Polish: inlay hints and codelens performance #634
Conversation
aspeddro
changed the title
|
Having a look at this soon, sorry for the wait! |
|
@aspeddro we don't have any real benchmarks unfortunately. When I've tested I've clocked how long refreshing takes for us in that large file vs TS in a large TS file. |
There was a problem hiding this comment.
I think this looks good. @cristianoc might want to have a glance. @aspeddro mind pulling in the latest master too? There's a conflict.
analysis/src/Hint.ml
Outdated
| }; | ||
| }) | ||
| | _ -> None)) | ||
| match Cmt.fullFromPath ~path with |
There was a problem hiding this comment.
I think this is great. Question for the future: does it make sense to have some sort of cache inside of Cmt.fullFromPath directly?
There was a problem hiding this comment.
The change looks good. Perhaps it's possible to rename fullFromPath to something that makes it clear it's an expensive operation. E.g. loadFullCmtFromPath.
Cmt.fullFromPathfor each hint.A noticeable change to
analysis/examples/large-project/src/res_core.resIs there any way to measure?
Part of #511
cc @zth