Snippets!

[clason]

What?
What it says on the tin: a pure Lua snippet engine.

Why?
The major snippet plugins (UltiSnips, neosnippets) are remote python plugins, which adds a dependency. There's also https://github.com/hrsh7th/vim-vsnip in pure vimscript, though, but Lua should have significantly better performance and be more hackable. It should support the basic features: snippet management (either in its own format, UltiSnip/neosnippet format, or LSP); snippet expansion (including nested placeholders) and ideally transformation.

Potential existing implementations:
https://github.com/norcalli/snippets.nvim, which is useable but still lacks a few of the above features.

Potential pitfalls:
Ideally, the expansion should be based on extmarks, but that requires the PR https://github.com/neovim/neovim/pull/12249/files to be merged first.

[clason]

I think snippets.nvim deserves being moved to the "In Progress" tab :)

[wbthomason]

Good point; I've updated the state of all the ideas for which we had a documented partial/in-progress implementation.

[clason]

Actually, snippets.nvim is arguably full-featured enough now to merit being put into the "Done" category 👍

(Otherwise it'd be worth specifying what precisely counts as a "Done" plugin...)

[wbthomason]

Cool, sounds good.

[smolck]

(Otherwise it'd be worth specifying what precisely counts as a "Done" plugin...)

That should probably be done regardless, right?

[wbthomason]

Yeah, that's a good idea. I'd suggest something like "A wish is 'granted' when there exists a feature-complete, stable plugin implementing the wish's idea". snippets.nvim calls itself beta-quality software, so it's not clear that it would meet the "stable" criterion there, but in practice it has seemed stable and feature-complete.

Does that seem reasonable?

[clason]

Yeah, that makes sense. (The readme isn't up-to-date, but @norcalli called it basically done on Gitter today.)

[runiq]

It has a wiki now, too.