Right now, the main thing that tree-sitter is doing for us is helping with variable-scoping for things like renames. This is quite useful, but not worth the overhead anymore now that we have a proper tokenizer. Now, obviously there's a gap between a flat set of tokens and a properly parsed tree, but I think there's value is doing this!
Right now, the main thing that tree-sitter is doing for us is helping with variable-scoping for things like renames. This is quite useful, but not worth the overhead anymore now that we have a proper tokenizer. Now, obviously there's a gap between a flat set of tokens and a properly parsed tree, but I think there's value is doing this!