Custom statement_range() LSP message
#85
Conversation
| #[macro_export] | ||
| macro_rules! backend_trace { |
There was a problem hiding this comment.
Does it seem reasonable to do this? Is there anything else I need to do besides this tag?
| // TODO: Uncommenting this causes a compile error??? | ||
| // backend_trace!(self, "statement_range(): No child associated with point."); |
There was a problem hiding this comment.
@kevinushey I think we should talk about this and see if the backend_trace!() macro has a bug in it somehow
There was a problem hiding this comment.
It seems to work above; maybe the macro doesn't handle invocations lacking any arguments?
There was a problem hiding this comment.
Or perhaps it's because logging here requires a call to .await and that doesn't play nicely when you have treesitter nodes on the stack?
There was a problem hiding this comment.
It has something to do with the latter. If i put the message before the creation of root then everything is fine
There was a problem hiding this comment.
The exact error is this horrible thing
There was a problem hiding this comment.
That thing gives C++ template compilation errors a run for its money 💀
We could probably adjust the macro to either avoid the await, or spawn a separate execution context wherein it could be awaited safely and independently from the current call... not sure if you have any strong feelings.
There was a problem hiding this comment.
Definitely no strong feelings
DavisVaughan
changed the title
statement_range() implementationstatement_range() LSP message
Joint with posit-dev/positron#1227 - see there for full details
goto_first_child_for_point()doesn't work as expected, see tree-sitter/tree-sitter#2012, so I'm using a Rust implementation of the patch they used in Emacs tree-sitter/tree-sitter#2012 (comment)