@bcoconni Done
std::toupper is overloaded so there is the need to help the compiler choose the correct one.
See https://stackoverflow.com/questions/7131858/stdtransform-and-toupper-no-matching-function

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (af7910e) 24.87% compared to head (71d7cf1) 24.87%.
Report is 3 commits behind head on master.

@@           Coverage Diff           @@
##           master    #1020   +/-   ##
=======================================
  Coverage   24.87%   24.87%           
=======================================
  Files         168      168           
  Lines       18908    18908           
=======================================
  Hits         4704     4704           
  Misses      14204    14204           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

I reran the github build process and no compiler errors now across all compilers and platforms.

Out of interest what are you using JSBSim for and is there a particular need to use the fairly old MSVC 2015 compiler?

Awesome 😎
MSVC 14.0 (2015) is the minimum supported compiler version (ABI compatibility), therefore I started with it.
JSBSim is fully compatible with MSVC 14.0 except for this minor fix (I've created a PR for others).

After you added the #include <cctype> did the following original code making use of a lambda not compile with MSVC 2015?

        std::transform(str.begin(), str.end(), str.begin(),
                       [](unsigned char c){ return std::toupper(c); }
                      );

After you added the #include <cctype> did the following original code making use of a lambda not compile with MSVC 2015?

        std::transform(str.begin(), str.end(), str.begin(),
                       [](unsigned char c){ return std::toupper(c); }
                      );

I was about to make the exact same point: the current code in the master branch just works. It's fine to add #include <cctype> if it is needed for VS2015. But I'm struggling to see the point in replacing a lambda function by a function pointer conversion. Both of them look equally ugly to me so let's stick with the current code.

And if it's done for the sake of C++ pedantry then I'll argue that you must use static_cast<> for type conversion 😉

Done.
It works! 🥳🤯
Let's see the workflows...