In windows shell it's possible to pause the output just by selecting a text in the shell window (a QuickEdit mode must be enabled). The same is possible in the cygwin shell (again with a QuickEdit mode).
How can I pause the output in emacs shell?
I need to pause it because some tasks are very verbose and it's difficult to search the necessary text in a big buffer.
When I execute M-x comint-stop-subjob I get a message "No SIGTSTP support"
As an example try to pause the following:
find . -type f -name "*.txt" -exec cat {} ';'
Not exactly what you asked for, but clone-buffer will get you close to what you want. After binding it to a convenient key and running your find command, calling clone-buffer will copy the contents of the buffer to a new buffer. The find command in the original buffer keeps running, but the cloned buffer contains only the output of find up to the point that you cloned it.
You can invoke this command multiple times in the original window, giving you a series of stop points to examine. I don't know how to actually pause the original window though.
Note that the newly cloned buffers are fully functional shell buffers.
You can stop the job by pressing C-c C-z. Then just enter fg to make it work again.
I don't know about pausing the output, but you can presumably just narrow the buffer?
C-xnn to narrow-to-region
C-xnn to narrow-to-page
C-xnw to widen to full buffer
Related
In emacsclient, is there a way to open a file from the command line such that its buffer will remain open on the server after I close the client?
Alternatively, is there a command I can run from in a client to tell it to effectively detach from a buffer, so that the buffer does not get killed when the client exits?
Normally when closing emacsclient either with C-x C-c or C-x #, the buffer(s) associated with that client get killed, which is usually convenient behavior, but sometimes I would like for buffers to stay alive after closing. So far the only way I have found to accomplish this is to run the client without specifying any files, then visit them with C-x C-f, but I'm wondering if there's a better way to do this.
You should be able to do this by using the -n option. That means that it won't wait for you to "finish" the buffer and it'll just stay in the buffer list. I use this with emacsclient myself.
So, one option is to use emacsclient's eval command line option to run a lisp command to find the file you want.
emacsclient -c -e '(find-file "my_file")'
Obviously this is a lot more to type than the command sequence emacsclient -c, C-x C-f, my_file, but it could pretty easily be wrapped in a script that takes an extra argument to tell it whether to just choose the file or use find-file.
Someone more well versed in elisp than I could probably just add the option directly into emacs.
According the info manual, if you never want to kill emacsclient buffers, when you're done with them, you can customise the server-kill-new-buffers variable (more information at C-h v server-kill-new-buffers).
For you use-case, depending on how often you want the buffers killed or not, you could set the above variable to nil and then manually kill the buffers that you do want killed.
With Emacs, if the current buffer is one that's "visiting" a normal file (for example), whose full pathname is /path/to/somefile, and one runs find-file (C-x C-f), the prompt that appears in the mini-buffer is something like
Find file: /path/to/▮
...with the cursor placed as indicated above by ▮. IOW, the suggested path shown by default is always to the directory containing the file that the current buffer is visiting.
If, however, the current buffer is an Emacs shell process, and one runs find-file, then, AFAICT, the path shown in the prompt remains fixed at the value of $PWD when the shell process was started, irrespective of the current value of $PWD:
Find file: /pwd/at/startup/▮
This behavior is not so useful, because the $PWD at startup often becomes irrelevant later on. It would be really nice if the directory shown in find-file's prompt were instead the shell process's current $PWD.
Is there a simple way to modify find-file to behave this way whenever the current buffer is a shell process?
You want "shell directory tracking". E.g. check dirtrack-mode or shell-dirtrack-mode.
shell-dirtrack-mode tries to parse "cd" commands, (event_jr: which in my experience does not work consistently). dirtrack-mode uses the prompt regexp, which works very well.
There are a number of ways to manage this. As Stefan notes, there are a couple of built in packages that manage it.
My preferred way is to alter your prompt (when in Emacs) to have the $PWD embedded in it, and then Emacs strips it out and uses it. This has the benefit of always being up to date. I've found that dirtrack-mode sometimes gets out of sync.
Check out my solution here, which is a modification of a similar implementation on the Emacs Wiki.
Say I have long debugging session in Perl with perldb on Emacs (M-x perldb). The output piles up, and at some point it becomes difficult to navigate through it.
I can restart the debugger with R, but that doesn't clear up the previous output. Is there a command to clear up the output of the buffer of the Perl debugger without having to kill the debugger and starting a new session?
You can run comint-truncate-buffer. This will remove all but the last 1024 lines from the buffer. The size can be customized by changing comint-buffer-maximum-size.
If you'd like to have the buffer truncated automatically, run this snippet of elisp:
(add-hook 'comint-output-filter-functions 'comint-truncate-buffer)
Debugger input and output is buffer contents like any other — if you want to delete it, just do it. For example, C-x h C-w (mark-whole-buffer followed by kill-region) works in perldb buffers just fine, and is the closest equivalent to a "clear screen" command in a text terminal.
I would like to write a small script in ELisp that would:
send a command to a given buffer
get its output
parse it
send it to another buffer
I am struggling with point 2: I cant get the output of a command. For example, if I have a shell buffer on, I can use
(process-send-string "shell" "help\n")
to send "help" to my shell buffer. It will then show the list of the commands available. But how can I get this list to use it somewhere else?
Thanks,
S4m
(buffer-string) returns the contents of the current buffer, so (with-current-buffer <buf> (buffer-string)) will return the contents of <buf>.
I don't know the exact emacs commands for this off the top of my head, but one option would be to do the following:
Set the mark in the shell buffer right below the command line
Execute the command.
Move the point to the end of the file and kill the text between there and the mark.
Move to the destination buffer and yank the text into there.
Have you considered using the shell-command or shell-command-to-string functions?
The don't "send a command to a buffer" like you asked, but they do both allow running a command through a process that will be started just for that purpose and either dumping the output into a target buffer or collecting it into a string.
I'm using zsh in OS X Terminal.app and for quite a while, I've been longing for a way to jump back and forth between prev/next prompts in the terminal's output.
One convenience with this would be to be able to review (and track errors at) the end of each command's output; eg. when you building stuff from source with ./configure; make; make install. Note: I'm obviously not referring to jumping back and forth in the command-history, but for a way to take a peek at the endings of each command's output.
Has anyone heard of such functionality in the *nix (preferrably also Mac) world? Would it require some sort of OS-centric Terminal plugin, or can it be programmatically done via a shell script which can be tied to a keyboard shortcut? Maybe I'm the only one thinking about this? :)
Edit: Here's an example scenario: Let's say I want to compile and install some program (using standard ./configure && make && make install procedure) and after the make command, I run into some errors. Now, the way I understand it (I may be completely wrong), the crucial error causing the make command to fail usually shows up in the last line(s) in the output, no? Anyway, at this point, I might do something like cat INSTALL to read up on the INSTALL document to check whether there's something I've missed. NOW, if I want to go back to see what the error was, that caused my initial make command to fail, I then have to manually scroll up to that position again, since my cat INSTALL command printed a ton of text after it.
I don't know if this scenario is the most elucidative – but there are many other situations where I wish I could just "jump" back to previous prompt lines and check up on previous command output; whether it was a simple ls command, make, git status, or whatever it was – swapping positions in the window by means of using prompt lines as "bookmark" positions seems an interesting idea to me.
command + left or right goes between tabs in iterm. is this what you are asking?
Emacs has a shell-mode that runs a shell inside the Emacs editor, providing a rich environment of additional commands for navigating and working with shell commands. This includes commands for going to the previous/next prompt, and deleting the output from commands so you can "clean up" and issue another command.
If you aren't familiar with Emacs: to start a shell inside Emacs, run emacs from the shell, then type Esc-x (or Meta-x, if you have "Use option as meta key" enabled in Terminal > Preferences > Settings > [profile] > Keyboard). This will ask for a command to execute. Enter shell.
To see a list of commands you can use in Shell Mode, enter Control-h m. Here are the ones for moving the cursor to the previous/next prompt:
C-c C-n comint-next-prompt
C-c C-p comint-previous-prompt
These commands would also be useful:
C-c C-r comint-show-output
C-c C-o comint-delete-output