"Fossies" - the Fresh Open Source Software Archive

Member "emacs-26.1/doc/emacs/misc.texi" (25 May 2018, 126050 Bytes) of package /linux/misc/emacs-26.1.tar.xz:

Caution: As a special service "Fossies" has tried to format the requested Texinfo source page into HTML format but that may be not always succeeeded perfectly. Alternatively you can here view or download the uninterpreted Texinfo source code. A member file download can also be achieved by clicking within a package contents listing on the according byte size field. See also the latest Fossies "Diffs" side-by-side code changes report for "misc.texi": 25.3_vs_26.1.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

1 Email and Usenet News with Gnus

Gnus is an Emacs package primarily designed for reading and posting Usenet news. It can also be used to read and respond to messages from a number of other sources—email, remote directories, digests, and so on. Here we introduce Gnus and describe several basic features. For full details, see Gnus in The Gnus Manual.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

1.1 Gnus Buffers

Gnus uses several buffers to display information and to receive commands. The three most commonly-used Gnus buffers are the group buffer, the summary buffer and the article buffer.

The group buffer contains a list of article sources (e.g., newsgroups and email inboxes), which are collectively referred to as groups. This is the first buffer Gnus displays when it starts up. It normally displays only the groups to which you subscribe and that contain unread articles. From this buffer, you can select a group to read.

The summary buffer lists the articles in a single group, showing one article per line. By default, it displays each article’s author, subject, and line number, but this is customizable; See Summary Buffer Format in The Gnus Manual. The summary buffer is created when you select a group in the group buffer, and is killed when you exit the group.

From the summary buffer, you can choose an article to view. The article is displayed in the article buffer. In normal Gnus usage, you view this buffer but do not select it—all useful Gnus commands can be invoked from the summary buffer. But you can select the article buffer, and execute Gnus commands from it, if you wish.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

1.2 When Gnus Starts Up

If your system has been set up for reading Usenet news, getting started with Gnus is easy—just type M-x gnus.

On starting up, Gnus reads your news initialization file: a file named ‘.newsrc’ in your home directory which lists your Usenet newsgroups and subscriptions (this file is not unique to Gnus; it is used by many other newsreader programs). It then tries to contact the system’s default news server, which is typically specified by the NNTPSERVER environment variable.

If your system does not have a default news server, or if you wish to use Gnus for reading email, then before invoking M-x gnus you need to tell Gnus where to get news and/or mail. To do this, customize the variables gnus-select-method and/or gnus-secondary-select-methods. See Finding the News in The Gnus Manual.

Once Gnus has started up, it displays the group buffer. By default, the group buffer shows only a small number of subscribed groups. Groups with other statuses—unsubscribed, killed, or zombie—are hidden. The first time you start Gnus, any group to which you are not subscribed is made into a killed group; any group that subsequently appears on the news server becomes a zombie group.

To proceed, you must select a group in the group buffer to open the summary buffer for that group; then, select an article in the summary buffer to view its article buffer in a separate window. The following sections explain how to use the group and summary buffers to do this.

To quit Gnus, type q in the group buffer. This automatically records your group statuses in the files ‘.newsrc’ and ‘.newsrc.eld’, so that they take effect in subsequent Gnus sessions.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

1.3 Using the Gnus Group Buffer

The following commands are available in the Gnus group buffer:


Switch to the summary buffer for the group on the current line (gnus-group-read-group).

A s

In the group buffer, list only the groups to which you subscribe and which contain unread articles (gnus-group-list-groups; this is the default listing).

A u

List all subscribed and unsubscribed groups, but not killed or zombie groups (gnus-group-list-all-groups).

A k

List killed groups (gnus-group-list-killed).

A z

List zombie groups (gnus-group-list-zombies).


Toggle the subscription status of the group (gnus-group-unsubscribe-current-group) on the current line (i.e., turn a subscribed group into an unsubscribed group, or vice versa). Invoking this on a killed or zombie group turns it into an unsubscribed group.


Kill the group on the current line (gnus-group-kill-group). Killed groups are not recorded in the ‘.newsrc’ file, and they are not shown in the l or L listings.


Move point to the previous group containing unread articles (gnus-group-prev-unread-group).


Move point to the next unread group (gnus-group-next-unread-group).


Move point to the previous unread group (gnus-group-prev-unread-group).


Update your Gnus settings, and quit Gnus (gnus-group-exit).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

1.4 Using the Gnus Summary Buffer

The following commands are available in the Gnus summary buffer:


If there is no article selected, select the article on the current line and display its article buffer. Otherwise, try scrolling the selected article buffer in its window; on reaching the end of the buffer, select the next unread article (gnus-summary-next-page).

Thus, you can read through all articles by repeatedly typing <SPC>.


Scroll the text of the article backwards (gnus-summary-prev-page).


Select the next unread article (gnus-summary-next-unread-article).


Select the previous unread article (gnus-summary-prev-unread-article).


Do an incremental search on the selected article buffer (gnus-summary-isearch-article), as if you switched to the buffer and typed C-s (@pxref{Incremental Search}).

M-s regexp <RET>

Search forward for articles containing a match for regexp (gnus-summary-search-article-forward).

M-r regexp <RET>

Search back for articles containing a match for regexp (gnus-summary-search-article-backward).


Exit the summary buffer and return to the group buffer (gnus-summary-exit).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

2 Host Security

Emacs runs inside an operating system such as GNU/Linux, and relies on the operating system to check security constraints such as accesses to files. The default settings for Emacs are designed for typical use; they may require some tailoring in environments where security is more of a concern, or less of a concern, than usual. For example, file-local variables can be risky, and you can set the variable enable-local-variables to :safe or (even more conservatively) to nil; conversely, if your files can all be trusted and the default checking for these variables is irritating, you can set enable-local-variables to :all. @xref{Safe File Variables}.

See Security Considerations in The Emacs Lisp Reference Manual, for more information about security considerations when using Emacs as part of a larger application.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

3 Network Security

Whenever Emacs establishes any network connection, it passes the established connection to the Network Security Manager (NSM). NSM is responsible for enforcing the network security under your control. Currently, this works by using the Transport Layer Security (TLS) features.

The network-security-level variable determines the security level that NSM enforces. If its value is low, no security checks are performed.

If this variable is medium (which is the default), a number of checks will be performed. If as result NSM determines that the network connection might not be trustworthy, it will make you aware of that, and will ask you what to do about the network connection.

You can decide to register a permanent security exception for an unverified connection, a temporary exception, or refuse the connection entirely.

Below is a list of the checks done on the medium level.

unable to verify a TLS certificate

If the connection is a TLS, SSL or STARTTLS connection, NSM will check whether the certificate used to establish the identity of the server we’re connecting to can be verified.

While an invalid certificate is often the cause for concern (there could be a Man-in-the-Middle hijacking your network connection and stealing your password), there may be valid reasons for going ahead with the connection anyway. For instance, the server may be using a self-signed certificate, or the certificate may have expired. It’s up to you to determine whether it’s acceptable to continue with the connection.

a self-signed certificate has changed

If you’ve previously accepted a self-signed certificate, but it has now changed, that could mean that the server has just changed the certificate, but it might also mean that the network connection has been hijacked.

previously encrypted connection now unencrypted

If the connection is unencrypted, but it was encrypted in previous sessions, this might mean that there is a proxy between you and the server that strips away STARTTLS announcements, leaving the connection unencrypted. This is usually very suspicious.

talking to an unencrypted service when sending a password

When connecting to an IMAP or POP3 server, these should usually be encrypted, because it’s common to send passwords over these connections. Similarly, if you’re sending email via SMTP that requires a password, you usually want that connection to be encrypted. If the connection isn’t encrypted, NSM will warn you.

If network-security-level is high, the following checks will be made, in addition to the above:

a validated certificate changes the public key

Servers change their keys occasionally, and that is normally nothing to be concerned about. However, if you are worried that your network connections are being hijacked by agencies who have access to pliable Certificate Authorities which issue new certificates for third-party services, you may want to keep track of these changes.

Diffie-Hellman low prime bits

When doing the public key exchange, the number of prime bits should be high to ensure that the channel can’t be eavesdropped on by third parties. If this number is too low, you will be warned.

RC4 stream cipher

The RC4 stream cipher is believed to be of low quality and may allow eavesdropping by third parties.

SSL1, SSL2 and SSL3

The protocols older than TLS1.0 are believed to be vulnerable to a variety of attacks, and you may want to avoid using these if what you’re doing requires higher security.

Finally, if network-security-level is paranoid, you will also be notified the first time NSM sees any new certificate. This will allow you to inspect all the certificates from all the connections that Emacs makes.

The following additional variables can be used to control details of NSM operation:


This is the file where NSM stores details about connections. It defaults to ‘~/.emacs.d/network-security.data’.


By default, host names will not be saved for non-STARTTLS connections. Instead a host/port hash is used to identify connections. This means that one can’t casually read the settings file to see what servers the user has connected to. If this variable is t, NSM will also save host names in the nsm-settings-file.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4 Document Viewing

DocView mode is a major mode for viewing DVI, PostScript (PS), PDF, OpenDocument, and Microsoft Office documents. It provides features such as slicing, zooming, and searching inside documents. It works by converting the document to a set of images using the gs (GhostScript) or mudraw/pdfdraw (MuPDF) commands and other external tools (1), and displaying those images.

When you visit a document file that can be displayed with DocView mode, Emacs automatically uses that mode (2). As an exception, when you visit a PostScript file, Emacs switches to PS mode, a major mode for editing PostScript files as text; however, it also enables DocView minor mode, so you can type C-c C-c to view the document with DocView. In either DocView mode or DocView minor mode, repeating C-c C-c (doc-view-toggle-display) toggles between DocView and the underlying file contents.

When you visit a file which would normally be handled by DocView mode but some requirement is not met (e.g., you operate in a terminal frame or Emacs has no PNG support), you are queried if you want to view the document’s contents as plain text. If you confirm, the buffer is put in text mode and DocView minor mode is activated. Thus, by typing C-c C-c you switch to the fallback mode. With another C-c C-c you return to DocView mode. The plain text contents can also be displayed from within DocView mode by typing C-c C-t (doc-view-open-text).

You can explicitly enable DocView mode with the command M-x doc-view-mode. You can toggle DocView minor mode with M-x doc-view-minor-mode.

When DocView mode starts, it displays a welcome screen and begins formatting the file, page by page. It displays the first page once that has been formatted.

To kill the DocView buffer, type k (doc-view-kill-proc-and-buffer). To bury it, type q (quit-window).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.1 DocView Navigation

In DocView mode, you can scroll the current page using the usual Emacs movement keys: C-p, C-n, C-b, C-f, and the arrow keys.

By default, the line-motion keys C-p and C-n stop scrolling at the beginning and end of the current page, respectively. However, if you change the variable doc-view-continuous to a non-nil value, then C-p displays the previous page if you are already at the beginning of the current page, and C-n displays the next page if you are at the end of the current page.

You can also display the next page by typing n, <PageDown>, <next> or C-x ] (doc-view-next-page). To display the previous page, type p, <PageUp>, <prior> or C-x [ (doc-view-previous-page).

<SPC> (doc-view-scroll-up-or-next-page) is a convenient way to advance through the document. It scrolls within the current page or advances to the next. <DEL> moves backwards in a similar way (doc-view-scroll-down-or-previous-page).

To go to the first page, type M-< (doc-view-first-page); to go to the last one, type M-> (doc-view-last-page). To jump to a page by its number, type M-g M-g or M-g g (doc-view-goto-page).

You can enlarge or shrink the document with + (doc-view-enlarge) and - (doc-view-shrink). These commands work by reconverting the document at the new size. To specify the default size for DocView, customize the variable doc-view-resolution.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.2 DocView Searching

In DocView mode, you can search the file’s text for a regular expression (@pxref{Regexps}). The interface for searching is inspired by isearch (@pxref{Incremental Search}).

To begin a search, type C-s (doc-view-search) or C-r (doc-view-search-backward). This reads a regular expression using a minibuffer, then echoes the number of matches found within the document. You can move forward and back among the matches by typing C-s and C-r. DocView mode has no way to show the match inside the page image; instead, it displays a tooltip (at the mouse position) listing all matching lines in the current page. To force display of this tooltip, type C-t (doc-view-show-tooltip).

To start a new search, use the search command with a prefix argument; i.e., C-u C-s for a forward search or C-u C-r for a backward search.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.3 DocView Slicing

Documents often have wide margins for printing. They are annoying when reading the document on the screen, because they use up screen space and can cause inconvenient scrolling.

With DocView you can hide these margins by selecting a slice of pages to display. A slice is a rectangle within the page area; once you specify a slice in DocView, it applies to whichever page you look at.

To specify the slice numerically, type s s (doc-view-set-slice); then enter the top left pixel position and the slice’s width and height.

A more convenient graphical way to specify the slice is with s m (doc-view-set-slice-using-mouse), where you use the mouse to select the slice. Simply press and hold the left mouse button at the upper-left corner of the region you want to have in the slice, then move the mouse pointer to the lower-right corner and release the button.

The most convenient way is to set the optimal slice by using BoundingBox information automatically determined from the document by typing s b (doc-view-set-slice-from-bounding-box).

To cancel the selected slice, type s r (doc-view-reset-slice). Then DocView shows the entire page including its entire margins.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

4.4 DocView Conversion

For efficiency, DocView caches the images produced by gs. The name of the directory where it caches images is given by the variable doc-view-cache-directory. You can clear the cache directory by typing M-x doc-view-clear-cache.

To force reconversion of the currently viewed document, type r or g (revert-buffer). To kill the converter process associated with the current buffer, type K (doc-view-kill-proc). The command k (doc-view-kill-proc-and-buffer) kills the converter process and the DocView buffer.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5 Running Shell Commands from Emacs

Emacs has commands for passing single command lines to shell subprocesses, and for running a shell interactively with input and output to an Emacs buffer, and for running a shell in a terminal emulator window.

M-! cmd <RET>

Run the shell command cmd and display the output (shell-command).

M-| cmd <RET>

Run the shell command cmd with region contents as input; optionally replace the region with the output (shell-command-on-region).

M-& cmd <RET>

Run the shell command cmd asynchronously, and display the output (async-shell-command).

M-x shell

Run a subshell with input and output through an Emacs buffer. You can then give commands interactively.

M-x term

Run a subshell with input and output through an Emacs buffer. You can then give commands interactively. Full terminal emulation is available.

Whenever you specify a relative file name for an executable program (either in the cmd argument to one of the above commands, or in other contexts), Emacs searches for the program in the directories specified by the variable exec-path. The value of this variable must be a list of directories; the default value is initialized from the environment variable PATH when Emacs is started (@pxref{General Variables}).

M-x eshell invokes a shell implemented entirely in Emacs. It is documented in its own manual. See Eshell in Eshell: The Emacs Shell.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.1 Single Shell Commands

M-! (shell-command) reads a line of text using the minibuffer and executes it as a shell command, in a subshell made just for that command. Standard input for the command comes from the null device. If the shell command produces any output, the output appears either in the echo area (if it is short), or in an Emacs buffer named ‘*Shell Command Output*’, displayed in another window (if the output is long). The variables resize-mini-windows and max-mini-window-height (@pxref{Minibuffer Edit}) control when Emacs should consider the output to be too long for the echo area.

For instance, one way to decompress a file named ‘foo.gz’ is to type M-! gunzip foo.gz <RET>. That shell command normally creates the file ‘foo’ and produces no terminal output.

A numeric argument to shell-command, e.g., M-1 M-!, causes it to insert terminal output into the current buffer instead of a separate buffer. It puts point before the output, and sets the mark after the output. For instance, M-1 M-! gunzip < foo.gz <RET> would insert the uncompressed form of the file ‘foo.gz’ into the current buffer.

Provided the specified shell command does not end with ‘&’, it runs synchronously, and you must wait for it to exit before continuing to use Emacs. To stop waiting, type C-g to quit; this sends a SIGINT signal to terminate the shell command (this is the same signal that C-c normally generates in the shell). Emacs then waits until the command actually terminates. If the shell command doesn’t stop (because it ignores the SIGINT signal), type C-g again; this sends the command a SIGKILL signal, which is impossible to ignore.

A shell command that ends in ‘&’ is executed asynchronously, and you can continue to use Emacs as it runs. You can also type M-& (async-shell-command) to execute a shell command asynchronously; this is exactly like calling M-! with a trailing ‘&’, except that you do not need the ‘&’. The default output buffer for asynchronous shell commands is named ‘*Async Shell Command*’. Emacs inserts the output into this buffer as it comes in, whether or not the buffer is visible in a window.

If you want to run more than one asynchronous shell command at the same time, they could end up competing for the output buffer. The option async-shell-command-buffer specifies what to do about this; e.g., whether to rename the pre-existing output buffer, or to use a different buffer for the new command. Consult the variable’s documentation for more possibilities.

If you want the output buffer for asynchronous shell commands to be displayed only when the command generates output, set async-shell-command-display-buffer to nil.

M-| (shell-command-on-region) is like M-!, but passes the contents of the region as the standard input to the shell command, instead of no input. With a numeric argument, it deletes the old region and replaces it with the output from the shell command.

For example, you can use M-| with the gpg program to see what keys are in the buffer. If the buffer contains a GnuPG key, type C-x h M-| gpg <RET> to feed the entire buffer contents to gpg. This will output the list of keys to the ‘*Shell Command Output*’ buffer.

The above commands use the shell specified by the variable shell-file-name. Its default value is determined by the SHELL environment variable when Emacs is started. If the file name is relative, Emacs searches the directories listed in exec-path (see section Running Shell Commands from Emacs).

To specify a coding system for M-! or M-|, use the command C-x <RET> c immediately beforehand. @xref{Communication Coding}.

By default, error output is intermixed with the regular output in the output buffer. But if you change the value of the variable shell-command-default-error-buffer to a string, error output is inserted into a buffer of that name.

By default, the output buffer is erased between shell commands. If you change the value of the variable shell-command-dont-erase-buffer to a non-nil value, the output buffer is not erased. This variable also controls where to set the point in the output buffer after the command completes; see the documentation of the variable for details.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.2 Interactive Subshell

To run a subshell interactively, type M-x shell. This creates (or reuses) a buffer named ‘*shell*’, and runs a shell subprocess with input coming from and output going to that buffer. That is to say, any terminal output from the subshell goes into the buffer, advancing point, and any terminal input for the subshell comes from text in the buffer. To give input to the subshell, go to the end of the buffer and type the input, terminated by <RET>.

By default, when the subshell is invoked interactively, the ‘*shell*’ buffer is displayed in a new window, unless the current window already shows the ‘*shell*’ buffer. This behavior can be customized via display-buffer-alist (@pxref{Window Choice}).

While the subshell is waiting or running a command, you can switch windows or buffers and perform other editing in Emacs. Emacs inserts the output from the subshell into the Shell buffer whenever it has time to process it (e.g., while waiting for keyboard input).

In the Shell buffer, prompts are displayed with the face comint-highlight-prompt, and submitted input lines are displayed with the face comint-highlight-input. This makes it easier to distinguish input lines from the shell output. @xref{Faces}.

To make multiple subshells, invoke M-x shell with a prefix argument (e.g., C-u M-x shell). Then the command will read a buffer name, and create (or reuse) a subshell in that buffer. You can also rename the ‘*shell*’ buffer using M-x rename-uniquely, then create a new ‘*shell*’ buffer using plain M-x shell. Subshells in different buffers run independently and in parallel.

To specify the shell file name used by M-x shell, customize the variable explicit-shell-file-name. If this is nil (the default), Emacs uses the environment variable ESHELL if it exists. Otherwise, it usually uses the variable shell-file-name (see section Single Shell Commands); but if the default directory is remote (@pxref{Remote Files}), it prompts you for the shell file name. @xref{Minibuffer File}, for hints how to type remote file names effectively.

Emacs sends the new shell the contents of the file ‘~/.emacs_shellname’ as input, if it exists, where shellname is the name of the file that the shell was loaded from. For example, if you use bash, the file sent to it is ‘~/.emacs_bash’. If this file is not found, Emacs tries with ‘~/.emacs.d/init_shellname.sh’.

To specify a coding system for the shell, you can use the command C-x <RET> c immediately before M-x shell. You can also change the coding system for a running subshell by typing C-x <RET> p in the shell buffer. @xref{Communication Coding}.

Emacs sets the environment variable INSIDE_EMACS in the subshell to ‘version,comint’, where version is the Emacs version (e.g., ‘24.1’). Programs can check this variable to determine whether they are running inside an Emacs subshell.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.3 Shell Mode

The major mode for Shell buffers is Shell mode. Many of its special commands are bound to the C-c prefix, and resemble the usual editing and job control characters present in ordinary shells, except that you must type C-c first. Here is a list of Shell mode commands:


Send the current line as input to the subshell (comint-send-input). Any shell prompt at the beginning of the line is omitted (see section Shell Prompts). If point is at the end of buffer, this is like submitting the command line in an ordinary interactive shell. However, you can also invoke <RET> elsewhere in the shell buffer to submit the current line as input.


Complete the command name or file name before point in the shell buffer (completion-at-point). This uses the usual Emacs completion rules (@pxref{Completion}), with the completion alternatives being file names, environment variable names, the shell command history, and history references (see section Shell History References). For options controlling the completion, see section Shell Mode Options.


Display temporarily a list of the possible completions of the file name before point (comint-dynamic-list-filename-completions).


Either delete a character or send EOF (comint-delchar-or-maybe-eof). Typed at the end of the shell buffer, this sends EOF to the subshell. Typed at any other position in the buffer, this deletes a character as usual.

C-c C-a

Move to the beginning of the line, but after the prompt if any (comint-bol-or-process-mark). If you repeat this command twice in a row, the second time it moves back to the process mark, which is the beginning of the input that you have not yet sent to the subshell. (Normally that is the same place—the end of the prompt on this line—but after C-c <SPC> the process mark may be in a previous line.)

C-c <SPC>

Accumulate multiple lines of input, then send them together (comint-accumulate). This command inserts a newline before point, but does not send the preceding text as input to the subshell—at least, not yet. Both lines, the one before this newline and the one after, will be sent together (along with the newline that separates them), when you type <RET>.

C-c C-u

Kill all text pending at end of buffer to be sent as input (comint-kill-input). If point is not at end of buffer, this only kills the part of this text that precedes point.

C-c C-w

Kill a word before point (backward-kill-word).

C-c C-c

Interrupt the shell or its current subjob if any (comint-interrupt-subjob). This command also kills any shell input pending in the shell buffer and not yet sent.

C-c C-z

Stop the shell or its current subjob if any (comint-stop-subjob). This command also kills any shell input pending in the shell buffer and not yet sent.

C-c C-\

Send quit signal to the shell or its current subjob if any (comint-quit-subjob). This command also kills any shell input pending in the shell buffer and not yet sent.

C-c C-o

Delete the last batch of output from a shell command (comint-delete-output). This is useful if a shell command spews out lots of output that just gets in the way.

C-c C-s

Write the last batch of output from a shell command to a file (comint-write-output). With a prefix argument, the file is appended to instead. Any prompt at the end of the output is not written.

C-c C-r

Scroll to display the beginning of the last batch of output at the top of the window; also move the cursor there (comint-show-output).

C-c C-e

Scroll to put the last line of the buffer at the bottom of the window (comint-show-maximum-output).

C-c C-f

Move forward across one shell command, but not beyond the current line (shell-forward-command). The variable shell-command-regexp specifies how to recognize the end of a command.

C-c C-b

Move backward across one shell command, but not beyond the current line (shell-backward-command).

M-x dirs

Ask the shell for its working directory, and update the Shell buffer’s default directory. See section Directory Tracking.

M-x send-invisible <RET> text <RET>

Send text as input to the shell, after reading it without echoing. This is useful when a shell command runs a program that asks for a password.

Please note that Emacs will not echo passwords by default. If you really want them to be echoed, evaluate (@pxref{Lisp Eval}) the following Lisp expression:

(remove-hook 'comint-output-filter-functions
M-x comint-continue-subjob

Continue the shell process. This is useful if you accidentally suspend the shell process.(3)

M-x comint-strip-ctrl-m

Discard all control-M characters from the current group of shell output. The most convenient way to use this command is to make it run automatically when you get output from the subshell. To do that, evaluate this Lisp expression:

(add-hook 'comint-output-filter-functions
M-x comint-truncate-buffer

This command truncates the shell buffer to a certain maximum number of lines, specified by the variable comint-buffer-maximum-size. Here’s how to do this automatically each time you get output from the subshell:

(add-hook 'comint-output-filter-functions

Shell mode is a derivative of Comint mode, a general-purpose mode for communicating with interactive subprocesses. Most of the features of Shell mode actually come from Comint mode, as you can see from the command names listed above. The special features of Shell mode include the directory tracking feature, and a few user commands.

Other Emacs features that use variants of Comint mode include GUD (@pxref{Debuggers}) and M-x run-lisp (@pxref{External Lisp}).

You can use M-x comint-run to execute any program of your choice in a subprocess using unmodified Comint mode—without the specializations of Shell mode.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.4 Shell Prompts

A prompt is text output by a program to show that it is ready to accept new user input. Normally, Comint mode (and thus Shell mode) automatically figures out which part of the buffer is a prompt, based on the output of the subprocess. (Specifically, it assumes that any received output line which doesn’t end with a newline is a prompt.)

Comint mode divides the buffer into two types of fields: input fields (where user input is typed) and output fields (everywhere else). Prompts are part of the output fields. Most Emacs motion commands do not cross field boundaries, unless they move over multiple lines. For instance, when point is in the input field on a shell command line, C-a puts point at the beginning of the input field, after the prompt. Internally, the fields are implemented using the field text property (see Text Properties in the Emacs Lisp Reference Manual).

If you change the variable comint-use-prompt-regexp to a non-nil value, then Comint mode will recognize prompts using a regular expression (@pxref{Regexps}). In Shell mode, the regular expression is specified by the variable shell-prompt-pattern. The default value of comint-use-prompt-regexp is nil, because this method for recognizing prompts is unreliable, but you may want to set it to a non-nil value in unusual circumstances. In that case, Emacs does not divide the Comint buffer into fields, so the general motion commands behave as they normally do in buffers without special text properties. However, you can use the paragraph motion commands to conveniently navigate the buffer (@pxref{Paragraphs}); in Shell mode, Emacs uses shell-prompt-pattern as paragraph boundaries.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.5 Shell Command History

Shell buffers support three ways of repeating earlier commands. You can use keys like those used for the minibuffer history; these work much as they do in the minibuffer, inserting text from prior commands while point remains always at the end of the buffer. You can move through the buffer to previous inputs in their original place, then resubmit them or copy them to the end. Or you can use a ‘!’-style history reference.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.5.1 Shell History Ring


Fetch the next earlier old shell command (comint-previous-input).


Fetch the next later old shell command (comint-next-input).


Begin an incremental regexp search of old shell commands (comint-history-isearch-backward-regexp).

C-c C-x

Fetch the next subsequent command from the history (comint-get-next-from-history).

C-c .

Fetch one argument from an old shell command (comint-input-previous-argument).

C-c C-l

Display the buffer’s history of shell commands in another window (comint-dynamic-list-input-ring).

Shell buffers provide a history of previously entered shell commands. To reuse shell commands from the history, use the editing commands M-p, M-n, and M-r. These work similar to the minibuffer history commands (@pxref{Minibuffer History}), except that they operate within the Shell buffer rather than the minibuffer, and M-r in a Shell buffer invokes incremental search through shell command history.

M-p fetches an earlier shell command to the end of the shell buffer. Successive use of M-p fetches successively earlier shell commands, each replacing any text that was already present as potential shell input. M-n does likewise except that it finds successively more recent shell commands from the buffer. C-<UP> works like M-p, and C-<DOWN> like M-n.

The history search command M-r begins an incremental regular expression search of previous shell commands. After typing M-r, start typing the desired string or regular expression; the last matching shell command will be displayed in the current line. Incremental search commands have their usual effects—for instance, C-s and C-r search forward and backward for the next match (@pxref{Incremental Search}). When you find the desired input, type <RET> to terminate the search. This puts the input in the command line. Any partial input you were composing before navigating the history list is restored when you go to the beginning or end of the history ring.

Often it is useful to reexecute several successive shell commands that were previously executed in sequence. To do this, first find and reexecute the first command of the sequence. Then type C-c C-x; that will fetch the following command—the one that follows the command you just repeated. Then type <RET> to reexecute this command. You can reexecute several successive commands by typing C-c C-x <RET> over and over.

The command C-c . (comint-input-previous-argument) copies an individual argument from a previous command, like <ESC> . in Bash. The simplest use copies the last argument from the previous shell command. With a prefix argument n, it copies the nth argument instead. Repeating C-c . copies from an earlier shell command instead, always using the same value of n (don’t give a prefix argument when you repeat the C-c . command).

These commands get the text of previous shell commands from a special history list, not from the shell buffer itself. Thus, editing the shell buffer, or even killing large parts of it, does not affect the history that these commands access.

Some shells store their command histories in files so that you can refer to commands from previous shell sessions. Emacs reads the command history file for your chosen shell, to initialize its own command history. The file name is ‘~/.bash_history’ for bash, ‘~/.sh_history’ for ksh, and ‘~/.history’ for other shells.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.5.2 Shell History Copying

C-c C-p

Move point to the previous prompt (comint-previous-prompt).

C-c C-n

Move point to the following prompt (comint-next-prompt).

C-c <RET>

Copy the input command at point, inserting the copy at the end of the buffer (comint-copy-old-input). This is useful if you move point back to a previous command. After you copy the command, you can submit the copy as input with <RET>. If you wish, you can edit the copy before resubmitting it. If you use this command on an output line, it copies that line to the end of the buffer.


If comint-use-prompt-regexp is nil (the default), copy the old input command that you click on, inserting the copy at the end of the buffer (comint-insert-input). If comint-use-prompt-regexp is non-nil, or if the click is not over old input, just yank as usual.

Moving to a previous input and then copying it with C-c <RET> or mouse-2 produces the same results—the same buffer contents—that you would get by using M-p enough times to fetch that previous input from the history list. However, C-c <RET> copies the text from the buffer, which can be different from what is in the history list if you edit the input text in the buffer after it has been sent.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.5.3 Shell History References

Various shells, including csh and bash, support history references that begin with ‘!’ and ‘^’. Shell mode recognizes these constructs, and can perform the history substitution for you.

If you insert a history reference and type <TAB>, this searches the input history for a matching command, performs substitution if necessary, and places the result in the buffer in place of the history reference. For example, you can fetch the most recent command beginning with ‘mv’ with ! m v <TAB>. You can edit the command if you wish, and then resubmit the command to the shell by typing <RET>.

Shell mode can optionally expand history references in the buffer when you send them to the shell. To request this, set the variable comint-input-autoexpand to input. You can make <SPC> perform history expansion by binding <SPC> to the command comint-magic-space. @xref{Rebinding}.

Shell mode recognizes history references when they follow a prompt. See section Shell Prompts, for how Shell mode recognizes prompts.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.6 Directory Tracking

Shell mode keeps track of ‘cd’, ‘pushd’ and ‘popd’ commands given to the subshell, in order to keep the Shell buffer’s default directory (@pxref{File Names}) the same as the shell’s working directory. It recognizes these commands by examining lines of input that you send.

If you use aliases for these commands, you can tell Emacs to recognize them also, by setting the variables shell-pushd-regexp, shell-popd-regexp, and shell-cd-regexp to the appropriate regular expressions (@pxref{Regexps}). For example, if shell-pushd-regexp matches the beginning of a shell command line, that line is regarded as a pushd command. These commands are recognized only at the beginning of a shell command line.

If Emacs gets confused about changes in the working directory of the subshell, type M-x dirs. This command asks the shell for its working directory and updates the default directory accordingly. It works for shells that support the most common command syntax, but may not work for unusual shells.

You can also use Dirtrack mode, a buffer-local minor mode that implements an alternative method of tracking the shell’s working directory. To use this method, your shell prompt must contain the working directory at all times, and you must supply a regular expression for recognizing which part of the prompt contains the working directory; see the documentation of the variable dirtrack-list for details. To use Dirtrack mode, type M-x dirtrack-mode in the Shell buffer, or add dirtrack-mode to shell-mode-hook (@pxref{Hooks}).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.7 Shell Mode Options

If the variable comint-scroll-to-bottom-on-input is non-nil, insertion and yank commands scroll the selected window to the bottom before inserting. The default is nil.

If comint-scroll-show-maximum-output is non-nil, then arrival of output when point is at the end tries to scroll the last line of text to the bottom line of the window, showing as much useful text as possible. (This mimics the scrolling behavior of most terminals.) The default is t.

By setting comint-move-point-for-output, you can opt for having point jump to the end of the buffer whenever output arrives—no matter where in the buffer point was before. If the value is this, point jumps in the selected window. If the value is all, point jumps in each window that shows the Comint buffer. If the value is other, point jumps in all nonselected windows that show the current buffer. The default value is nil, which means point does not jump to the end.

If you set comint-prompt-read-only, the prompts in the Comint buffer are read-only.

The variable comint-input-ignoredups controls whether successive identical inputs are stored in the input history. A non-nil value means to omit an input that is the same as the previous input. The default is nil, which means to store each input even if it is equal to the previous input.

Three variables customize file name completion. The variable comint-completion-addsuffix controls whether completion inserts a space or a slash to indicate a fully completed file or directory name (non-nil means do insert a space or slash). comint-completion-recexact, if non-nil, directs <TAB> to choose the shortest possible completion if the usual Emacs completion algorithm cannot add even a single character. comint-completion-autolist, if non-nil, says to list all the possible completions whenever completion is not exact.

Command completion normally considers only executable files. If you set shell-completion-execonly to nil, it considers nonexecutable files as well.

The variable shell-completion-fignore specifies a list of file name extensions to ignore in Shell mode completion. The default setting is nil, but some users prefer ("~" "#" "%") to ignore file names ending in ‘~’, ‘#’ or ‘%’. Other related Comint modes use the variable comint-completion-fignore instead.

Some implementation details of the shell command completion may also be found in the lisp documentation of the shell-dynamic-complete-command function.

You can configure the behavior of ‘pushd’. Variables control whether ‘pushd’ behaves like ‘cd’ if no argument is given (shell-pushd-tohome), pop rather than rotate with a numeric argument (shell-pushd-dextract), and only add directories to the directory stack if they are not already on it (shell-pushd-dunique). The values you choose should match the underlying shell, of course.

Comint mode sets the TERM environment variable to a safe default value, but this value disables some useful features. For example, color is disabled in applications that use TERM to determine if color is supported. Therefore, Emacs provides an option comint-terminfo-terminal, which you can set to a terminal that is present in your system’s terminfo database, in order to take advantage of advanced features of that terminal.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.8 Emacs Terminal Emulator

To run a subshell in a text terminal emulator, use M-x term. This creates (or reuses) a buffer named ‘*terminal*’, and runs a subshell with input coming from your keyboard, and output going to that buffer.

The terminal emulator uses Term mode, which has two input modes. In line mode, Term basically acts like Shell mode (see section Shell Mode). In char mode, each character is sent directly to the subshell, as terminal input; the sole exception is the terminal escape character, which by default is C-c (see section Term Mode). Any echoing of your input is the responsibility of the subshell; any terminal output from the subshell goes into the buffer, advancing point.

Some programs (such as Emacs itself) need to control the appearance of the terminal screen in detail. They do this by emitting special control codes. Term mode recognizes and handles ANSI-standard VT100-style escape sequences, which are accepted by most modern terminals, including xterm. (Hence, you can actually run Emacs inside an Emacs Term window.)

The term face specifies the default appearance of text in the terminal emulator (the default is the same appearance as the default face). When terminal control codes are used to change the appearance of text, these are represented in the terminal emulator by the faces term-color-black, term-color-red, term-color-green, term-color-yellow term-color-blue, term-color-magenta, term-color-cyan, term-color-white, term-color-underline, and term-color-bold. @xref{Faces}.

You can also use Term mode to communicate with a device connected to a serial port. See section Serial Terminal.

The file name used to load the subshell is determined the same way as for Shell mode. To make multiple terminal emulators, rename the buffer ‘*terminal*’ to something different using M-x rename-uniquely, just as with Shell mode.

Unlike Shell mode, Term mode does not track the current directory by examining your input. But some shells can tell Term what the current directory is. This is done automatically by bash version 1.15 and later.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.9 Term Mode

To switch between line and char mode in Term mode, use these commands:

C-c C-j

Switch to line mode (term-line-mode). Do nothing if already in line mode.

C-c C-k

Switch to char mode (term-char-mode). Do nothing if already in char mode.

The following commands are only available in char mode:

C-c C-c

Send a literal C-c to the sub-shell (term-interrupt-subjob).

C-c char

This is equivalent to C-x char in normal Emacs. For example, C-c o invokes the global binding of C-x o, which is normally ‘other-window’.

Term mode has a page-at-a-time feature. When enabled, it makes output pause at the end of each screenful:

C-c C-q

Toggle the page-at-a-time feature (term-pager-toggle). This command works in both line and char modes. When the feature is enabled, the mode-line displays the word ‘page’, and each time Term receives more than a screenful of output, it pauses and displays ‘**MORE**’ in the mode-line. Type <SPC> to display the next screenful of output, or ? to see your other options. The interface is similar to the more program.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.10 Remote Host Shell

You can login to a remote computer, using whatever commands you would from a regular terminal (e.g., using the ssh or telnet or rlogin commands), from a Term window.

A program that asks you for a password will normally suppress echoing of the password, so the password will not show up in the buffer. This will happen just as if you were using a real terminal, if the buffer is in char mode. If it is in line mode, the password is temporarily visible, but will be erased when you hit return. (This happens automatically; there is no special password processing.)

When you log in to a different machine, you need to specify the type of terminal you’re using, by setting the TERM environment variable in the environment for the remote login command. (If you use bash, you do that by writing the variable assignment before the remote login command, without a separating comma.) Terminal types ‘ansi’ or ‘vt100’ will work on most systems.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

5.11 Serial Terminal

If you have a device connected to a serial port of your computer, you can communicate with it by typing M-x serial-term. This command asks for a serial port name and speed, and switches to a new Term mode buffer. Emacs communicates with the serial device through this buffer just like it does with a terminal in ordinary Term mode.

The speed of the serial port is measured in bits per second. The most common speed is 9600 bits per second. You can change the speed interactively by clicking on the mode line.

A serial port can be configured even more by clicking on ‘8N1’ in the mode line. By default, a serial port is configured as ‘8N1’, which means that each byte consists of 8 data bits, No parity check bit, and 1 stopbit.

If the speed or the configuration is wrong, you cannot communicate with your device and will probably only see garbage output in the window.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

6 Using Emacs as a Server

Various programs can invoke your choice of editor to edit a particular piece of text. For instance, version control programs invoke an editor to enter version control logs (@pxref{Version Control}), and the Unix mail utility invokes an editor to enter a message to send. By convention, your choice of editor is specified by the environment variable EDITOR. If you set EDITOR to ‘emacs’, Emacs would be invoked, but in an inconvenient way—by starting a new Emacs process. This is inconvenient because the new Emacs process doesn’t share buffers, a command history, or other kinds of information with any existing Emacs process.

You can solve this problem by setting up Emacs as an edit server, so that it “listens” for external edit requests and acts accordingly. There are various ways to start an Emacs server:

Once an Emacs server is started, you can use a shell command called emacsclient to connect to the Emacs process and tell it to visit a file. You can then set the EDITOR environment variable to ‘emacsclient’, so that external programs will use the existing Emacs process for editing.(4)

You can run multiple Emacs servers on the same machine by giving each one a unique server name, using the variable server-name. For example, M-x set-variable <RET> server-name <RET> "foo" <RET> sets the server name to ‘foo’. The emacsclient program can specify a server by name, using the ‘-s’ option (see section emacsclient Options).

If you want to run multiple Emacs daemons (@pxref{Initial Options}), you can give each daemon its own server name like this:

  emacs --daemon=foo

If you have defined a server by a unique server name, it is possible to connect to the server from another Emacs instance and evaluate Lisp expressions on the server, using the server-eval-at function. For instance, (server-eval-at "foo" '(+ 1 2)) evaluates the expression (+ 1 2) on the ‘foo’ server, and returns 3. (If there is no server with that name, an error is signaled.) Currently, this feature is mainly useful for developers.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

6.1 TCP Emacs server

An Emacs server usually listens to connections on a local Unix domain socket. Some operating systems, such as MS-Windows, do not support local sockets; in that case, the server uses TCP sockets instead. In some cases it is useful to have the server listen on a TCP socket even if local sockets are supported, e.g., if you need to contact the Emacs server from a remote machine. You can set server-use-tcp to non-nil to have Emacs listen on a TCP socket instead of a local socket. This is the default if your OS does not support local sockets.

If the Emacs server is set to use TCP, it will by default listen on a random port on the localhost interface. This can be changed to another interface and/or a fixed port using the variables server-host and server-port.

A TCP socket is not subject to file system permissions. To retain some control over which users can talk to an Emacs server over TCP sockets, the emacsclient program must send an authorization key to the server. This key is normally randomly generated by the Emacs server. This is the recommended mode of operation.

If needed, you can set the authorization key to a static value by setting the server-auth-key variable. The key must consist of 64 ASCII printable characters except for space (this means characters from ‘!’ to ‘~’, or from decimal code 33 to 126). You can use M-x server-generate-key to get a random key.

When you start a TCP Emacs server, Emacs creates a server file containing the TCP information to be used by emacsclient to connect to the server. The variable server-auth-dir specifies the directory containing the server file; by default, this is ‘~/.emacs.d/server/’. In the absence of a local socket with file permissions, the permissions of this directory determine which users can have their emacsclient processes talk to the Emacs server.

To tell emacsclient to connect to the server over TCP with a specific server file, use the ‘-f’ or ‘--server-file’ option, or set the EMACS_SERVER_FILE environment variable (see section emacsclient Options). If server-auth-dir is set to a non-standard value, emacsclient needs an absolute file name to the server file, as the default server-auth-dir is hard-coded in emacsclient to be used as the directory for resolving relative filenames.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

6.2 Invoking emacsclient

The simplest way to use the emacsclient program is to run the shell command ‘emacsclient file’, where file is a file name. This connects to an Emacs server, and tells that Emacs process to visit file in one of its existing frames—either a graphical frame, or one in a text terminal (@pxref{Frames}). You can then select that frame to begin editing.

If there is no Emacs server, the emacsclient program halts with an error message (you can prevent this from happening by using the ‘--alternate-editor=""’ option to emacsclient, see section emacsclient Options). If the Emacs process has no existing frame—which can happen if it was started as a daemon (see section Using Emacs as a Server)—then Emacs opens a frame on the terminal in which you called emacsclient.

You can also force emacsclient to open a new frame on a graphical display using the ‘-c’ option, or on a text terminal using the ‘-t’ option. See section emacsclient Options.

If you are running on a single text terminal, you can switch between emacsclient’s shell and the Emacs server using one of two methods: (i) run the Emacs server and emacsclient on different virtual terminals, and switch to the Emacs server’s virtual terminal after calling emacsclient; or (ii) call emacsclient from within the Emacs server itself, using Shell mode (see section Interactive Subshell) or Term mode (see section Term Mode); emacsclient blocks only the subshell under Emacs, and you can still use Emacs to edit the file.

When you finish editing file in the Emacs server, type C-x # (server-edit) in its buffer. This saves the file and sends a message back to the emacsclient program, telling it to exit. Programs that use EDITOR usually wait for the editor—in this case emacsclient—to exit before doing something else.

You can also call emacsclient with multiple file name arguments: ‘emacsclient file1 file2 ...’ tells the Emacs server to visit file1, file2, and so forth. Emacs selects the buffer visiting file1, and buries the other buffers at the bottom of the buffer list (@pxref{Buffers}). The emacsclient program exits once all the specified files are finished (i.e., once you have typed C-x # in each server buffer).

Finishing with a server buffer also kills the buffer, unless it already existed in the Emacs session before the server was asked to create it. However, if you set server-kill-new-buffers to nil, then a different criterion is used: finishing with a server buffer kills it if the file name matches the regular expression server-temp-file-regexp. This is set up to distinguish certain temporary files.

Each C-x # checks for other pending external requests to edit various files, and selects the next such file. You can switch to a server buffer manually if you wish; you don’t have to arrive at it with C-x #. But C-x # is the way to tell emacsclient that you are finished.

If you set the value of the variable server-window to a window or a frame, C-x # always displays the next server buffer in that window or in that frame.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

6.3 emacsclient Options

You can pass some optional arguments to the emacsclient program, such as:

emacsclient -c +12 file1 +4:3 file2

The ‘+line’ or ‘+line:column’ arguments specify line numbers, or line and column numbers, for the next file argument. These behave like the command line arguments for Emacs itself. @xref{Action Arguments}.

The other optional arguments recognized by emacsclient are listed below:

-a command

Specify a shell command to run if emacsclient fails to contact Emacs. This is useful when running emacsclient in a script. The command may include arguments, which may be quoted "like this". Currently, escaping of quotes is not supported.

As a special exception, if command is the empty string, then emacsclient starts Emacs in daemon mode (as ‘emacs --daemon’) and then tries connecting again.

The environment variable ALTERNATE_EDITOR has the same effect as the ‘-a’ option. If both are present, the latter takes precedence.


Create a new graphical client frame, instead of using an existing Emacs frame. See below for the special behavior of C-x C-c in a client frame. If Emacs cannot create a new graphical frame (e.g., if it cannot connect to the X server), it tries to create a text terminal client frame, as though you had supplied the ‘-t’ option instead.

On MS-Windows, a single Emacs session cannot display frames on both graphical and text terminals, nor on multiple text terminals. Thus, if the Emacs server is running on a text terminal, the ‘-c’ option, like the ‘-t’ option, creates a new frame in the server’s current text terminal. @xref{Windows Startup}.

If you omit a filename argument while supplying the ‘-c’ option, the new frame displays the ‘*scratch*’ buffer by default. You can customize this behavior with the variable initial-buffer-choice (@pxref{Entering Emacs}).

-F alist

Set the parameters for a newly-created graphical frame (@pxref{Frame Parameters}).

-d display

Tell Emacs to open the given files on the X display display (assuming there is more than one X display available).


Tell Emacs to evaluate some Emacs Lisp code, instead of visiting some files. When this option is given, the arguments to emacsclient are interpreted as a list of expressions to evaluate, not as a list of files to visit.

-f server-file

Specify a server file (see section TCP Emacs server) for connecting to an Emacs server via TCP. Alternatively, you can set the EMACS_SERVER_FILE environment variable to point to the server file. (The command-line option overrides the environment variable.)

An Emacs server usually uses a local socket to listen for connections, but also supports connections over TCP. To connect to a TCP Emacs server, emacsclient needs to read a server file containing the connection details of the Emacs server. The name of this file is specified with this option, either as a file name relative to ‘~/.emacs.d/server’ or as an absolute file name. See section TCP Emacs server.


Let emacsclient exit immediately, instead of waiting until all server buffers are finished. You can take as long as you like to edit the server buffers within Emacs, and they are not killed when you type C-x # in them.

--parent-id id

Open an emacsclient frame as a client frame in the parent X window with id id, via the XEmbed protocol. Currently, this option is mainly useful for developers.


Do not let emacsclient display messages about waiting for Emacs or connecting to remote server sockets.


Do not let emacsclient display results returned from the server. Mostly useful in combination with ‘-e’ when the evaluation performed is for side-effect rather than result.

-s server-name

Connect to the Emacs server named server-name. The server name is given by the variable server-name on the Emacs server. If this option is omitted, emacsclient connects to the first server it finds. (This option is not supported on MS-Windows.)


Create a new client frame on the current text terminal, instead of using an existing Emacs frame. This behaves just like the ‘-c’ option, described above, except that it creates a text terminal frame (@pxref{Non-Window Terminals}).

On MS-Windows, ‘-t’ behaves just like ‘-c’ if the Emacs server is using the graphical display, but if the Emacs server is running on a text terminal, it creates a new frame in the current text terminal.

-T tramp-prefix

Set the prefix to add to filenames for Emacs to locate files on remote machines (@pxref{Remote Files}) using TRAMP (see The Tramp Manual in The Tramp Manual). This is mostly useful in combination with using the Emacs server over TCP (see section TCP Emacs server). By ssh-forwarding the listening port and making the server-file available on a remote machine, programs on the remote machine can use emacsclient as the value for the EDITOR and similar environment variables, but instead of talking to an Emacs server on the remote machine, the files will be visited in the local Emacs session using TRAMP.

Setting the environment variable EMACSCLIENT_TRAMP has the same effect as using the ‘-T’ option. If both are specified, the command-line option takes precedence.

For example, assume two hosts, ‘local’ and ‘remote’, and that the local Emacs listens on tcp port 12345. Assume further that ‘/home’ is on a shared file system, so that the server file ‘~/.emacs.d/server/server’ is readable on both hosts.

local$ ssh -R12345:localhost:12345 remote
remote$ export EDITOR="emacsclient \
        --server-file=server \
remote$ $EDITOR /tmp/foo.txt #Should open in local emacs.

The new graphical or text terminal frames created by the ‘-c’ or ‘-t’ options are considered client frames. Any new frame that you create from a client frame is also considered a client frame. If you type C-x C-c (save-buffers-kill-terminal) in a client frame, that command does not kill the Emacs session as it normally does (@pxref{Exiting}). Instead, Emacs deletes the client frame; furthermore, if the client frame has an emacsclient waiting to regain control (i.e., if you did not supply the ‘-n’ option), Emacs deletes all other frames of the same client, and marks the client’s server buffers as finished, as though you had typed C-x # in all of them. If it so happens that there are no remaining frames after the client frame(s) are deleted, the Emacs session exits.

As an exception, when Emacs is started as a daemon, all frames are considered client frames, and C-x C-c never kills Emacs. To kill a daemon session, type M-x kill-emacs.

Note that the ‘-t’ and ‘-n’ options are contradictory: ‘-t’ says to take control of the current text terminal to create a new client frame, while ‘-n’ says not to take control of the text terminal. If you supply both options, Emacs visits the specified files(s) in an existing frame rather than a new client frame, negating the effect of ‘-t’.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

7 Printing Hard Copies

Emacs provides commands for printing hardcopies of either an entire buffer or part of one. You can invoke the printing commands directly, as detailed below, or using the ‘File’ menu on the menu bar.

Aside from the commands described in this section, you can also print hardcopies from Dired (@pxref{Operating on Files}) and the diary (@pxref{Displaying the Diary}). You can also “print” an Emacs buffer to HTML with the command M-x htmlfontify-buffer, which converts the current buffer to a HTML file, replacing Emacs faces with CSS-based markup. Furthermore, Org mode allows you to print Org files to a variety of formats, such as PDF (@pxref{Org Mode}).

M-x print-buffer

Print hardcopy of current buffer with page headings containing the file name and page number.

M-x lpr-buffer

Print hardcopy of current buffer without page headings.

M-x print-region

Like print-buffer but print only the current region.

M-x lpr-region

Like lpr-buffer but print only the current region.

On most operating systems, the above hardcopy commands submit files for printing by calling the lpr program. To change the printer program, customize the variable lpr-command. To specify extra switches to give the printer program, customize the list variable lpr-switches. Its value should be a list of option strings, each of which should start with ‘-’ (e.g., the option string "-w80" specifies a line width of 80 columns). The default is the empty list, nil.

To specify the printer to use, set the variable printer-name. The default, nil, specifies the default printer. If you set it to a printer name (a string), that name is passed to lpr with the ‘-P’ switch; if you are not using lpr, you should specify the switch with lpr-printer-switch.

The variable lpr-headers-switches similarly specifies the extra switches to use to make page headers. The variable lpr-add-switches controls whether to supply ‘-T’ and ‘-J’ options (suitable for lpr) to the printer program: nil means don’t add them (this should be the value if your printer program is not compatible with lpr).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

7.1 PostScript Hardcopy

These commands convert buffer contents to PostScript, either printing it or leaving it in another Emacs buffer.

M-x ps-print-buffer

Print hardcopy of the current buffer in PostScript form.

M-x ps-print-region

Print hardcopy of the current region in PostScript form.

M-x ps-print-buffer-with-faces

Print hardcopy of the current buffer in PostScript form, showing the faces used in the text by means of PostScript features.

M-x ps-print-region-with-faces

Print hardcopy of the current region in PostScript form, showing the faces used in the text.

M-x ps-spool-buffer

Generate and spool a PostScript image for the current buffer text.

M-x ps-spool-region

Generate and spool a PostScript image for the current region.

M-x ps-spool-buffer-with-faces

Generate and spool a PostScript image for the current buffer, showing the faces used.

M-x ps-spool-region-with-faces

Generate and spool a PostScript image for the current region, showing the faces used.

M-x ps-despool

Send the spooled PostScript to the printer.

M-x handwrite

Generate/print PostScript for the current buffer as if handwritten.

The ps-print-buffer and ps-print-region commands print buffer contents in PostScript form. One command prints the entire buffer; the other, just the region. The commands ps-print-buffer-with-faces and ps-print-region-with-faces behave similarly, but use PostScript features to show the faces (fonts and colors) of the buffer text.

Interactively, when you use a prefix argument (C-u), these commands prompt the user for a file name, and save the PostScript image in that file instead of sending it to the printer.

The commands whose names have ‘spool’ instead of ‘print’, generate the PostScript output in an Emacs buffer instead of sending it to the printer.

Use the command ps-despool to send the spooled images to the printer. This command sends the PostScript generated by ‘-spool-’ commands (see commands above) to the printer. With a prefix argument (C-u), it prompts for a file name, and saves the spooled PostScript image in that file instead of sending it to the printer.

M-x handwrite is more frivolous. It generates a PostScript rendition of the current buffer as a cursive handwritten document. It can be customized in group handwrite. This function only supports ISO 8859-1 characters.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

7.2 Variables for PostScript Hardcopy

All the PostScript hardcopy commands use the variables ps-lpr-command and ps-lpr-switches to specify how to print the output. ps-lpr-command specifies the command name to run, ps-lpr-switches specifies command line options to use, and ps-printer-name specifies the printer. If you don’t set the first two variables yourself, they take their initial values from lpr-command and lpr-switches. If ps-printer-name is nil, printer-name is used.

The variable ps-print-header controls whether these commands add header lines to each page—set it to nil to turn headers off.

If your printer doesn’t support colors, you should turn off color processing by setting ps-print-color-p to nil. By default, if the display supports colors, Emacs produces hardcopy output with color information; on black-and-white printers, colors are emulated with shades of gray. This might produce illegible output, even if your screen colors only use shades of gray.

Alternatively, you can set ps-print-color-p to black-white to print colors on black/white printers.

By default, PostScript printing ignores the background colors of the faces, unless the variable ps-use-face-background is non-nil. This is to avoid unwanted interference with the zebra stripes and background image/text.

The variable ps-paper-type specifies which size of paper to format for; legitimate values include a4, a3, a4small, b4, b5, executive, ledger, legal, letter, letter-small, statement, tabloid. The default is letter. You can define additional paper sizes by changing the variable ps-page-dimensions-database.

The variable ps-landscape-mode specifies the orientation of printing on the page. The default is nil, which stands for portrait mode. Any non-nil value specifies landscape mode.

The variable ps-number-of-columns specifies the number of columns; it takes effect in both landscape and portrait mode. The default is 1.

The variable ps-font-family specifies which font family to use for printing ordinary text. Legitimate values include Courier, Helvetica, NewCenturySchlbk, Palatino and Times. The variable ps-font-size specifies the size of the font for ordinary text and defaults to 8.5 points. The value of ps-font-size can also be a cons of 2 floats: one for landscape mode, the other for portrait mode.

Emacs supports more scripts and characters than a typical PostScript printer. Thus, some of the characters in your buffer might not be printable using the fonts built into your printer. You can augment the fonts supplied with the printer with those from the GNU Intlfonts package, or you can instruct Emacs to use Intlfonts exclusively. The variable ps-multibyte-buffer controls this: the default value, nil, is appropriate for printing ASCII and Latin-1 characters; a value of non-latin-printer is for printers which have the fonts for ASCII, Latin-1, Japanese, and Korean characters built into them. A value of bdf-font arranges for the BDF fonts from the Intlfonts package to be used for all characters. Finally, a value of bdf-font-except-latin instructs the printer to use built-in fonts for ASCII and Latin-1 characters, and Intlfonts BDF fonts for the rest.

To be able to use the BDF fonts, Emacs needs to know where to find them. The variable bdf-directory-list holds the list of directories where Emacs should look for the fonts; the default value includes a single directory ‘/usr/local/share/emacs/fonts/bdf’.

Many other customization variables for these commands are defined and described in the Lisp files ‘ps-print.el’ and ‘ps-mule.el’.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

7.3 Printing Package

The basic Emacs facilities for printing hardcopy can be extended using the Printing package. This provides an easy-to-use interface for choosing what to print, previewing PostScript files before printing, and setting various printing options such as print headers, landscape or portrait modes, duplex modes, and so forth. On GNU/Linux or Unix systems, the Printing package relies on the ‘gs’ and ‘gv’ utilities, which are distributed as part of the GhostScript program. On MS-Windows, the ‘gstools’ port of Ghostscript can be used.

To use the Printing package, add (require 'printing) to your init file (@pxref{Init File}), followed by (pr-update-menus). This function replaces the usual printing commands in the menu bar with a ‘Printing’ submenu that contains various printing options. You can also type M-x pr-interface <RET>; this creates a ‘*Printing Interface*’ buffer, similar to a customization buffer, where you can set the printing options. After selecting what and how to print, you start the print job using the ‘Print’ button (click mouse-2 on it, or move point over it and type <RET>). For further information on the various options, use the ‘Interface Help’ button.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

8 Sorting Text

Emacs provides several commands for sorting text in the buffer. All operate on the contents of the region. They divide the text of the region into many sort records, identify a sort key for each record, and then reorder the records into the order determined by the sort keys. The records are ordered so that their keys are in alphabetical order, or, for numeric sorting, in numeric order. In alphabetic sorting, all upper-case letters ‘A’ through ‘Z’ come before lower-case ‘a’, in accordance with the ASCII character sequence (but sort-fold-case, described below, can change that).

The various sort commands differ in how they divide the text into sort records and in which part of each record is used as the sort key. Most of the commands make each line a separate sort record, but some commands use paragraphs or pages as sort records. Most of the sort commands use each entire sort record as its own sort key, but some use only a portion of the record as the sort key.

M-x sort-lines

Divide the region into lines, and sort by comparing the entire text of a line. A numeric argument means sort into descending order.

M-x sort-paragraphs

Divide the region into paragraphs, and sort by comparing the entire text of a paragraph (except for leading blank lines). A numeric argument means sort into descending order.

M-x sort-pages

Divide the region into pages, and sort by comparing the entire text of a page (except for leading blank lines). A numeric argument means sort into descending order.

M-x sort-fields

Divide the region into lines, and sort by comparing the contents of one field in each line. Fields are defined as separated by whitespace, so the first run of consecutive non-whitespace characters in a line constitutes field 1, the second such run constitutes field 2, etc.

Specify which field to sort by with a numeric argument: 1 to sort by field 1, etc.; the default is 1. A negative argument means count fields from the right instead of from the left; thus, minus 1 means sort by the last field. If several lines have identical contents in the field being sorted, they keep the same relative order that they had in the original buffer.

M-x sort-numeric-fields

Like M-x sort-fields except the specified field is converted to an integer for each line, and the numbers are compared. ‘10’ comes before ‘2’ when considered as text, but after it when considered as a number. By default, numbers are interpreted according to sort-numeric-base, but numbers beginning with ‘0x’ or ‘0’ are interpreted as hexadecimal and octal, respectively.

M-x sort-columns

Like M-x sort-fields except that the text within each line used for comparison comes from a fixed range of columns. With a prefix argument, sort in reverse order. See below for more details on this command.

M-x reverse-region

Reverse the order of the lines in the region. This is useful for sorting into descending order by fields, since those sort commands do not have a feature for doing that.

For example, if the buffer contains this:

On systems where clash detection (locking of files being edited) is
implemented, Emacs also checks the first time you modify a buffer
whether the file has changed on disk since it was last visited or
saved.  If it has, you are asked to confirm that you want to change
the buffer.

applying M-x sort-lines to the entire buffer produces this:

On systems where clash detection (locking of files being edited) is
implemented, Emacs also checks the first time you modify a buffer
saved.  If it has, you are asked to confirm that you want to change
the buffer.
whether the file has changed on disk since it was last visited or

where the upper-case ‘O’ sorts before all lower-case letters. If you use C-u 2 M-x sort-fields instead, you get this:

implemented, Emacs also checks the first time you modify a buffer
saved.  If it has, you are asked to confirm that you want to change
the buffer.
On systems where clash detection (locking of files being edited) is
whether the file has changed on disk since it was last visited or

where the sort keys were ‘Emacs’, ‘If’, ‘buffer’, ‘systems’ and ‘the’.

M-x sort-columns requires more explanation. You specify the columns by putting point at one of the columns and the mark at the other column. Because this means you cannot put point or the mark at the beginning of the first line of the text you want to sort, this command uses an unusual definition of “region”: all of the line point is in is considered part of the region, and so is all of the line the mark is in, as well as all the lines in between.

For example, to sort a table by information found in columns 10 to 15, you could put the mark on column 10 in the first line of the table, and point on column 15 in the last line of the table, and then run sort-columns. Equivalently, you could run it with the mark on column 15 in the first line and point on column 10 in the last line.

This can be thought of as sorting the rectangle specified by point and the mark, except that the text on each line to the left or right of the rectangle moves along with the text inside the rectangle. @xref{Rectangles}.

Many of the sort commands ignore case differences when comparing, if sort-fold-case is non-nil.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

9 Editing Pictures

To edit a picture made out of text characters (for example, a picture of the division of a register into fields, as a comment in a program), use the command M-x picture-mode to enter Picture mode.

In Picture mode, editing is based on the quarter-plane model of text, according to which the text characters lie studded on an area that stretches infinitely far to the right and downward. The concept of the end of a line does not exist in this model; the most you can say is where the last nonblank character on the line is found.

Of course, Emacs really always considers text as a sequence of characters, and lines really do have ends. But Picture mode replaces the most frequently-used commands with variants that simulate the quarter-plane model of text. They do this by inserting spaces or by converting tabs to spaces.

Most of the basic editing commands of Emacs are redefined by Picture mode to do essentially the same thing but in a quarter-plane way. In addition, Picture mode defines various keys starting with the C-c prefix to run special picture editing commands.

One of these keys, C-c C-c, is particularly important. Often a picture is part of a larger file that is usually edited in some other major mode. Picture mode records the name of the previous major mode so you can use the C-c C-c command (picture-mode-exit) later to go back to that mode. C-c C-c also deletes spaces from the ends of lines, unless given a numeric argument.

The special commands of Picture mode all work in other modes (provided the ‘picture’ library is loaded), but are not bound to keys except in Picture mode. The descriptions below talk of moving “one column” and so on, but all the picture mode commands handle numeric arguments as their normal equivalents do.

Turning on Picture mode runs the hook picture-mode-hook. Additional extensions to Picture mode can be found in ‘artist.el’.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

9.1 Basic Editing in Picture Mode

Most keys do the same thing in Picture mode that they usually do, but do it in a quarter-plane style. For example, C-f is rebound to run picture-forward-column, a command which moves point one column to the right, inserting a space if necessary so that the actual end of the line makes no difference. C-b is rebound to run picture-backward-column, which always moves point left one column, converting a tab to multiple spaces if necessary. C-n and C-p are rebound to run picture-move-down and picture-move-up, which can either insert spaces or convert tabs as necessary to make sure that point stays in exactly the same column. C-e runs picture-end-of-line, which moves to after the last nonblank character on the line. C-a runs picture-beginning-of-line. (The choice of screen model does not affect beginnings of lines; the only extra thing this command does is update the current picture column to 0.)

Insertion of text is adapted to the quarter-plane screen model through the use of Overwrite mode (@pxref{Minor Modes}.) Self-inserting characters replace existing text, column by column, rather than pushing existing text to the right. <RET> runs picture-newline, which just moves to the beginning of the following line so that new text will replace that line.

In Picture mode, the commands that normally delete or kill text, instead erase text (replacing it with spaces). <DEL> (picture-backward-clear-column) replaces the preceding character with a space rather than removing it; this moves point backwards. C-d (picture-clear-column) replaces the next character or characters with spaces, but does not move point. (If you want to clear characters to spaces and move forward over them, use <SPC>.) C-k (picture-clear-line) really kills the contents of lines, but does not delete the newlines from the buffer.

To do actual insertion, you must use special commands. C-o (picture-open-line) creates a blank line after the current line; it never splits a line. C-M-o (split-line) makes sense in Picture mode, so it is not changed. C-j (picture-duplicate-line) inserts another line with the same contents below the current line.

To do actual deletion in Picture mode, use C-w, C-c C-d (which is defined as delete-char, as C-d is in other modes), or one of the picture rectangle commands (see section Picture Mode Rectangle Commands).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

9.2 Controlling Motion after Insert

Since self-inserting characters in Picture mode overwrite and move point, there is no essential restriction on how point should be moved. Normally point moves right, but you can specify any of the eight orthogonal or diagonal directions for motion after a self-inserting character. This is useful for drawing lines in the buffer.

C-c <
C-c <LEFT>

Move left after insertion (picture-movement-left).

C-c >

Move right after insertion (picture-movement-right).

C-c ^
C-c <UP>

Move up after insertion (picture-movement-up).

C-c .
C-c <DOWN>

Move down after insertion (picture-movement-down).

C-c `
C-c <Home>

Move up and left (“northwest”) after insertion (picture-movement-nw).

C-c '
C-c <PageUp>
C-c <prior>

Move up and right (“northeast”) after insertion (picture-movement-ne).

C-c /
C-c <End>

Move down and left (“southwest”) after insertion

C-c \
C-c <PageDown>
C-c <next>

Move down and right (“southeast”) after insertion

Two motion commands move based on the current Picture insertion direction. The command C-c C-f (picture-motion) moves in the same direction as motion after insertion currently does, while C-c C-b (picture-motion-reverse) moves in the opposite direction.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

9.3 Picture Mode Tabs

Two kinds of tab-like action are provided in Picture mode. Use M-<TAB> (picture-tab-search) for context-based tabbing. With no argument, it moves to a point underneath the next “interesting” character that follows whitespace in the previous nonblank line. “Next” here means “appearing at a horizontal position greater than the one point starts out at”. With prefix argument, as in C-u M-<TAB>, this command moves to the next such interesting character in the current line. M-<TAB> does not change the text; it only moves point. “Interesting” characters are defined by the variable picture-tab-chars, which should define a set of characters. The syntax for this variable is like the syntax used inside of ‘[…]’ in a regular expression—but without the ‘[’ and the ‘]’. Its default value is "!-~".

<TAB> itself runs picture-tab, which operates based on the current tab stop settings; it is the Picture mode equivalent of tab-to-tab-stop. Normally it just moves point, but with a numeric argument it clears the text that it moves over.

The context-based and tab-stop-based forms of tabbing are brought together by the command C-c <TAB> (picture-set-tab-stops). This command sets the tab stops to the positions which M-<TAB> would consider significant in the current line. The use of this command, together with <TAB>, can get the effect of context-based tabbing. But M-<TAB> is more convenient in the cases where it is sufficient.

It may be convenient to prevent use of actual tab characters in pictures. For example, this prevents C-x <TAB> from messing up the picture. You can do this by setting the variable indent-tabs-mode to nil.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

9.4 Picture Mode Rectangle Commands

Picture mode defines commands for working on rectangular pieces of the text in ways that fit with the quarter-plane model. The standard rectangle commands may also be useful. @xref{Rectangles}.

C-c C-k

Clear out the region-rectangle with spaces (picture-clear-rectangle). With a prefix argument, delete the text.

C-c C-w r

Similar, but save rectangle contents in register r first (picture-clear-rectangle-to-register). @xref{Registers}.

C-c C-y

Copy last killed rectangle into the buffer by overwriting, with upper left corner at point (picture-yank-rectangle). With argument, insert instead.

C-c C-x r

Similar, but use the rectangle in register r (picture-yank-rectangle-from-register).

The picture rectangle commands C-c C-k (picture-clear-rectangle) and C-c C-w (picture-clear-rectangle-to-register) differ from the standard rectangle commands in that they normally clear the rectangle instead of deleting it; this is analogous with the way C-d is changed in Picture mode.

However, deletion of rectangles can be useful in Picture mode, so these commands delete the rectangle if given a numeric argument. C-c C-k either with or without a numeric argument saves the rectangle for C-c C-y.

The Picture mode commands for yanking rectangles differ from the standard ones in that they overwrite instead of inserting. This is the same way that Picture mode insertion of other text differs from other modes. C-c C-y (picture-yank-rectangle) inserts (by overwriting) the rectangle that was most recently killed, while C-c C-x (picture-yank-rectangle-from-register) does likewise for the rectangle found in a specified register.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

10 Editing Binary Files

There is a special major mode for editing binary files: Hexl mode. To use it, use M-x hexl-find-file instead of C-x C-f to visit the file. This command converts the file’s contents to hexadecimal and lets you edit the translation. When you save the file, it is converted automatically back to binary.

You can also use M-x hexl-mode to translate an existing buffer into hex. This is useful if you visit a file normally and then discover it is a binary file.

Ordinary text characters overwrite in Hexl mode. This is to reduce the risk of accidentally spoiling the alignment of data in the file. There are special commands for insertion. Here is a list of the commands of Hexl mode:


Insert a byte with a code typed in decimal.


Insert a byte with a code typed in octal.


Insert a byte with a code typed in hex.

C-x [

Move to the beginning of a 1k-byte page.

C-x ]

Move to the end of a 1k-byte page.


Move to an address specified in hex.


Move to an address specified in decimal.

C-c C-c

Leave Hexl mode, going back to the major mode this buffer had before you invoked hexl-mode.

Other Hexl commands let you insert strings (sequences) of binary bytes, move by shorts or ints, etc.; type C-h a hexl-<RET> for details.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

11 Saving Emacs Sessions

Use the desktop library to save the state of Emacs from one session to another. Once you save the Emacs desktop—the buffers, their file names, major modes, buffer positions, and so on—then subsequent Emacs sessions reload the saved desktop. By default, the desktop also tries to save the frame and window configuration. To disable this, set desktop-restore-frames to nil. (See that variable’s documentation for some related options that you can customize to fine-tune this behavior.)

Information about buffers visiting remote files is not saved by default. Customize the variable desktop-files-not-to-save to change this.

When the desktop restores the frame and window configuration, it uses the recorded values of frame parameters, disregarding any settings for those parameters you have in your init file (@pxref{Init File}). This means that frame parameters such as fonts and faces for the restored frames will come from the desktop file, where they were saved when you exited your previous Emacs session; any settings for those parameters in your init file will be ignored. To disable this, customize the value of frameset-filter-alist to filter out the frame parameters you don’t want to be restored.

You can save the desktop manually with the command M-x desktop-save. You can also enable automatic saving of the desktop when you exit Emacs, and automatic restoration of the last saved desktop when Emacs starts: use the Customization buffer (@pxref{Easy Customization}) to set desktop-save-mode to t for future sessions, or add this line in your init file (@pxref{Init File}):

(desktop-save-mode 1)

If you turn on desktop-save-mode in your init file, then when Emacs starts, it looks for a saved desktop in the current directory. (More precisely, it looks in the directories specified by desktop-path, and uses the first desktop it finds.) Thus, you can have separate saved desktops in different directories, and the starting directory determines which one Emacs reloads. You can save the current desktop and reload one saved in another directory by typing M-x desktop-change-dir. Typing M-x desktop-revert reverts to the desktop previously reloaded.

Specify the option ‘--no-desktop’ on the command line when you don’t want it to reload any saved desktop. This turns off desktop-save-mode for the current session. Starting Emacs with the ‘--no-init-file’ option also disables desktop reloading, since it bypasses the init file, where desktop-save-mode is usually turned on.

By default, all the buffers in the desktop are restored in one go. However, this may be slow if there are a lot of buffers in the desktop. You can specify the maximum number of buffers to restore immediately with the variable desktop-restore-eager; the remaining buffers are restored lazily, when Emacs is idle.

Type M-x desktop-clear to empty the Emacs desktop. This kills all buffers except for internal ones, and clears the global variables listed in desktop-globals-to-clear. If you want this to preserve certain buffers, customize the variable desktop-clear-preserve-buffers-regexp, whose value is a regular expression matching the names of buffers not to kill.

If you want to save minibuffer history from one session to another, use the savehist library.

While Emacs runs with desktop-save-mode turned on, it by default auto-saves the desktop whenever any of it changes. The variable desktop-auto-save-timeout determines how frequently Emacs checks for modifications to your desktop.

The file in which Emacs saves the desktop is locked while the session runs, to avoid inadvertently overwriting it from another Emacs session. That lock is normally removed when Emacs exits, but if Emacs or your system crashes, the lock stays, and when you restart Emacs, it will by default ask you whether to use the locked desktop file. You can avoid the question by customizing the variable desktop-load-locked-desktop to either nil, which means never load the desktop in this case, or t, which means load the desktop without asking.

When Emacs starts in daemon mode, it cannot ask you any questions, so if it finds the desktop file locked, it will not load it, unless desktop-load-locked-desktop is t. Note that restoring the desktop in daemon mode is somewhat problematic for other reasons: e.g., the daemon cannot use GUI features, so parameters such as frame position, size, and decorations cannot be restored. For that reason, you may wish to delay restoring the desktop in daemon mode until the first client connects, by calling desktop-read in a hook function that you add to after-make-frame-functions (see Creating Frames in The Emacs Lisp Reference Manual).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

12 Recursive Editing Levels

A recursive edit is a situation in which you are using Emacs commands to perform arbitrary editing while in the middle of another Emacs command. For example, when you type C-r inside of a query-replace, you enter a recursive edit in which you can change the current buffer. On exiting from the recursive edit, you go back to the query-replace. @xref{Query Replace}.

Exiting the recursive edit means returning to the unfinished command, which continues execution. The command to exit is C-M-c (exit-recursive-edit).

You can also abort the recursive edit. This is like exiting, but also quits the unfinished command immediately. Use the command C-] (abort-recursive-edit) to do this. @xref{Quitting}.

The mode line shows you when you are in a recursive edit by displaying square brackets around the parentheses that always surround the major and minor mode names. Every window’s mode line shows this in the same way, since being in a recursive edit is true of Emacs as a whole rather than any particular window or buffer.

It is possible to be in recursive edits within recursive edits. For example, after typing C-r in a query-replace, you may type a command that enters the debugger. This begins a recursive editing level for the debugger, within the recursive editing level for C-r. Mode lines display a pair of square brackets for each recursive editing level currently in progress.

Exiting the inner recursive edit (such as with the debugger c command) resumes the command running in the next level up. When that command finishes, you can then use C-M-c to exit another recursive editing level, and so on. Exiting applies to the innermost level only. Aborting also gets out of only one level of recursive edit; it returns immediately to the command level of the previous recursive edit. If you wish, you can then abort the next recursive editing level.

Alternatively, the command M-x top-level aborts all levels of recursive edits, returning immediately to the top-level command reader. It also exits the minibuffer, if it is active.

The text being edited inside the recursive edit need not be the same text that you were editing at top level. It depends on what the recursive edit is for. If the command that invokes the recursive edit selects a different buffer first, that is the buffer you will edit recursively. In any case, you can switch buffers within the recursive edit in the normal manner (as long as the buffer-switching keys have not been rebound). You could probably do all the rest of your editing inside the recursive edit, visiting files and all. But this could have surprising effects (such as stack overflow) from time to time. So remember to exit or abort the recursive edit when you no longer need it.

In general, we try to minimize the use of recursive editing levels in GNU Emacs. This is because they constrain you to go back in a particular order—from the innermost level toward the top level. When possible, we present different activities in separate buffers so that you can switch between them as you please. Some commands switch to a new major mode which provides a command to switch back. These approaches give you more flexibility to go back to unfinished tasks in the order you choose.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

13 Hyperlinking and Web Navigation Features

The following subsections describe convenience features for handling URLs and other types of links occurring in Emacs buffer text.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

13.1 Web Browsing with EWW

EWW, the Emacs Web Wowser, is a web browser package for Emacs. It allows browsing URLs within an Emacs buffer. The command M-x eww will open a URL or search the web. You can open a file using the command M-x eww-open-file. You can use EWW as the web browser for browse-url, see section Following URLs. For full details, see EWW in The Emacs Web Wowser Manual.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

13.2 Embedded WebKit Widgets

If Emacs was compiled with the appropriate support packages, it is able to show browser widgets in its buffers. The command M-x xwidget-webkit-browse-url asks for a URL to display in the browser widget. The URL normally defaults to the URL at or before point, but if there is an active region (@pxref{Mark}), the default URL comes from the region instead, after removing any whitespace from it. The command then creates a new buffer with the embedded browser showing the specified URL. The buffer is put in the Xwidget-WebKit mode (similar to Image mode, @pxref{File Conveniences}), which provides one-key commands for scrolling the widget, changing its size, and reloading it. Type C-h b in that buffer to see the key bindings.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

13.3 Following URLs

M-x browse-url <RET> url <RET>

Load a URL into a Web browser.

The Browse-URL package allows you to easily follow URLs from within Emacs. Most URLs are followed by invoking a web browser; ‘mailto:’ URLs are followed by invoking the compose-mail Emacs command to send mail to the specified address (@pxref{Sending Mail}).

The command M-x browse-url prompts for a URL, and follows it. If point is located near a plausible URL, that URL is offered as the default. The Browse-URL package also provides other commands which you might like to bind to keys, such as browse-url-at-point and browse-url-at-mouse.

You can customize Browse-URL’s behavior via various options in the browse-url Customize group. In particular, the option browse-url-mailto-function lets you define how to follow ‘mailto:’ URLs, while browse-url-browser-function lets you define how to follow other types of URLs. For more information, view the package commentary by typing C-h P browse-url <RET>.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

13.4 Activating URLs

M-x goto-address-mode

Activate URLs and e-mail addresses in the current buffer.

You can make Emacs mark out URLs specially in the current buffer, by typing M-x goto-address-mode. When this buffer-local minor mode is enabled, it finds all the URLs in the buffer, highlights them, and turns them into clickable buttons. You can follow the URL by typing C-c <RET> (goto-address-at-point) while point is on its text; or by clicking with mouse-2, or by clicking mouse-1 quickly (@pxref{Mouse References}). Following a URL is done by calling browse-url as a subroutine (see section Following URLs).

It can be useful to add goto-address-mode to mode hooks and hooks for displaying an incoming message (e.g., rmail-show-message-hook for Rmail, and mh-show-mode-hook for MH-E). This is not needed for Gnus, which has a similar feature of its own.

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

13.5 Finding Files and URLs at Point

The FFAP package replaces certain key bindings for finding files, such as C-x C-f, with commands that provide more sensible defaults. These commands behave like the ordinary ones when given a prefix argument. Otherwise, they get the default file name or URL from the text around point. If what is found in the buffer has the form of a URL rather than a file name, the commands use browse-url to view it (see section Following URLs).

This feature is useful for following references in mail or news buffers, ‘README’ files, ‘MANIFEST’ files, and so on. For more information, view the package commentary by typing C-h P ffap <RET>.

To enable FFAP, type M-x ffap-bindings. This makes the following key bindings, and also installs hooks for additional FFAP functionality in Rmail, Gnus and VM article buffers.

C-x C-f filename <RET>

Find filename, guessing a default from text around point (find-file-at-point).

C-x C-r filename <RET>

ffap-read-only, analogous to find-file-read-only.

C-x C-v filename <RET>

ffap-alternate-file, analogous to find-alternate-file.

C-x d directory <RET>

Start Dired on directory, defaulting to the directory at point (dired-at-point).

C-x C-d directory <RET>

ffap-list-directory, analogous to list-directory.

C-x 4 f filename <RET>

ffap-other-window, analogous to find-file-other-window.

C-x 4 r filename <RET>

ffap-read-only-other-window, analogous to find-file-read-only-other-window.

C-x 4 d directory <RET>

ffap-dired-other-window, like dired-other-window.

C-x 5 f filename <RET>

ffap-other-frame, analogous to find-file-other-frame.

C-x 5 r filename <RET>

ffap-read-only-other-frame, analogous to find-file-read-only-other-frame.

C-x 5 d directory <RET>

ffap-dired-other-frame, analogous to dired-other-frame.

M-x ffap-next

Search buffer for next file name or URL, then find that file or URL.


ffap-at-mouse finds the file guessed from text around the position of a mouse click.


Display a menu of files and URLs mentioned in current buffer, then find the one you select (ffap-menu).

[ << ] [ < ] [ Up ] [ > ] [ >> ]         [Top] [Contents] [Index] [ ? ]

14 Games and Other Amusements

The animate package makes text dance (e.g., M-x animate-birthday-present).

M-x blackbox, M-x mpuz and M-x 5x5 are puzzles. blackbox challenges you to determine the location of objects inside a box by tomography. mpuz displays a multiplication puzzle with letters standing for digits in a code that you must guess—to guess a value, type a letter and then the digit you think it stands for. The aim of 5x5 is to fill in all the squares.

M-x bubbles is a game in which the object is to remove as many bubbles as you can in the smallest number of moves.

M-x decipher helps you to cryptanalyze a buffer which is encrypted in a simple monoalphabetic substitution cipher.

M-x dissociated-press scrambles the text in the current Emacs buffer, word by word or character by character, writing its output to a buffer named ‘*Dissociation*’. A positive argument tells it to operate character by character, and specifies the number of overlap characters. A negative argument tells it to operate word by word, and specifies the number of overlap words. Dissociated Press produces results fairly like those of a Markov chain, but is however, an independent, ignoriginal invention; it techniquitously copies several consecutive characters from the sample text between random jumps, unlike a Markov chain which would jump randomly after each word or character. Keep dissociwords out of your documentation, if you want it to be well userenced and properbose.

M-x dunnet runs a text-based adventure game.

If you want a little more personal involvement, try M-x gomoku, which plays the game Go Moku with you.

If you are a little bit bored, you can try M-x hanoi. If you are considerably bored, give it a numeric argument. If you are very, very bored, try an argument of 9. Sit back and watch.

M-x life runs Conway’s Game of Life cellular automaton.

M-x morse-region converts the text in the region to Morse code; M-x unmorse-region converts it back. M-x nato-region converts the text in the region to NATO phonetic alphabet; M-x denato-region converts it back.

M-x pong, M-x snake and M-x tetris are implementations of the well-known Pong, Snake and Tetris games.

M-x solitaire plays a game of solitaire in which you jump pegs across other pegs.

The command M-x zone plays games with the display when Emacs is idle.

“Real Programmers” deploy M-x butterfly, which uses butterflies to flip a bit on the drive platter, see https://xkcd.com/378.

Finally, if you find yourself frustrated, try describing your problems to the famous psychotherapist Eliza. Just do M-x doctor. End each input by typing <RET> twice.

[Top] [Contents] [Index] [ ? ]



For PostScript files, GhostScript is a hard requirement. For DVI files, dvipdf or dvipdfm is needed. For OpenDocument and Microsoft Office documents, the unoconv tool is needed.


The needed external tools for the document type must be available, and Emacs must be running in a graphical frame and have PNG image support. If these requirements is not fulfilled, Emacs falls back to another major mode.


You should not suspend the shell process. Suspending a subjob of the shell is a completely different matter—that is normal practice, but you must use the shell to continue the subjob; this command won’t do it.


Some programs use a different environment variable; for example, to make TeX use ‘emacsclient’, set the TEXEDIT environment variable to ‘emacsclient +%d %s’.

[Top] [Contents] [Index] [ ? ]

About This Document

This document was generated on May 30, 2018 using texi2html.

The buttons in the navigation panels have the following meaning:

Button Name Go to From 1.2.3 go to
[ << ] FastBack Beginning of this chapter or previous chapter 1
[ < ] Back Previous section in reading order 1.2.2
[ Up ] Up Up section 1.2
[ > ] Forward Next section in reading order 1.2.4
[ >> ] FastForward Next chapter 2
[Top] Top Cover (top) of document  
[Contents] Contents Table of contents  
[Index] Index Index  
[ ? ] About About (help)  

where the Example assumes that the current position is at Subsubsection One-Two-Three of a document of the following structure:

This document was generated on May 30, 2018 using texi2html.