I always use Emacs's shell. I create 2 buffers (C-x 2) and use one of them as a terminal (M-x shell).
Sometimes I use a command a lot of times (e. g. python3 test.py). I use shortcut <C-up> to repeat the last command. And it's very inconvenient that Emacs saves duplicates of commands. I have to press <C-up> many times so that I can run any other command.
The problems is only in Emacs's shell. I have setting HISTCONTROL=ignoreboth in .bashrc and in usual terminal it works very well.
Is it possible to configure Emacs so that it ignores duplicates?
It's customizable
(setq comint-input-ignoredups t)
Setting 'comint-input-ignoredups' as suggested by Jürgen Hötzel will of course solve the immediate, but here are some alternatives.
You could choose to use the command 'ansi-term' instead of 'shell', this will run a bash (or whatever you choose) so you get the same control as in (say) an xterm. As an added bonus, you also get a rather capable terminal so applications that depend on this will work.
Another possibility, if you have something you want to repeatedly run, is to use the 'compile' command. This can run any shell command, it does not have to be a compiler. The advantage here is that you keep the running of the test out of your shell command history (no matter how you run your shell) and you get the output in a separate buffer. If the output is suitably organised and/or the compilation buffer is suitably configured, you could also use the 'next-error' command to jump back to the appropriate part of the source.
Finally, I would like to mention that the 'shell' command supports searching the command history. It is by default bound to M-r (comint-history-isearch-backward-regexp).
Related
Is there a possability in Emacs to run custom commands or a set of commands (eg. shell) with a user defined emacs shortcut/hook?
To make this clearer. I am working with an embedded system (target), but developing on the host. After writing and compiling code (eg using emacs compile command), I would like to copy (eg. scp) the binary to the system right away.
Furthermore it would be good if the custom shorcut/hook would be easy to adjust (eg. ip address of the target).
Btw: I am aware of the "shell-command" in emacs, but that is not quite what I am looking for.
Would appreciate any advice. Thanks!
I ended up using C-x C-f /root#my-target-ip:/tmp/myfile to edit the file directly on the target. Then you can run it with shell-command ./myfile from within emacs.
I have emacs as my defualt editor in linux, and I also have alias in my .cshrc file.
alias e "emacs -mm"
Sometime I just want to hit the file name in the command line and open it in emacs directly with out the editor beign metioned.
Example instead of
$ e foo.cc&
What I want is to open
$foo.cc
May be this is lazy to do but it saves a lot time if you have so many files to handle. Thanks for the help.
You probably cannot open a file with $foo.cc (and that would be ambiguous for a shell script script.sh: would script.sh means "edit the file script.sh" or "run the shell script script.sh" ?). However,
You might want to use xdg-open, or the $EDITOR variable (see environ(7)). If you always have a single emacs running, you might set EDITOR to emacsclient in your ~/.bashrc (if using /bin/bash) or your ~/.zshrc (if using /bin/zsh)
BTW many editors (including emacs, gedit, vim) are able to edit several files, i.e. $EDITOR *.c
And depending upon your login shell (zsh, fish, or bash) you could set up a shell function or alias to simply type e foo.c; I feel that it is not worth the effort, since with autocompletion I just have to type 3 keys e m tab to get emacs (and often the up arrows are enough)
Actually I start only once every day emacs then open many files inside it (and also I compile inside emacs)
BTW, you should avoid csh since it is considered harmful. Install a good interactive shell (e.g. with sudo aptitude install zsh zsh-doc) and use once chsh(1) to make it your login shell.
I'm somewhat new to gnu emacs, and so perhaps this is a noob question, but I have a few batch files I use a lot when coding in emacs to compile/build/execute/debug/etc. I am wondering how I could A) run these batch files from emacs without having to keep opening a cmd prompt window or going to windows explorer and B) bind this to a key shortcut (perhaps I could specify the file?) I have seen several things online about running emacs in batch-mode, but I don't believe this is what I'm looking for. And I know it is possible because I have seen others run batch from emacs (output and everything would appear in a new buffer adjacent to the current as if you did C-x 3)
Thanks in advance!
To run an arbitrary shell command in Emacs, you call shell-command which is bound to M-!
See C-hf shell-command (or C-hkM-!) for details.
I believe in the Windows-native Emacs, the default shell is cmd (or some alias thereof), so I'm reasonably confident that this is what you're thinking of.
I'm not sure whether all of the following work in Windows, but related commands are:
M-& - async-shell-command
M-| - shell-command-on-region
And with a prefix argument (e.g. C-uM-!) any of those commands will insert the shell command's output into the current buffer. (In the case of shell-command-on-region, that replaces the region.)
A very useful time-saver for working with shell is that you can search for command history using Ctrl - r. Emacs has the same functionality in its shell mode: C-S-r. However, the shell's history is cleared every time emacs is restarted. Any idea how to persist the history?
In general the answer would be:
M-x customize-group RET savehist RET
Enable savehist-mode and configure the variables you wish to persist between sessions.
However in this case I think you'll find that it's your shell which is tracking its command history; not Emacs.
e.g. for bash, commands are stored in $HISTFILE (by default ~/.bash_history), but depending on your settings, the history from one shell may be clobbered by another shell, depending on which one exits last. For bash, read the HISTORY section of its man page.
Edit: For comint-mode and derivatives (including shell-mode) the searchable history is read in via comint-read-input-ring, which uses comint-input-ring-file-name which you can set in a mode hook. However I would suggest that you actually set your HISTFILE environment variable to ~/.zsh_history because shell-mode automatically defers to that.
It seems a little odd to me that zsh doesn't already do this, mind (assuming you're starting Emacs from a zsh shell?). If you use some kind of GUI launcher, OTOH, the environment Emacs starts in may not be what you want? I couldn't say for sure.
Failing that, the documentation suggests that this ought to work:
(add-hook 'shell-mode-hook 'my-shell-mode-hook)
(defun my-shell-mode-hook ()
(setq comint-input-ring-file-name "~/.zsh_history")
(comint-read-input-ring t))
(but using HISTFILE would be better).
n.b. I'm using Emacs 24.3 and the keybindings are different to those you have described (M-r to search the history in shell-mode; C-S-r is unbound for me), so YMMV. I always run my shells inside ansi-term, so I couldn't tell you whether this functionality has changed recently.
This occurs when I set the vim's shell to be interactive:
set shellcmdflag=-ic
or
set shell=/bin/bash\ -i
I like these because they give syntax highlighting to the output (eg.: !ls)
But the cost is that I have to type fg # every time.
Is this a default behavior?
How can I get interactive shell in vim without having to make it run foreground?
You can't. This behavior is perfectly normal and expected and in line with Vim's author's philosophy. It's very unlikely to change in the future.
If you want a shell inside Vim, you'll have to install a plugin like Conque or Vimshell.
I usually just use tmux instead. You can split the terminal and have a normal interactive shell and an instance of vim running side by side - very handy.