Michael Förderer
Michael Förderer
You're right, this part is commented in https://github.com/gcovr/gcovr/pull/1059#issuecomment-2631549943. For now this should be done in a separate PR. I'll check on it.
I'm on it. Is here `signature="() const"` correct?
@Pesa Please can you check the linked PR?
> Btw what does "gcovr could not infer a working directory that resolved it" mean? The merge errors are catched and added to the default message. This doesn't make sense...
Both seems to be constructors but we only normalise at the end: _ZN3ndn4util14IndentedStream _C1_ ERSoSt17basic_string_viewIcSt11char_traitsIcEE _ZN3ndn4util14IndentedStream _C2_ ERSoSt17basic_string_viewIcSt11char_traitsIcEE Maybe we should use the mangled name from GCOV and always detangle...
That was the next idea. If we have a demangled name use the first mangled name when merged. If we do not have a demangled name we check that the...
That was intended to merge them already in the data model but now that you're asking the lines are still present for all mangled names and now the function definition...
@Pesa Do we need to merge the functions or is it also ok to add the mangled name to the signature to have unique entries.
In data model and JSON I would revert and keep it separate. In Cobertura and HTML I'm working on merging the data but it's quite complicated to create a merged...
@mikedld I'm still working on it. If you've problems this problem you can downgrade to `7.2`. Is it possible for you to test the PR branch after I push my...