Add support for std::span (C++20)

[AbitTheGray]

Main goal of this Pull Request is to add support for converting glm::vec, glm::mat and glm::quat to <span>'s std::span but there are few more changes. The goal is to have a a single non-owning object to access underlying data (similar to std::string_view).

Changes

• New Experimental glm/gtx/span.hpp
  • std::span for C++20 (tested with __cpp_lib_span)
  • std::mdspan for C++23 (tested with __cpp_lib_mdspan) - untested as GCC does not support it right now (see C++23 support list)
  • std::valarray to have something for older C++ versions, but it copies the underlying data
  • Tests (test/gtx/gtx_span.cpp) inspired by test/gtx/gtx_range.cpp
• Update glm/gtx/range.hpp to use glm::type from glm/gtx/type_trait.hpp
  • Mark functions as constexpr and [[nodiscard]]
  • Use glm::type to get number of elements instead if calculating it just for ranges
  • Add support for quaternions (they may have worked before using the matrix fallback but it would be incorrect)
  • Expand tests (test/gtx/gtx_range.cpp) to include more types
• Make glm::type from glm/gtx/type_trait.hpp easier to use
  • All specializations have all fields, including new elements (calculated as cols * rows for convenience)
    • glm::type<glm::vec<...>> did not have cols and rows
    • glm::type<glm::mat<...>> still has componenets=cols, which is the reason for elements
      • Watch out for glm::components<T>(T const&) from glm/gtx/range.hpp as it returns glm::type<T>::elements (which is previous behaviour)
  • Checks for const, volatile and & in type provided to glm::type<...> as those would use the fallback zeroes
    • glm::type<glm::vec3> returns correct values but glm::type<const glm::vec3> does not
      • Use can get the const there by using decltype like const glm::vec3 value{}; const auto components = glm::type<decltype(value)>::components; assert(components == 3);
• CMake option GLM_ENABLE_CXX_23
  • Defines GLM_FORCE_CXX23 but otherwise works like GLM_ENABLE_CXX_20
  • Also added GLM_LANG_CXX23_FLAG and GLM_LANG_CXX23 in glm/detail/setup.hpp

I was also thinking about adding concepts but it would not work for older C++ versions. It may be something for future Pull Request as you can use it like void someFunc(glm::concept::vec auto v) where glm::vec3 and glm::i32vec3 would work but not glm::vec2.

Missing or problems

• std::span is used as std::span<T> (with dynamic extent) while it may be std::span<T,N>
  • This is something I forgot about but is a good idea to add before merging this as it would allow libraries to have std::span<float, 3> as an argument and pass glm::vec3 but not glm::vec2
• Untested with compilers other than GCC 14
  • I've just tested it with different GLM_ENABLE_CXX_?? values so hopefully C++ versions are covered
  • Did not try to run MSVC or CLang
• There are still TODOs in std::span test (test/gtx/gtx_span.cpp) - there are no tests for std::valarray and std::mdspan
  • I will add those based on feedback on this MR (just did not want to waste more time in case there is a problem with my proposal to add the std::span support)
• glm::type should probably handle the const, volatile and references correctly without the assertion, but it does not do it now so I feel like it should at least tell you
  • It could be hidden behind #ifndef in case someone depends on the unintuitive behavior
    • Right now it has only C++11 check (as static_assert and other used features did not exist before)

DISCLAIMER

This is my first Pull Request to project this big and I am proposing new functionality that, as far as I know, was not requested. There were some guesses from me about how some thing should be done.
I've tried to make it compatible for any compiler and C++ version but I have little experience there as well.

I hope there is no problem with me also changing the glm/gtx/range.hpp header (and its test) as I feel it is easier to read now.

No, this is not a "school project". It is something I found is missing in glm and took it as a good reason to learn more about glm's internals.