Issue 278 - Refactor of domain folder.

[Rohit-Kakodkar]

Description

Sorry for the big PR 😣

This PR refactors the domain folder.

• Promotes element_types from impl namespace to specfem::compute namespace
  • get_element_on_device functions are now part of element_types
• Refactor kokkos_kernels so that the functions are using the macros
• promote boundary conditions from specfem::domain::impl namespace to specfem::boundary_conditions
• Delete domain folder and its namespace

Issue Number

closes #278

Checklist

Please make sure to check developer documentation on specfem docs.

• I ran the code through pre-commit to check style
• 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]

Sorry for the big PR 😣

LOL this a big boi

[icui]

How do you decide between using file name include/compute/element_types/element_types.hpp and include/compute/compute_element_types.hpp?

[Rohit-Kakodkar]

It should always be element_types/element_types.hpp imho. I initially started naming files compute_xxx.hpp/cpp thinking that compute would be a small struct. Now its grown much more detailed.

[icui]

Looks like there a few filed that are no longer needed in cmake:
src

• algorithms/interpolate.cpp
• mesh/mpi_interfaces/mpi_interfaces.cpp
• parameter_parser/solver/solver.cpp
• parameter_parser/time_scheme/solver.cpp
• point/boundaries.cpp
• source/utilities.cpp
• source_time_function/utilities.cpp
• utilities/utilities.cpp

include

• compute/properties/properties.tpp

[icui]

Otherwise LGTM

[lsawade]

I have nothing to add to congyue's remarks, except the minor formatting issue.

I think we need a testing hackathon. This behemoth of a PR would be less daunting if we had more coverage (mildly motivated by finding seismogram iterator bug earlier today).

Also, it is wild to me that this PR cuts 2700 lines.

[lsawade]

clang-format being a butt

[lsawade]

This i nice, no need to loop over all elements no moe

[Rohit-Kakodkar]

I have nothing to add to congyue's remarks, except the minor formatting issue.

I think we need a testing hackathon. This behemoth of a PR would be less daunting if we had more coverage (mildly motivated by finding seismogram iterator bug earlier today).

Also, it is wild to me that this PR cuts 2700 lines.

A testing hackathon or a testing spring would be great in the future. We definitely need some of that.

[Rohit-Kakodkar]

Looks like there a few filed that are no longer needed in cmake: src

• algorithms/interpolate.cpp
• mesh/mpi_interfaces/mpi_interfaces.cpp
• parameter_parser/solver/solver.cpp
• parameter_parser/time_scheme/solver.cpp
• point/boundaries.cpp
• source/utilities.cpp
• source_time_function/utilities.cpp
• utilities/utilities.cpp

include

• compute/properties/properties.tpp

Fixed