Daan Schipper

I've done some more investigation into this issue. The `locateDependenciesMain` task does not take the capabilities of a project dependency into account when determining all declared dependencies. Some information about...

I'm using something like ``, not ideal as you have to check in the jar in git but it works.

For anyone interested, this is how you can add the dependencies in the version catalog: ``` dependencyAnalysis { dependencies { // Adds the defined aliases in the version catalog to...

I'm currently investigating the problem while parsing many files with JavaParser. I'm running it against a project with ~290 source roots and ~20k .java files. My findings are: - The...

Fortunately the cache only take up memory. As you say these caches are needed to solve lambdas, but I have the idea many entries are not needed. I think it...

Another question, `SourceRoot` provides a way to parallel parse the individual `CompilationUnit`s, but I'm constantly asking `expression.calculateResolvedType`. To my knowledge there only exist one `TypeSolver` and `JavaParserFacade` instance, does this...

Regarding the threading, I'm only visiting the individual nodes and for certain conditions I'm calling `expression.resolve` or `expression.calculateResolvedType` and saving them in a db. I know for certain that the...

I'm now trying to profile the sequential one. If I really want to use the parallel parsing I'll have to implement some queue of nodes that JavaParserFacade can resolve.

@ftomassetti On your question, VisualVM did not list the method parameters in the sampling mode. I'm now running it with 'Profiler' and that does show the method parameters

Apparently you can't invoke JavaParserFacade in parallel, so I'm now running everything in serial, which seems to perform a lot better. However, it's still not enough. I've tried to optimise...