I just took the parsers from the first group plus python and tried them against our unit tests to see what happens. Fortran seems to produce different output in many situations so I dropped it for now. The asm parser uses the cpreprocessor.c/h that the new cxx parser uses and depends on some things from the new cxx parser so I dropped asm from now too. Finally, the new python parser generates different tags for imports and we should make some modifications in Geany to adjust for that so it's dropped for now too.
The rest of the parsers seem to work fine and the updates in unit tests seem to do the right thing. I pushed the updates to unit tests language by language including comments about the changes to simplify review.
I can of course drop some parser from this pull request if someone knows some updates were made to our parser that should go upstream first.
To summarize, this is the result of this pull request:
Parsers included here: - basic - diff - go - html - jscript - lua - make - objc - pascal - perl - php - ruby - rust - sql - nsis
Different parsers (longer-term stuff): - iniconf - matlab - r - sh - tcl - tex - verilog - vhdl - fortran - c.c
Parsers for which I will make separate PRs soon: - python (needs some more adjustments) - cxx - asm (depends on cxx)