JSON line information output from .dat coverage

[dariuskia]

I've been using bazel coverage to generate a combined coverage report .dat file, and then genhtml to produce the viewable report. However, I'd like to have a parsed output of the lines that were hit for each test target, e.g. in JSON like gcov's functionality. My intent is to produce a dataset of mappings between test targets and a structured (machine-readable) view of the lines that were hit. Is this possible?

According to this, lcov automatically erases the intermediate .gcov files, so gcov -j isn't an option. The most information I've been able to get is from lcov -l; however, this doesn't provide the exact list of lines that were hit.

[henry2cox]

I don't think I understand what you are looking for - though it sounds as if you want to use lcov as a driver - to run gcov for you.
If that is the case: you lcov --preserve --capture ... might do something similar.
You probably also want to provide the temp directory - so you don't get a random one in /tmp.

More generally: if your actual final application worked perfectly: what did it do?
Not the data collection part - the actual application.
There may be another way or an easier way to accomplish your goal.