Lukas Atkinson

@pustekuchen91 Thank you for that example! Which compiler are you using? Constexpr-lambdas are certainly an interesting pattern that I have not seen before, but I'm not quite sure why that...

> How about using the mangled symbol name to identify a function? @Pesa We are not just using the function “name”, which, as you correctly point out, is meaningless in...

In other news, I've been experimenting with supporting different function coverage merging modes, but this is rather tricky since this implies dealing with a set of objects that change their...

@Spacetown Honestly, I don't quite remember. I know I added the `MergeOptions()` concept to make it *possible* to merge otherwise conflicting data. But so far, that code is *almost* bug-compatible...

Thank you for suggesting this. In principle this would be possible. But I don't think this is in-scope for the gcovr project. * PyInstaller is great for end-user oriented applications....

Probably the same issue as in #386. Problem is probably a combination of the following factors. 1. Gcovr doesn't know how to invoke `gcov` and just makes multiple guesses until...

@kcudnik Branch coverage on C++ code is a really tricky subject. Please refer to our FAQ entry *[Why does C++ code have so many uncovered branches?](https://gcovr.com/en/5.0/faq.html#why-does-c-code-have-so-many-uncovered-branches)*. Note that `SWSS_LOG_ENTER()` is...

> Since the methods only make sense with the lines (needed to get the coverage) Oh right. Hmm. This is more difficult than expected. So I think this is blocked...

Thank you, this looks great! There's nothing wrong with the previous intermediate format, but moving to JSON will make future improvements easier (both for GCC and downstream users like gcovr)....

The JaCoCo site has a [DTD schema for download](https://www.jacoco.org/jacoco/trunk/coverage/report.dtd). The format looks simple enough. So creating a new report format might be a good first issue. The Cobertura and Sonarqube...