The underlying vim-textobj-user doesn't support count, so this could be a challenge.

In Markdown paragraphs with inline code, the inline code should be exempted from replacement, education, etc.

(This is probably a feature request rather than a bug.)

The Installation section refers to a dependency on "kana/vim-textobject-user". It should instead say "kana/vim-textobj-user".

The link is correct; only the text is wrong. But this is still a problem for those who use package managers that take just the repo name, and who therefore just copy and paste it from the README.

It'd be nice to use the https://github.com/rhysd/vim-operator-surround to add/replace quotes around text objects.

It likely involves an entry to g:operator#surround#blocks.

e.g. with this buffer (^ indicating cursor):

hello
^

daq or diq will delete the h. The behaviour with built-in ' and " is to do nothing.

vim-textobj-quote: 32cd093 (latest, I believe)
neovim: v0.2.0-1089-g8d139553 (full build info here)
macOS: 10.12.4

Problem: If I try to use quotes, I get raw escapes mixed with text. For example, if I type It's, I end up with this result: It<t_x30x128><t_x30s>. (Note that the 's' ends up inside the second bit of escape code.)

Please let me know if you need me to do anything else to debug the problem. Thanks.

Hello!

I'm considering to switch from Emacs to (neo)vim and looking for a decent replacement for Emacs' typoel package which allows me to e.g. enter '...' and get ellipsis. Same with '--'/'---' for en/em-dash along with entering appropriate quotes like 'educate' with this plugin, so wonder if this plugin can help with it or some other hint to have a proper support when writing content (articles, blog posts etc.) using Markdown Extra markup?

When using r" the " is not replaced with the 'correct' quotation mark
When using xi" it is

This is at least unexpected

Same as Shougo/dein.vim#345, because there are non-english characters in https://github.com/preservim/vim-textobj-quote/blob/master/doc/textobj-quote.txt (e.g. “‘curly’”), but vim help files which has an extension named 'txt' are allowed to have only english character, so the helptags will fail.

A simple solution is to rename the file extension from txt to ??x. (?? should be a two-letter language code).

thanks!

Under Manjaro Linux, I have installed your package through vundle, pasted the configuration into my .vimrc, and I have the following error message every time I launch vim:

Error detected while processing function textobj#quote#init:
line    16:
E117: Unknown function: textobj#user#plugin

I know this plugin is designed primarily for prose, but I wanted smart quotes in comments in my code.

I've managed to do this. Currently it looks like this in my .vimrc:

" Smart quotes toggle {{{4
function! s:ToggleEducate()
    if g:textobj#quote#educate
        silent NoEducate
        silent let g:textobj#quote#educate = 0 " For smart quotes in comments
        echom "Smart quotes off"
    else
        silent Educate
        silent let g:textobj#quote#educate = 1 " For smart quotes in comments
        echom "Smart quotes on"
    endif
endfunction
nnoremap <Leader>' :call <SID>ToggleEducate()<Cr>

" Smart quotes in comments {{{4
function! s:SmartQuotesInComments()
    " Respect the setting above, only do smart quotes in comments
    " If the educate variable is truthy
    if g:textobj#quote#educate
        if synIDattr(synID(line('.'),col('.')-1,1),'name') =~? 'comment'
            exec 'silent Educate'
        else
            exec 'silent NoEducate'
        endif
    endif
endfunction
augroup smartquotes
    autocmd!
    autocmd InsertCharPre * if index(textlikeft, &filetype) < 0 |
        \   call <SID>SmartQuotesInComments()
    \ | endif
    autocmd InsertLeave * if index(textlikeft, &filetype) < 0 |
        \   exec 'silent NoEducate'
    \ | endif
augroup END

I've achieved this by loading the plugin for both text-like and non-text-like file types. For text-like file types, Educate is enabled when loading the plugin.

For non-text-like file types (basically, for source code), the plugin is loaded, but Educate is not enabled. However, the configuration variable is still set to 1 (after plugin init). I figured that the plugin only read it on initialization anyway (i.e. changing it after init doesn't change the plugin's behaviour), so I've reused it for the following purpose:

In non-text-like file types:

• if the configuration variable is 1, a function is run before a character is inserted in insert mode. This function checks to see whether we're currently in a piece of text that has a syntax group whose name contains the string comment (case-insentive). If it does, enable Educate. If it does not, disable Educate.
• if the configuration variable is 0, don't even check what the current syntax group is, and therefore don't enable Educate
• always disable Educate if we leave insert mode.

The result is that I have smart quotes in my code's comments automatically. I am also able to toggle that off in case I need ASCII quotes in my comments (e.g. when typing a string in VimL, before the closing double quote is entered, Vim thinks we're in a comment—if I couldn't toggle Educate for comments I'd go mad).

I developed and tested this on an Asus Transformer Book (T100TA), which has a wimpy little Atom processor. I didn't notice any performance impact.

I thought this might be of interest to others. Perhaps you'd like to add this functionality to your plugin, or mention this in the docs. ☺

Not that this is exactly textobj-quote's fault, but if easyclip is configured as follows (which is what the easyclip documentation reccomends):

set clipboard=unnamed

then ReplaceWithCurly and friends don't make any change to the text because pasting no longer uses the " register. It seems that using set clipboard=unnamedplus fixes this problem, although I haven't fully worked out if it has other knock-on effects.

I guess textobj-quote could check for easyclip's existence and alter the paste register it uses; or maybe even documenting the issue might be enough?

A feature request: If Educate is set, it changes quotes according to the current &spelllang. That is, if for example &spelllang = "de", the setting

let g:textobj#quote#doubleDefault = '„“' " „doppel“
let g:textobj#quote#singleDefault = '‚‘' " ‚einzel‘

is used.

In French also there is a non breakable space between the quote mark and text.
I just tried as a test to put u00a0 twice between the french quotes as below but it did' work.
autocmd FileType *.adoc call textobj#quote#init({ 'double':'« »', 'single':'«»', 'educate': 1 })
Can textobj help me out here? Or any other suggestions?

How about a textobject, let's call it u, going to the quotes closest to the current cursor position.

That is, viu selects everything in between any two curly, single, double or any other kind of quotes.

There should be a simple way to turn on and off the "educate" feature on the fly, interactively, globally or in the current buffer.

The documented let g:textobj#quote#educate=0 doesn't do it, which I take to be either a bug or not intended to work in an already-loaded buffer.

When starting vim with the file which activates call textobj#quote#init (actually, neovim from git (commit neovim/neovim@e46534b42 ) I get this error:

Error detected while processing \
    /home/matej/.vim/pack/minpac/start/vim-textobj-quote/autoload/textobj/quote.vim:                                                                      
line    9:                                                                                 
E48: Not allowed in sandbox: scriptencoding utf-8

I don't know if it matters, we could use "\u02a4" instead of switching scriptencoding.

After that something is wrong, e.g., :w says E13: File exists (add ! to override) , I have to run :w! to save the file.

No matter whether I rely on default (which is supposed to be that educate is on), or whether I add 'educate': 1 explicitly to the configuration of the plugin), it is always off, and I have to run command :Edcuate every time I start new instance of vim (neovim from the master branch of the git, or vim 8.2.0348, but I have observed it for some time already).

I don’t see any error messages anywhere.

Nothing happens when I viq or viQ inside a quoted string.
vim-textobj-user and vim-textobj-quote are installed.
I'm using NeoVim v0.6.1.

My config:

Plug 'kana/vim-textobj-user'
Plug 'preservim/vim-textobj-quote'
" ...

augroup textobj_quote
  autocmd!
  autocmd FileType ruby call textobj#quote#init({'educate': 0})
augroup END

I use vim-surround for manipulating matching pairs of quotes and other syntax. Adding vim-textobj-quote to the mix works great for getting motion commands for curly quoted strings. It also works out of the box to add a surround operator for curly quotes. Changing other surround symbols to curly quotes even works: using "this‸ string", typing cs"q will produce “this‸ string”. The issue I have is that it doesn't work the other direction: starting with “this‸ string” and typing csq" does nothing.

What's going on here? Is this a bug or do I have something setup wrong?

Using https://github.com/FooSoft/md2vim or similar.

When you have 'something' and you do cs'q on it, you do get “something”, but when you have “something” and you do csq' on it, you don't get any change.

Hi, I've noticed that Vim's native spell check doesn't recognize the ’ in words. This sucks.

I've researched several approaches for getting it to recognize it: some included using plugins to use different spell check engines. I thought that was overkill.

So, it turns out there's a better way, create a custom spelling file with the correct entries and get Vim to use it:

1. Open Vim, with spell check enabled
2. Run :spelldump. Remove any comment lines (lines starting with #)
3. Save this file somewhere, e.g. ~/spelldump.txt
4. Replace all the ' characters with ’ characters, e.g. by running grep "'" ~/spelldump.txt | sed "s/'/’/g" >> ~/spelldump_smartquotes.txt
5. Place this file in a location Vim expects, like ~/.vim/spell/en.utf-8.add or ~/.config/nvim/spell/en.utf-8.add (NeoVim). If that file already exists, append to it instead of overwriting it, but the first line (the region line) ought to be at the top of the file, instead of appended at the bottom. If there's already a region line at the top and it's different to the one in the new file, consult the Vim spelling docs to understand the spell format and make the appropriate changes.
6. Start Vim, run :mkspell! ~/.vim/spell/en.utf-8.add
7. Restart Vim

Now Vim should recognize the closing single quote mark in words. Phew.

Perhaps you'd like to add this to the docs.

This plugin is based on US locale, where the opening quotes are in superscript, e.g., “Hello there”. Other locales, e.g., German or Polish, use subscript quotes: „Hello there”.

Could you add support for subscript curly quotes to this plugin? Probably making them interchangeable with superscript ones should be enough.

After doing map <silent> <leader>qc <Plug>ReplaceWithCurly per the README, I tried \qc. Instead of replacing pairs of straight quotes with matching curlies, it instead replaced all straight quotes with right curly quotes.