highlight calls to erlang modules as types

[the-mikedavis]

:wave: hello!

I'm integrating the new parser and highlight queries into the helix editor (see helix-editor/helix#830) and I noticed that a function call on an atom highlights differently than an alias (module). This PR adds a clause to the highlights to recognize that case and highlight the atom the same as an alias:

^{before | after}

What do you think?

[jonatanklosko]

Hey @the-mikedavis! I intentionally didn't do it, because I'm not sure if atom should be colored differently based on context. On one side both tokens refer to a module, but in the end atoms and aliases are different beings. It's certainly subjective, but for some reference makup_elixir does the same.

@josevalim any preference?

[josevalim]

I am thinking they could be the same. It is already syntactically distinct so having them visually the same can help drive the point home they are all modules?

[jonatanklosko]

I'm not really opinionated here, so as you two agree it sounds perfectly fine to me!

[the-mikedavis]

I think the consistency with makeup is pretty desirable. I'd be willing to try out a PR to makeup to align the behavior.

here are some more snippets from a real code-base for comparison...

  ^{before | after}

[jonatanklosko]

Ah, actually this matches the behaviour in https://github.com/nvim-treesitter/nvim-treesitter/pull/1904 too :)

[jonatanklosko]

Thanks! :cat:

[connorlay]

Ah, actually this matches the behaviour in nvim-treesitter/nvim-treesitter#1904 too :)

Yep, which was me following what I saw in exdocs snippets, bringing things around full circle 😆