Right now each provider typically grabs the file, parses it, runs some analyses (UDF finder, sym finder, etc) each on their own. This results in significant overhead due to lots of duplicated parsing/analysis.
We should have a system in place that caches the tree and various analyses and invalidates that cache when the document's version changes.
Right now each provider typically grabs the file, parses it, runs some analyses (UDF finder, sym finder, etc) each on their own. This results in significant overhead due to lots of duplicated parsing/analysis.
We should have a system in place that caches the tree and various analyses and invalidates that cache when the document's version changes.