specfem::Context struct and ContextGuard for initialization of kokkos, mpi, and handling initialization from python side etc.

[lsawade]

Description

When I was trying out the Kokkos::View extents syntax, I asked claude to create a basic cpp project for me to just play around with the dependencies, and not having to wrestle with specfem compilation.

It ended up creating a core initialization framework to handle initialization and ending of Kokkos.

I created this PR that implements everything. Everything is tested include the Python interface, so no issues there.

Claude's sample implementation (Click me)

namespace specfem {

/**
 * @brief Minimal SPECFEM++ core class for testing dependencies
 * 
 * This class provides a simple interface to test that all major
 * dependencies (Kokkos, YAML-CPP, Boost) are working correctly.
 */
class Core {
public:
    Core();
    ~Core() {if (initialized) finalize()};
    void initialize(int argc, char* argv[]);
    void finalize();
    void print_system_info();
    void run_all_tests(argc, argv);
private:
    bool kokkos_initialized_;
    std::string execution_space_;
};

main function that uses it

int main(int argc, char* argv[]) {
    specfem::Core core;
    
    try {
        core.initialize(argc, argv);
        bool success = core.run_all_tests(argc, argv);

        return success ? 0 : 1;
        
    } catch (const std::exception& e) {
        std::cerr << "Fatal error: " << e.what() << std::endl;
        return 1;
    }
}

What's interesting is that it a allows for a pretty clean single executable without a lot of multiplied code:

running 2d:

specfem 2d -p specfem_config.yaml

running 3d:

specfem 3d -p specfem_config.yaml

Let me know what you think.

Issue Number

If there is an issue created for these changes, link it here

Checklist

Please make sure to check developer documentation on specfem docs.

• I ran the code through pre-commit to check style
• THE DOCUMENTATION BUILDS WITHOUT WARNINGS/ERRORS
• I have added labels to the PR (see right hand side of the PR page)
• My code passes all the integration tests
• I have added sufficient unittests to test my changes
• I have added/updated documentation for the changes I am proposing
• I have updated CMakeLists to ensure my code builds
• My code builds across all platforms

[lsawade]

If approved, this will need additional file edits for the documentation.

[Rohit-Kakodkar]

I am mostly on board with this.

• I think we need a new enumeration that is different from dimension. How about

namespace specfem {
enum class execution_model {
dim2, dim3, globe, ...
}
}

• Use std::unique_ptr where you can if the ptr is owned exclusively by core (which I think is the case). Or use std::shared_ptr with weak_ptr which might be appropriate as well. I recomment seeing Cherno's videos on those if you haven't.

[lsawade]

I am mostly on board with this.

Ok! I'm going to wait for @icui 's review as well before making the requested updates

Use std::unique_ptr where you can if the ptr is owned exclusively by core (which I think is the case). Or use std::shared_ptr with weak_ptr which might be appropriate as well. I recomment seeing Cherno's videos on those if you haven't.

I don't think I watched that one yet, will do!

[icui]

Looks good to me. Shall we rename it to Context since Core is already a folder that includes everything?

[lsawade]

I still have to update the documentation. But I thought since the updates were larger, I still wait for it.

[Rohit-Kakodkar]

Suggested change

	namespace specfem {
	namespace specfem::simulation {

[lsawade]

I merged devel into #1416 by accident, so we just gotta wait for that one to merge into here, for the conflicts to be resolved.