i want to create a shortcut to run typescript in vim how can i access the filename in .vimrc
i want to run short scripts.
nmap <leader>r :! ts-node current-file<return>
First try %, as this gets substituted by the file name (as explained in :help :!):
nmap <Leader>r :! ts-node %<CR>
But if the file name contains whitespaces or special characters, you might see some errors. In that case, you need to escape the file name as follows:
nmap <Leader>r :execute '!ts-node ' . shellescape(expand("%"))<CR>
You're strongly adviced to read :h cmdline.txt, section 6 :h cmdline-special from top down to bottom.
nnoremap <leader>r :!ts-node %:S<CR>
Related
How can I configure Vim to set
"require 'pry'; binding.pry"
in Ruby and
"debugger;"
in JavaScript when pressing F2 via key mapping?
You can set this in your .vimrc as follows:
autocmd FileType ruby map <F2> orequire 'pry'; binding.pry<ESC>
autocmd FileType javascript map <F2> odebugger;<ESC>
When the F2 key is pressed in a *.rb file, "require pry" will be set and "debugger" is set in a *.js file.
The other answer is correct, but not completely correct. You should use the noremap variant of map (see :h noremap), and the proper noremap for whatever mode your are in. If that's insert mode, then it's inoremap <F2> require..., or nnoremap for normal mode, etc.
You can also put those mappings into their own file instead of your vimrc so that you don't need to use autocommands (see :h ftplugin). And (thanks to the comments for reminding me) use <buffer> mappings so they only apply to the file you set them on (see :h <buffer>). In all, this is a good setup for you:
In ~/vim/after/ftplugin/ruby.vim, put the line:
inoremap <buffer> <F2> require 'pry'; binding.pry
and in ~/vim/after/ftplugin/javascript.vim, put the line:
inoremap <buffer> <F2> defbugger;
On windows, the vim directory is instead the vimfiles directory. If you want those mappings in normal mode instead of insert mode, you need to put i or O or another character like that at the front to go into insert mode and put <Esc> on the end to exit insert mode.
In my vim editor, I can find two mappings of through following commands:
:imap <CR>
and it outputs:
i <CR> &#<SNR>60_AutoPairsOldCRWrapper73<SNR>60_AutoPairsReturn
i <CR> <CR><Plug>DiscretionaryEnd
I want to disable the first one, so I add it into my .vimrc file:
iunmap <buffer> <CR>
but vim shows error no such mapping error when I open my editor, but actually I can disable the mapping by typing command in editor:
:iunmap <buffer> <CR>
I want to know why I cann't make it work in my .vimrc configuration file.
Plugins are sourced after your vimrc so the mapping you want to disable is not defined when that comand is executed.
That said, the plugin's README tells you how to replace the default mappings. So… read it and experiment.
I've got these bindings in my .vimrc. They work most of the time, but sometimes they don't: they will just save the file but not run it.
Then I go into insert mode and exit it back to normal mode, and they work again. What could be the problem?
Thanks!
autocmd FileType python map <C-k> :write <CR> :! python % <CR>
autocmd FileType lisp map <C-k> :write <CR> :! clisp % <CR>
autocmd FileType scala map <C-k> :write <CR> :! scala % <CR>
There are many topics to discuss here so with out further ado:
Mappings
Your current mappings map <c-k> ... will only work in normal, visual, and operator-pending modes. However executing your mappings in visual mode or operator-pending modes would save the buffer with only a range of lines (read not good). I suggest you make your mappings for normal mode only.
Two general rules of thumb:
Always supply a mode like n for normal.
Always use noremp instead of map unless you are mapping to a <Plug> mapping.
So one of your mappings might look something similar to this:
nnoremap <c-k> :w<cr>:!python %<cr>
For more information:
:h :map-modes
:h map-overview
:h :nore
:h map-listing
:h map-verbose
Filetype based mappings
You need to do 2 things:
Create a mapping local to a specific buffer by using the <buffer> option for noremap.
Load the mappings for just a specific filetype.
This can be done via an autocmd and FileType event in your .vimrc like so:
autocmd FileType python nnoremap <buffer> <c-k> :w<cr>:!python %:p<cr>
The other way option is by creating a filetype plugin. (see :h ftplugin for more details)
A simple example is do create a file named, ~/.vim/ftplugin/python.vim and place your mappings inside like so:
nnoremap <buffer> <c-k> :w<cr>:!python %:p<cr>
I personally lean more towards the ftplugin approach but having a everything in your .vimrc file can be nice.
For more help see:
:h :au
:h FileType
:h map-local
:h ftplugin
:make
A more vim like way of doing this would be to use :make. (Assuming you want to lint vs execute your current buffer)
:make will execute the 'makeprg'. It defaults to make which is great of C projects
After running :make the quickfix list will be contain any errors.
Set your compiler via the :compiler command.
Extra parameter can be passed like so :make foo-command
Current filename can be represented by %. e.g. :make %
Often people set do :complier/'makeprg' in side of ftplugins e.g. ~/.vim/ftplugin/perl.vim or autocmd's e.g. autocmd FileType perl compiler perl.
Fop more help see:
:h :make
:h 'makeprg'
:h :compiler
:h c_%
quickfix list
Use :cnext and :cprev to move between your errors.
:copen to open up the quickfix list in a window (:cclose to close)
:cwindow to open quickfix list window only if there are errors
May want to use better mappings for :cnext and friends. I suggest Tim Pope's unimpaired plugin
For more help see the following:
:h quickfix
:h :cnext
:h :cope
Alternatives to using :make
Just use <c-z> to suspend vim and run your build system. (Cons: loose out on the quickfix list)
Use :! to compile. (Same cons as suspending) e.g. :!make
Syntastic is a syntax checking system that checks files on save
Ale (Asynchronous Lint Engine) is a plugin for providing linting in NeoVim and Vim 8 while you edit your text files
Dispatch can be used to run things in the background. Great for test suites
May want to consider terminal multiplexers like tmux or screen.
SingleComplile tries and takes some of the work out of using :make
Conclusion
Personally I would install ALE as this removes the need for your mappings. It is also a great idea to learn how to use the location (or quickfix) list. When you are ready to get your hands dirty then you can learn and use :make.
tl;dr
Install ALE.
The space inserted between <CR> and : could lead to errors, because this space is really entered during the execution of the mapping.
With :noremap or :nnoremap (which should be used most of the times, as Peter Rincker wrote in his answer), it wouldn't be a big problem, since it only would move the cursor one char to the right (see :h <space>).
But with :map, it could trigger a custom mapping (or a partial one).
Then, instead of:
autocmd FileType python map <C-k> :write <CR> :! python % <CR>
the following would be better:
autocmd FileType python map <C-k> :write <CR>:! python % <CR>
and this one, really better:
autocmd FileType python nnoremap <C-k> :write <CR>:! python % <CR>
I'm trying to bind F5 to run sass --update on the current file. Where I'm stuck is the output. My decided solution is to include a comment of the output file in the first line of the input file, e.g.:
//C:\Users\wuffie\Projects\output.css
How can I read the line, trim the beginning slashes, and put it into my command? Will it require a full-fledged function? Here is what I have so far:
au BufEnter,BufNew *.sass nmap <F5> :w<CR>:tabnew <bar> silent :r !sass --update expand('%:p'):
Let's first clean up your command (I'm omitting the :autocmd for brevity; alternatively, you could place the :map into ~/.vim/after/ftplugin/sass.vim, assuming the filetype is detected as sass):
nnoremap <buffer> <F5> :w<Bar>tabnew<Bar>silent :r !sass --update %:p<CR>
You should use :noremap; it makes the mapping immune to remapping and recursion.
With the <buffer> keyword, the mapping is only defined for the file's buffer, not polluting other buffers.
Use <Bar> everywhere, and add the missing concluding <CR>.
The expand() wouldn't evaluate right, and it's not necessary, the %:p is interpreted by :r already. (But we'll need it with the extension, read on):
Now, let's build the reading of the first line into that:
nnoremap <buffer> <F5> :execute 'w<Bar>tabnew<Bar>silent :r !sass --update ' . expand('%:p') . ':' . getline(1)[2:]<CR>
This uses :execute to build the command-line, and getline(1) to access the first line of the file. I've using String indexing to cut off the first two comment characters ([2:]); you could also use strpart() here, or substitute() for more elaborate cases. To make this bulletproof, wrap that final expression in shellescape(..., 1).
nnoremap <buffer> <F5> :execute 'w<Bar>tabnew<Bar>silent :r !sass --update ' . shellescape(expand('%:p') . ':' . getline(1)[2:], 1)<CR>
I realize that I can :nmap <leader>rc :!cat %<CR> to provide an easy set of triggers, but I would like to do this instead.
nmap <leader>rc :up :!cat %<CR> but it complains about needing only one filename. How do I get vim to recognize both commands, in series?
You are missing a <CR> after :up. <CR> tells vim you want a carriage return here.
nmap <leader>rc :up<CR> :!cat %<CR>
The reason up is complaining about multiple file names is that it sees :!cat and %<CR> as two arguments to up.
So the new macro executes
:up
:!cat %
instead of
:up :!cat %
(Side Note: you should probably use nnoremap instead of nmap)
ZyX recommends using the following mapping instead.
nnoremap ,rc :up\|execute "!cat" shellescape(#%, 1)<CR>
This uses | to separate commands and escapes the %. Escaping the % leads to a more robust mapping just incase the filename contains special characters.
Help for :h execute and :h shellescape