language server todos tracking list

[alex-s168]

• Incremental compilation #1254
• make the parser continue parsing on errors. important for when you are typing. this can be done by emitting error AST nodes instead of failing, and then skipping tokens. for example [a, b c] will generate a collection expression, but b c will generate a error node instead, but it will continue parsing because of nesting
• incremental parsing and tokenization: when typing more letters of an identifier or number, it shouldn't re-parse everything
• json parser for lib #1319 (for LSP protocol)
• base of the language server: incrementally re-parse and re-typecheck the inputted code and provide diagnostics
• feat: auto-complete
• feat: go-to-definition
• feat: find usages (also trough docs)
• feat: semantic highlighting
• feat: render documentation of function reference on-request
• action: extract to separate function
• action: wrap in Some { }
• action: generate remaining match arms

making a LS is hard, but it will make the language way more attractive to people on the outside

[alex-s168]

@andrew-johnson-4 are you okay with making the LSTS parser a lot more complicated for us to be able to use it in an language server?