Yochem van Rosmalen

Thanks! I'll definitely have a look at it. However, one of the goals is to have a language agnostic implementation that does not require new TS queries / captures, and...

As I said earlier, the `b:match_words` format is quite extensive and complex. For example, it supports multiple (nested) backrefs in the middle and end groups. These backrefs are like regular...

I combined the implementations for highlighting pairs and jumping between groups. Now you get highlighting keywords "for free". I don't think that many people want this, so it shouldn't be...

Thanks for the interest and your implementation @monkoose! I haven't worked on this PR because, like you mentioned, fully covering matchit, especially with treesitter, is quite hard and has a...

> There is no debounce (like in default matchparen), so in some cases it could visually slowdown cursor movement in some cases, especially in regex syntax buffers with heavy syntax...

Ah didn't know Vim supports the XDG base dirs as well now. Will do!

Reopened this to keep track if this issue when the Vim team decides to not implement this.

I don't entirely agree. Even though explicitly set by the user, the files contain data: the words that should be ignored. It does not configure Nvim and is not even...

> Also, "data" seems more suitable for potentially private info, whereas it's common to store "config" in a public repo. Spell .add files may contain quasi-private data. (Though XDG doesn't...

> (As long as both directories are searched, people can still decide for themselves by moving files manually.) I don't think I understand, as the directories are not really 'searched'....