AlecGhost
commented
5 months ago Hello there. I'm delighted to hear that you are interested in the project. I have no clear roadmap yet, so feel free to request and implement features, that you are interested in. I would lay out the main goals as follows:
- Provide a language-independent platform, to integrate Tree-sitter-based features into VSCode.
- Be an easy-to-use stepping stone for custom/new/small programming languages or DSLs.
- Make embedding/injecting languages into other languages easier.
These goals should also set it apart from the related projects that you mentioned. Specifically, I don't plan on shipping grammars/parsers for established languages with this extension. (Only possibly as examples or tests.)
The next tasks would be the following (I'll create issues for those in the upcoming days):
- Implement the injection rules described in the Tree-sitter docs. I couldn't find a real spec, so I'm not sure how exactly they need to be implemented. But the most important features should be dynamic language injection (opposed to the current @language syntax) and combining the queries instead of performing analysis on each separately.
- Add a configuration option to exclude languages with parsers from the top-level semantic token provider. E. g. I want to inject a language, but I don't want to highlight the files that are only of that language.
- Provide other features than highlighting (e. g. folding).
- Investigate whether the "hacks" described in the documentation are really necessary or if they can be covered by proper VSCode APIs.
Interesting project! I am looking to use/implement/contribute to something like this.
It seems not all the logic is implemented yet. Is that correct? Do you have a roadmap?
There are several other related projects