Open tdhock opened 6 months ago
Or is there a reason why code coverage for C files is not interesting?
No there is not.
Related for cython #25600
Our data also suggests (though we're not sure) the code may be considerably less covered than much of the rest, perhaps due to the lack of visibility in part. That's because the mutation score here is a lot lower (~50% vs. 70%+) than for the rest of the code.
Hi, I am doing mutation testing of pandas with @agroce
I see five C source code files that are maintained by pandas, is that right?
I expected that pandas devs should be interested in the code coverage of these files, but on current coverage reports https://app.codecov.io/gh/pandas-dev/pandas/tree/main/pandas?search=&displayType=list there is no coverage computed for C files (only for py files).
How can the build/test scripts be modified so that C code coverage could be computed?
Or is there a reason why code coverage for C files is not interesting?
I searched the issue tracker for similar keywords, and the only related issue I saw is about coverage for Cython: https://github.com/pandas-dev/pandas/pull/54453
Thanks in advance.