Since you've clearly familiar with the TreeSitter parsing approach, can you briefly explain how TreeSitter works, how it compares to existing VSCode Rust parser, and how it compares to IntelliJ Rust parser?
I'm at times very happy and at times not so happy with IntelliJ's Rust parser. If TreeSitter makes it easier to hack on the parser itself, it'd be quite cool.
I'm far from really knowing the subject deeply. What I know though, is that VSCode hightlighter is using TextMate grammar to describe the language to parse, and it seems to be regex based.
Tree-sitter, on the other hand, is semantic based, so I believe it parses the code and construct some kind of abstract representation of it in order to assign colours.
Anyone please feel free to correct me if I'm saying bad stuff here