Performance

[Jolanrensen]

Running the processor with Gradle can take a long while in a project the size of DataFrame. This can be felt most using the @include and arg processors, mainly, probably, because they're used the most in the project.

Some performance steps need to be researched.

[Jolanrensen]

#43 fixes a lot. For DataFrame it halves the build time!

Still, I think there's some potential speed to gain in the ArgDocProcessor; currently, all @set tags are processed before @get tags, but this happens in such a way that if a @get tag is found while @set tags are still present, it's returned without change, but still processed.

[Jolanrensen]

Now there might be some more time saving in caching the builds with outputs.uptToDateWhen {}.
Currently, they are rerun every build, but they only need to be updated when:

• the input has changed
• the output has changed (if we make the output Read only, we only need to detect file deletions

Partial updates could also be done, but that would require more research.

[Jolanrensen]

Also, ArgDocProcessor could run in parallel :)