[1.20.x] Trivial BackgroundScanHandler optimisations

[PaintNinja]

• Comment out three private lists that are updated but never queried, with no getters. Presumably some left over debug code?
• Use .toList() instead of .collect(Collectors.toList()) and !str.isEmpty() instead of str.length() > 0 where appropriate

[Technici4n]

Are there any numbers for these "optimizations"?

[LexManos]

This is more a code cleanup then an Optimization. You save a bit of memory removing the lists.
isEmpty vs length() saves you one rshift operation. So not important
But the toList one actually could be important depending on the stream implementation, could be a lot faster.

HOWEVER, the PROPER solution if you care about performance is to get rid of cpw's over-engineered code that does nothing but waste cycles, as the Filter matches everything , And the ModAnnoation intermediate object could be streamlines a lot. And you can bypass the entire stream if you use a basic for loop to bake the object. Or you used a builder system and baked it in the Annotation visitor's end.

This is one of the many places that I plan on cleaning up in the future when i have the free time to do the full project audit.

[PaintNinja]

Finally figured out profiling and got some numbers. Aside from Lex's comments on these changes, the removal of redundant list operations does make a measurable difference:

Most notably, addCompletedFile() is about 60% faster. Of course, doing larger changes like those suggested would yield in bigger improvements across the board, but considering how trivial these optimisations are, it's not bad.