I was considering supporting coverage reporting for Haskell by first pre-processing its native tix format to gcov, then using cc-test-reporter normally. To understand what I'd need to do, I looked at your gcov parser, it seems broken:
If I'm reading this correctly, it uses the gcov file name as the SourceFile name. Shouldn't it use the actual name of the source (e.g. examples/Calculator.swift in that test)?
I was considering supporting coverage reporting for Haskell by first pre-processing its native tix format to gcov, then using
cc-test-reporter
normally. To understand what I'd need to do, I looked at your gcov parser, it seems broken:https://github.com/codeclimate/test-reporter/blob/master/formatters/gcov/gcov_test.go#L26
If I'm reading this correctly, it uses the gcov file name as the SourceFile name. Shouldn't it use the actual name of the source (e.g.
examples/Calculator.swift
in that test)?If I'm correct, please also note you can't just
s/.gcov$//
to fix this, you have to read the filename out of the data itself: https://github.com/codeclimate/test-reporter/blob/master/formatters/gcov/examples/Calculator.swift.gcov#L1