Run checker only on code comments / strings (feature request)

[gabriel-v]

Love the plugin, thanks for the work!

When writing code with verbose comments and/or in-code documentation (docstrings), it would help to be able to run the tool only on some configurable sections of the file.

I've tried a few things:

• use a macro to select next relevant comment section and run :'<,'>LanguageToolCheck on the selection
  • need a different macro for every comment / string syntax / programming language combination
  • checking some entire file takes a long time (2s / request * 100 comment sections in some file) and is a manual process
• using global like :g/pattern/LanguageToolCheck
  • runs one request for every line
  • but the output buffer is scrapped every time, so it's not usable
• use some external script to replace irrelevant pieces of the file with spaces before running LanguageTool on it
  • replace LanguageTool executable with this script in g:languagetool_cmd
  • didn't actually try this one yet, is this the simplest fix?

On selecting what sections to be checked and what not to be, we pretty much have two options:

• include/exclude list of regex
• include/exclude list of syntax highlight types

With one of these available, the user would be able to set up different rule sets for different programming languages using autocmd FileType.

Edit: I see people trying to do this specifically for LaTeX: https://github.com/dpelle/vim-LanguageTool/pull/4

What do you think?