uni-algo VS tomlplusplus

Safe layer is just bounds checks that work in all cases that I need, before that I was coping with -D_GLIBCXX_DEBUG (doesn't have safe iterators for std::string and std::string_view and that I need the most) and MSVC debug iterators (better but slow as hell in debug). You can read more about the implementation here: https://github.com/uni-algo/uni-algo/blob/main/doc/SAFE_LAYER.md Nothing interesting it's possible to implement all of this even in C++98 but no one cared back then and it's a shame that it's not in C++ standard so we cannot choose to use safe or unsafe std::string for example and must rely on implementations in compilers that are simply incomplete in many cases or implement it from scratch.

Why call your files modules? "Modular programming" or "Modular architecture" is pretty standard term in programming I don't think there are a good synonym for "module" word so I plan to use "Modules" and "C++20 Modules" to avoid ambiguity. You have one cpp file in the project. Any reason for that? Including Unicode data files that may be pretty big into header files hurts compilation speed. C++20 Modules will help with that and I plan to support it. I'd also recommend supporting CMake I will support CMake it just I didn't need it in my development stage. Overall, looks very nice. Thank you. I made a post on GitHub with some of my plans: https://github.com/uni-algo/uni-algo/issues/3

Unless you want to make your own config file parser as an exercise (which is a good idea) I'd recommend using toml++.

It's a public repo - this one: https://github.com/marzer/tomlplusplus but thank you

For a live demo, see the toml++ homepage: https://marzer.github.io/tomlplusplus/

Added bonus, tomlplusplus is a solid C++ implementation, actively updated, and easy to drop into a project (header only, CMake, vcpkg, Conan, etc.). Very easy to use for what I've needed.

Key links: - Homepage: https://marzer.github.io/tomlplusplus/ - Repository: https://github.com/marzer/tomlplusplus - "What is TOML?": https://toml.io/

cmake_minimum_required(VERSION 3.23) project(testProj) set(CMAKE_CXX_STANDARD 23) set(FETCHCONTENT_BASE_DIR ${PROJECT_SOURCE_DIR}/${CMAKE_PROJECT_NAME}/libs CACHE PATH "Missing description." FORCE) include(FetchContent) FetchContent_Declare( tomlplusplus GIT_REPOSITORY https://github.com/marzer/tomlplusplus.git GIT_TAG v3.1.0 ) FetchContent_MakeAvailable(tomlplusplus) add_executable(testProj testProj/src/main.cpp) target_link_libraries(testProj tomlplusplus)

I use them in toml++ to store the path of the source document for all the data nodes parsed from that document. Storing it as a separate std::string allocation for every single object would have been very wasteful, and the alternative would have been to make an overall document class that owned everything, which I didn't want to do, so shared_ptr was a good compromise.

Key links: - Homepage: https://marzer.github.io/tomlplusplus/ - Repository: https://github.com/marzer/tomlplusplus - "What is TOML?": https://toml.io/

toml++ was both my first 'public'/open-source project, and my first ever attempt at a parser. There's a few things I'd do differently now if I were to write it from scratch, but I've learned a lot along the way (particularly about making things build 'clean' on multiple platforms).