Compilation database / compile_commands.json

[GPMueller]

TODO:

• [x] output compile_commands.json, with separate commands for the separate compile-stages (depfile and objfile)
• [x] reasonable performance, don't constantly re-parse the json
• [x] test (including printing to console)
• [x] mention the feature in the docs

Maybe

• [ ] update database before actually calling the commands
• [ ] find a different pattern than requiring a set of three return values from the Toolchain.generate_dependency_file and Toolchain.compile functions
• [ ] find a way to "prune" or "clean" the database of entries that are no longer needed or relevant but which will currently hurt performance

Closes #109.

[codecov[bot]]

Codecov Report

Merging #130 (60ebaa0) into master (50e872b) will increase coverage by 0.98%. The diff coverage is 77.41%.

@@            Coverage Diff             @@
##           master     #130      +/-   ##
==========================================
+ Coverage   78.58%   79.56%   +0.98%     
==========================================
  Files          17       17              
  Lines        1172     1199      +27     
==========================================
+ Hits          921      954      +33     
+ Misses        251      245       -6     

Impacted Files	Coverage Δ
clang_build/single_source.py	62.12% <63.15%> (+7.01%)	:arrow_up:
clang_build/environment.py	90.32% <100.00%> (+1.86%)	:arrow_up:
clang_build/target.py	80.51% <100.00%> (+1.50%)	:arrow_up:
clang_build/toolchain.py	82.65% <100.00%> (+5.56%)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 50e872b...60ebaa0. Read the comment docs.

[GPMueller]

I don't know why codecov thinks the coverage declined, as all code I changed is obviously executed in the tests (it is inside the Environment constructor and SingleSources generate_depfile and compile)...

[GPMueller]

Ok looks like the incorrect coverage may be due to https://github.com/pytest-dev/pytest-cov/issues/424 - not sure what to do about it, maybe just hope it disappears on its own over time?