediff-setup accepts an optional sixth
argument which has the form
((var-name-1 . val-1)
(var-name-2 . val-2) …). The function
ediff-setup sets the variables in the list to the respective
values, locally in the Ediff control buffer. This is an easy way to
throw in custom variables (which usually should be buffer-local) that
can then be tested in various hooks.
Make sure the variable
ediff-word-mode are set
properly in this case, as some things in Ediff depend on this.
When customizing Ediff, some other variables are useful, although they are
not user-definable. They are local to the Ediff control buffer, so this
buffer must be current when you access these variables. The control buffer
is accessible via the variable
ediff-control-buffer, which is also
local to that buffer. It is usually used for checking if the current buffer
is also the control buffer.
Other variables of interest are:
The first of the data buffers being compared.
The second of the data buffers being compared.
In three-way comparisons, this is the third buffer being compared. In merging, this is the merge buffer. In two-way comparison, this variable is
The window displaying buffer A. If buffer A is not visible, this variable is
nilor it may be a dead window.
The window displaying buffer B.
The window displaying buffer C, if any.
A dedicated frame displaying the control buffer, if it exists. It is non-
nilonly if Ediff uses the multiframe display, i.e., when the control buffer is in its own frame.
Ediff was written by Michael Kifer <firstname.lastname@example.org>. It was inspired by emerge.el written by Dale R. Worley <email@example.com>. An idea due to Boris Goldowsky <firstname.lastname@example.org> made it possible to highlight fine differences in Ediff buffers. Alastair Burt <email@example.com> ported Ediff to XEmacs, Eric Freudenthal <firstname.lastname@example.org> made it work with VC, Marc Paquette <email@example.com> wrote the toolbar support package for Ediff, and Hrvoje Nikšić <firstname.lastname@example.org> adapted it to the Emacs customization package.
Many people provided help with bug reports, feature suggestions, and advice. Without them, Ediff would not be nearly as useful as it is today. Here is a hopefully full list of contributors:
Adrian Aichner (aichner at ecf.teradyne.com), Drew Adams (drew.adams at oracle.com), Steve Baur (steve at xemacs.org), Neal Becker (neal at ctd.comsat.com), E. Jay Berkenbilt (ejb at ql.org), Lennart Borgman (ennart.borgman at gmail.com) Alastair Burt (burt at dfki.uni-kl.de), Paul Bibilo (peb at delcam.co.uk), Kevin Broadey (KevinB at bartley.demon.co.uk), Harald Boegeholz (hwb at machnix.mathematik.uni-stuttgart.de), Bradley A. Bosch (brad at lachman.com), Michael D. Carney (carney at ltx-tr.com), Jin S. Choi (jin at atype.com), Scott Cummings (cummings at adc.com), Albert Dvornik (bert at mit.edu), Eric Eide (eeide at asylum.cs.utah.edu), Paul Eggert (eggert at twinsun.com), Urban Engberg (ue at cci.dk), Kevin Esler (esler at ch.hp.com), Robert Estes (estes at ece.ucdavis.edu), Jay Finger (jayf at microsoft.com), Xavier Fornari (xavier at europe.cma.fr), Eric Freudenthal (freudent at jan.ultra.nyu.edu), Job Ganzevoort (Job.Ganzevoort at cwi.nl), Felix Heinrich Gatzemeier (felix.g at tzemeier.info), Boris Goldowsky (boris at cs.rochester.edu), Allan Gottlieb (gottlieb at allan.ultra.nyu.edu), Aaron Gross (aaron at bfr.co.il), Thorbjoern Hansen (thorbjoern.hansen at mchp.siemens.de), Marcus Harnisch (marcus_harnisch at mint-tech.com), Steven E. Harris (seh at panix.com), Aaron S. Hawley (Aaron.Hawley at uvm.edu), Xiaoli Huang (hxl at epic.com), Andreas Jaeger (aj at suse.de), Lars Magne Ingebrigtsen (larsi at ifi.uio.no), Larry Gouge (larry at itginc.com), Karl Heuer (kwzh at gnu.org), (irvine at lks.csi.com), (jaffe at chipmunk.cita.utoronto.ca), David Karr (dkarr at nmo.gtegsc.com), Norbert Kiesel (norbert at i3.informatik.rwth-aachen.de), Steffen Kilb (skilb at gmx.net), Leigh L Klotz (klotz at adoc.xerox.com), Fritz Knabe (Fritz.Knabe at ecrc.de), Heinz Knutzen (hk at informatik.uni-kiel.d400.de), Andrew Koenig (ark at research.att.com), Hannu Koivisto (azure at iki.fi), Ken Laprade (laprade at dw3f.ess.harris.com), Will C Lauer (wcl at cadre.com), Richard Levitte (levitte at e.kth.se), Mike Long (mike.long at analog.com), Dave Love (d.love at dl.ac.uk), Martin Maechler (maechler at stat.math.ethz.ch), Simon Marshall (simon at gnu.org), Paul C. Meuse (pmeuse at delcomsys.com), Richard Mlynarik (mly at adoc.xerox.com), Stefan Monnier (monnier at cs.yale.edu), Chris Murphy (murphycm at sun.aston.ac.uk), Erik Naggum (erik at naggum.no), Eyvind Ness (Eyvind.Ness at hrp.no), Ray Nickson (nickson at cs.uq.oz.au), Dan Nicolaescu (dann at ics.uci.edu), David Petchey (petchey_david at jpmorgan.com), Benjamin Pierce (benjamin.pierce at cl.cam.ac.uk), François Pinard (pinard at iro.umontreal.ca), Tibor Polgar (tlp00 at spg.amdahl.com), David Prince (dave0d at fegs.co.uk), Paul Raines (raines at slac.stanford.edu), Stefan Reicher (xsteve at riic.at), Charles Rich (rich at merl.com), Bill Richter (richter at math.nwu.edu), C.S. Roberson (roberson at aur.alcatel.com), Kevin Rodgers (kevin.rodgers at ihs.com), Sandy Rutherford (sandy at ibm550.sissa.it), Heribert Schuetz (schuetz at ecrc.de), Andy Scott (ascott at pcocd2.intel.com), Axel Seibert (axel at tumbolia.ppp.informatik.uni-muenchen.de), Vin Shelton (acs at xemacs.org), Scott O. Sherman (Scott.Sherman at mci.com), Nikolaj Schumacher (n_schumacher at web.de), Richard Stallman (rms at gnu.org), Richard Stanton (stanton at haas.berkeley.edu), Sam Steingold (sds at goems.com), Ake Stenhoff (etxaksf at aom.ericsson.se), Stig (stig at hackvan.com), Peter Stout (Peter_Stout at cs.cmu.edu), Chuck Thompson (cthomp at cs.uiuc.edu), Ray Tomlinson (tomlinso at bbn.com), Raymond Toy (toy at rtp.ericsson.se), Stephen J. Turnbull (stephen at xemacs.org), Jan Vroonhof (vroonhof at math.ethz.ch), Colin Walters (walters at cis.ohio-state.edu), Philippe Waroquiers (philippe.waroquiers at eurocontrol.be), Klaus Weber (gizmo at zork.north.de), Ben Wing (ben at xemacs.org), Tom Wurgler (twurgler at goodyear.com), Steve Youngs (youngs at xemacs.org), Ilya Zakharevich (ilya at math.ohio-state.edu), Eli Zaretskii (eliz at is.elta.co.il)
Appendix A GNU Free Documentation License
Copyright © 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc. https://fsf.org/ Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.
The purpose of this License is to make a manual, textbook, or other functional and useful document free in the sense of freedom: to assure everyone the effective freedom to copy and redistribute it, with or without modifying it, either commercially or noncommercially. Secondarily, this License preserves for the author and publisher a way to get credit for their work, while not being considered responsible for modifications made by others.
This License is a kind of “copyleft”, which means that derivative works of the document must themselves be free in the same sense. It complements the GNU General Public License, which is a copyleft license designed for free software.
We have designed this License in order to use it for manuals for free software, because free software needs free documentation: a free program should come with manuals providing the same freedoms that the software does. But this License is not limited to software manuals; it can be used for any textual work, regardless of subject matter or whether it is published as a printed book. We recommend this License principally for works whose purpose is instruction or reference.
- APPLICABILITY AND DEFINITIONS
This License applies to any manual or other work, in any medium, that contains a notice placed by the copyright holder saying it can be distributed under the terms of this License. Such a notice grants a world-wide, royalty-free license, unlimited in duration, to use that work under the conditions stated herein. The “Document”, below, refers to any such manual or work. Any member of the public is a licensee, and is addressed as “you”. You accept the license if you copy, modify or distribute the work in a way requiring permission under copyright law.
A “Modified Version” of the Document means any work containing the Document or a portion of it, either copied verbatim, or with modifications and/or translated into another language.
A “Secondary Section” is a named appendix or a front-matter section of the Document that deals exclusively with the relationship of the publishers or authors of the Document to the Document’s overall subject (or to related matters) and contains nothing that could fall directly within that overall subject. (Thus, if the Document is in part a textbook of mathematics, a Secondary Section may not explain any mathematics.) The relationship could be a matter of historical connection with the subject or with related matters, or of legal, commercial, philosophical, ethical or political position regarding them.
The “Invariant Sections” are certain Secondary Sections whose titles are designated, as being those of Invariant Sections, in the notice that says that the Document is released under this License. If a section does not fit the above definition of Secondary then it is not allowed to be designated as Invariant. The Document may contain zero Invariant Sections. If the Document does not identify any Invariant Sections then there are none.
The “Cover Texts” are certain short passages of text that are listed, as Front-Cover Texts or Back-Cover Texts, in the notice that says that the Document is released under this License. A Front-Cover Text may be at most 5 words, and a Back-Cover Text may be at most 25 words.
A “Transparent” copy of the Document means a machine-readable copy, represented in a format whose specification is available to the general public, that is suitable for revising the document straightforwardly with generic text editors or (for images composed of pixels) generic paint programs or (for drawings) some widely available drawing editor, and that is suitable for input to text formatters or for automatic translation to a variety of formats suitable for input to text formatters. A copy made in an otherwise Transparent file format whose markup, or absence of markup, has been arranged to thwart or discourage subsequent modification by readers is not Transparent. An image format is not Transparent if used for any substantial amount of text. A copy that is not “Transparent” is called “Opaque”.
Examples of suitable formats for Transparent copies include plain ASCII without markup, Texinfo input format, LaTeX input format, SGML or XML using a publicly available DTD, and standard-conforming simple HTML, PostScript or PDF designed for human modification. Examples of transparent image formats include PNG, XCF and JPG. Opaque formats include proprietary formats that can be read and edited only by proprietary word processors, SGML or XML for which the DTD and/or processing tools are not generally available, and the machine-generated HTML, PostScript or PDF produced by some word processors for output purposes only.
The “Title Page” means, for a printed book, the title page itself, plus such following pages as are needed to hold, legibly, the material this License requires to appear in the title page. For works in formats which do not have any title page as such, “Title Page” means the text near the most prominent appearance of the work’s title, preceding the beginning of the body of the text.
The “publisher” means any person or entity that distributes copies of the Document to the public.
A section “Entitled XYZ” means a named subunit of the Document whose title either is precisely XYZ or contains XYZ in parentheses following text that translates XYZ in another language. (Here XYZ stands for a specific section name mentioned below, such as “Acknowledgements”, “Dedications”, “Endorsements”, or “History”.) To “Preserve the Title” of such a section when you modify the Document means that it remains a section “Entitled XYZ” according to this definition.
The Document may include Warranty Disclaimers next to the notice which states that this License applies to the Document. These Warranty Disclaimers are considered to be included by reference in this License, but only as regards disclaiming warranties: any other implication that these Warranty Disclaimers may have is void and has no effect on the meaning of this License.
- VERBATIM COPYING
You may copy and distribute the Document in any medium, either commercially or noncommercially, provided that this License, the copyright notices, and the license notice saying this License applies to the Document are reproduced in all copies, and that you add no other conditions whatsoever to those of this License. You may not use technical measures to obstruct or control the reading or further copying of the copies you make or distribute. However, you may accept compensation in exchange for copies. If you distribute a large enough number of copies you must also follow the conditions in section 3.
You may also lend copies, under the same conditions stated above, and you may publicly display copies.
- COPYING IN QUANTITY
If you publish printed copies (or copies in media that commonly have printed covers) of the Document, numbering more than 100, and the Document’s license notice requires Cover Texts, you must enclose the copies in covers that carry, clearly and legibly, all these Cover Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on the back cover. Both covers must also clearly and legibly identify you as the publisher of these copies. The front cover must present the full title with all words of the title equally prominent and visible. You may add other material on the covers in addition. Copying with changes limited to the covers, as long as they preserve the title of the Document and satisfy these conditions, can be treated as verbatim copying in other respects.
If the required texts for either cover are too voluminous to fit legibly, you should put the first ones listed (as many as fit reasonably) on the actual cover, and continue the rest onto adjacent pages.
If you publish or distribute Opaque copies of the Document numbering more than 100, you must either include a machine-readable Transparent copy along with each Opaque copy, or state in or with each Opaque copy a computer-network location from which the general network-using public has access to download using public-standard network protocols a complete Transparent copy of the Document, free of added material. If you use the latter option, you must take reasonably prudent steps, when you begin distribution of Opaque copies in quantity, to ensure that this Transparent copy will remain thus accessible at the stated location until at least one year after the last time you distribute an Opaque copy (directly or through your agents or retailers) of that edition to the public.
It is requested, but not required, that you contact the authors of the Document well before redistributing any large number of copies, to give them a chance to provide you with an updated version of the Document.
You may copy and distribute a Modified Version of the Document under the conditions of sections 2 and 3 above, provided that you release the Modified Version under precisely this License, with the Modified Version filling the role of the Document, thus licensing distribution and modification of the Modified Version to whoever possesses a copy of it. In addition, you must do these things in the Modified Version:
- Use in the Title Page (and on the covers, if any) a title distinct from that of the Document, and from those of previous versions (which should, if there were any, be listed in the History section of the Document). You may use the same title as a previous version if the original publisher of that version gives permission.
- List on the Title Page, as authors, one or more persons or entities responsible for authorship of the modifications in the Modified Version, together with at least five of the principal authors of the Document (all of its principal authors, if it has fewer than five), unless they release you from this requirement.
- State on the Title page the name of the publisher of the Modified Version, as the publisher.
- Preserve all the copyright notices of the Document.
- Add an appropriate copyright notice for your modifications adjacent to the other copyright notices.
- Include, immediately after the copyright notices, a license notice giving the public permission to use the Modified Version under the terms of this License, in the form shown in the Addendum below.
- Preserve in that license notice the full lists of Invariant Sections and required Cover Texts given in the Document’s license notice.
- Include an unaltered copy of this License.
- Preserve the section Entitled “History”, Preserve its Title, and add to it an item stating at least the title, year, new authors, and publisher of the Modified Version as given on the Title Page. If there is no section Entitled “History” in the Document, create one stating the title, year, authors, and publisher of the Document as given on its Title Page, then add an item describing the Modified Version as stated in the previous sentence.
- Preserve the network location, if any, given in the Document for public access to a Transparent copy of the Document, and likewise the network locations given in the Document for previous versions it was based on. These may be placed in the “History” section. You may omit a network location for a work that was published at least four years before the Document itself, or if the original publisher of the version it refers to gives permission.
- For any section Entitled “Acknowledgements” or “Dedications”, Preserve the Title of the section, and preserve in the section all the substance and tone of each of the contributor acknowledgements and/or dedications given therein.
- Preserve all the Invariant Sections of the Document, unaltered in their text and in their titles. Section numbers or the equivalent are not considered part of the section titles.
- Delete any section Entitled “Endorsements”. Such a section may not be included in the Modified Version.
- Do not retitle any existing section to be Entitled “Endorsements” or to conflict in title with any Invariant Section.
- Preserve any Warranty Disclaimers.
If the Modified Version includes new front-matter sections or appendices that qualify as Secondary Sections and contain no material copied from the Document, you may at your option designate some or all of these sections as invariant. To do this, add their titles to the list of Invariant Sections in the Modified Version’s license notice. These titles must be distinct from any other section titles.
You may add a section Entitled “Endorsements”, provided it contains nothing but endorsements of your Modified Version by various parties—for example, statements of peer review or that the text has been approved by an organization as the authoritative definition of a standard.
You may add a passage of up to five words as a Front-Cover Text, and a passage of up to 25 words as a Back-Cover Text, to the end of the list of Cover Texts in the Modified Version. Only one passage of Front-Cover Text and one of Back-Cover Text may be added by (or through arrangements made by) any one entity. If the Document already includes a cover text for the same cover, previously added by you or by arrangement made by the same entity you are acting on behalf of, you may not add another; but you may replace the old one, on explicit permission from the previous publisher that added the old one.
The author(s) and publisher(s) of the Document do not by this License give permission to use their names for publicity for or to assert or imply endorsement of any Modified Version.
- COMBINING DOCUMENTS
You may combine the Document with other documents released under this License, under the terms defined in section 4 above for modified versions, provided that you include in the combination all of the Invariant Sections of all of the original documents, unmodified, and list them all as Invariant Sections of your combined work in its license notice, and that you preserve all their Warranty Disclaimers.
The combined work need only contain one copy of this License, and multiple identical Invariant Sections may be replaced with a single copy. If there are multiple Invariant Sections with the same name but different contents, make the title of each such section unique by adding at the end of it, in parentheses, the name of the original author or publisher of that section if known, or else a unique number. Make the same adjustment to the section titles in the list of Invariant Sections in the license notice of the combined work.
In the combination, you must combine any sections Entitled “History” in the various original documents, forming one section Entitled “History”; likewise combine any sections Entitled “Acknowledgements”, and any sections Entitled “Dedications”. You must delete all sections Entitled “Endorsements.”
- COLLECTIONS OF DOCUMENTS
You may make a collection consisting of the Document and other documents released under this License, and replace the individual copies of this License in the various documents with a single copy that is included in the collection, provided that you follow the rules of this License for verbatim copying of each of the documents in all other respects.
You may extract a single document from such a collection, and distribute it individually under this License, provided you insert a copy of this License into the extracted document, and follow this License in all other respects regarding verbatim copying of that document.
- AGGREGATION WITH INDEPENDENT WORKS
A compilation of the Document or its derivatives with other separate and independent documents or works, in or on a volume of a storage or distribution medium, is called an “aggregate” if the copyright resulting from the compilation is not used to limit the legal rights of the compilation’s users beyond what the individual works permit. When the Document is included in an aggregate, this License does not apply to the other works in the aggregate which are not themselves derivative works of the Document.
If the Cover Text requirement of section 3 is applicable to these copies of the Document, then if the Document is less than one half of the entire aggregate, the Document’s Cover Texts may be placed on covers that bracket the Document within the aggregate, or the electronic equivalent of covers if the Document is in electronic form. Otherwise they must appear on printed covers that bracket the whole aggregate.
Translation is considered a kind of modification, so you may distribute translations of the Document under the terms of section 4. Replacing Invariant Sections with translations requires special permission from their copyright holders, but you may include translations of some or all Invariant Sections in addition to the original versions of these Invariant Sections. You may include a translation of this License, and all the license notices in the Document, and any Warranty Disclaimers, provided that you also include the original English version of this License and the original versions of those notices and disclaimers. In case of a disagreement between the translation and the original version of this License or a notice or disclaimer, the original version will prevail.
If a section in the Document is Entitled “Acknowledgements”, “Dedications”, or “History”, the requirement (section 4) to Preserve its Title (section 1) will typically require changing the actual title.
You may not copy, modify, sublicense, or distribute the Document except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense, or distribute it is void, and will automatically terminate your rights under this License.
However, if you cease all violation of this License, then your license from a particular copyright holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and finally terminates your license, and (b) permanently, if the copyright holder fails to notify you of the violation by some reasonable means prior to 60 days after the cessation.
Moreover, your license from a particular copyright holder is reinstated permanently if the copyright holder notifies you of the violation by some reasonable means, this is the first time you have received notice of violation of this License (for any work) from that copyright holder, and you cure the violation prior to 30 days after your receipt of the notice.
Termination of your rights under this section does not terminate the licenses of parties who have received copies or rights from you under this License. If your rights have been terminated and not permanently reinstated, receipt of a copy of some or all of the same material does not give you any rights to use it.
- FUTURE REVISIONS OF THIS LICENSE
The Free Software Foundation may publish new, revised versions of the GNU Free Documentation License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. See https://www.gnu.org/licenses/.
Each version of the License is given a distinguishing version number. If the Document specifies that a particular numbered version of this License “or any later version” applies to it, you have the option of following the terms and conditions either of that specified version or of any later version that has been published (not as a draft) by the Free Software Foundation. If the Document does not specify a version number of this License, you may choose any version ever published (not as a draft) by the Free Software Foundation. If the Document specifies that a proxy can decide which future versions of this License can be used, that proxy’s public statement of acceptance of a version permanently authorizes you to choose that version for the Document.
“Massive Multiauthor Collaboration Site” (or “MMC Site”) means any World Wide Web server that publishes copyrightable works and also provides prominent facilities for anybody to edit those works. A public wiki that anybody can edit is an example of such a server. A “Massive Multiauthor Collaboration” (or “MMC”) contained in the site means any set of copyrightable works thus published on the MMC site.
“CC-BY-SA” means the Creative Commons Attribution-Share Alike 3.0 license published by Creative Commons Corporation, a not-for-profit corporation with a principal place of business in San Francisco, California, as well as future copyleft versions of that license published by that same organization.
“Incorporate” means to publish or republish a Document, in whole or in part, as part of another Document.
An MMC is “eligible for relicensing” if it is licensed under this License, and if all works that were first published under this License somewhere other than this MMC, and subsequently incorporated in whole or in part into the MMC, (1) had no cover texts or invariant sections, and (2) were thus incorporated prior to November 1, 2008.
The operator of an MMC Site may republish an MMC contained in the site under CC-BY-SA on the same site at any time before August 1, 2009, provided the MMC is eligible for relicensing.
ADDENDUM: How to use this License for your documents
To use this License in a document you have written, include a copy of the License in the document and put the following copyright and license notices just after the title page:
Copyright (C) year your name. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled ``GNU Free Documentation License''.
If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts, replace the “with…Texts.” line with this:
with the Invariant Sections being list their titles, with the Front-Cover Texts being list, and with the Back-Cover Texts being list.
If you have Invariant Sections without Cover Texts, or some other combination of the three, merge those two alternatives to suit the situation.
If your document contains nontrivial examples of program code, we recommend releasing these examples in parallel under your choice of free software license, such as the GNU General Public License, to permit their use in free software.
A B C D E F G H I J M N P Q R S U V W Z
A B C D E F G H I J M N P Q R S U V W Z