Viktor Kirilov

My original plan was indeed to use `abi::__cxa_demangle` but I avoided it because it required additional header includes in the interface part of doctest which is a no-no, but now...

I'd be interested in some numbers before considering such a change because moving all function definitions out of the interface part of the header was done explicitly in order to...

I used [these scripts](https://github.com/onqtam/doctest/tree/master/scripts/bench) to generate [the benchmarking page](https://github.com/onqtam/doctest/blob/master/doc/markdown/benchmarks.md) but the scripts are pretty horrible (python 2 vs 3 problems, etc) and the last time I ran the benchmarks locally...

Well... while I understand where you're coming from, doctest is designed so that it could be [embedded in production code](https://blog.jetbrains.com/rscpp/better-ways-testing-with-doctest/) & be used alongside it and that's why it's not...

One thing you could easily do is the following (if you provide your [own `main()`](https://github.com/onqtam/doctest/blob/master/doc/markdown/main.md)): ```c++ std::string test_cases_filter = ""; for(int i = 1; i < argc; ++i) if(argv[i][0] !=...

Actually now that I think about it what you are saying does make sense - this could indeed be the behavior when using `DOCTEST_CONFIG_IMPLEMENT_WITH_MAIN`. However with doctest I tend to...

I think Catch2 has something like this. It's a bit weird to be that reporting on runtime is desired if it already passed at compile time... I'll leave this open...

I'd advise using the #define locally for now (perhaps make a proxy doctest header which includes the actual one and defines extra struff) and I might add this in the...

From a brief glance I see that ```doctest_discover_tests``` uses ```--list-test-cases``` - doctest also has ```--list-test-suites```. Or perhaps there needs to be a new option to extract the names of the...

What you described should work :)