Legacy boundary conditions

[malamast]

Added a legacy implementation of 1st order Dirichlet and Neumann BCs.

• This is what we used to do in Bout++ version 3.
• The default BC remain unchanged.
• Higher order extrapolation can be unstable sometimes when using C4 so I added back the 1st order BC.
• You can use them in the input file by setting e.g. bndry_all = dirichlet_o1.
• Implementation for staggered grids is not done.

[github-actions]

clang-tidy made some suggestions

There were too many comments to post at once. Showing the first 25 out of 54. Check the log or trigger a new build to see more.

[dschwoerer]

* Implementation for staggered grids is not done.

Could you add an assert?

Also, please fix the clang-tidy and clang-format issues.

[malamast]

* Implementation for staggered grids is not done.

Could you add an assert?

Also, please fix the clang-tidy and clang-format issues.

I have this line here:

BOUT-dev/src/mesh/boundary_standard.cxx

Line 154 in a960447

throw BoutException("dirichlet_o1 BC is not implementated for staggered grids.");

throw BoutException("neumann_o1 BC is not implementated for staggered grids.");
That throws an Exception if this BC is used with staggered grids. Does this BoutException work as an error message or a an assert statement?

Also, I used: git clang-format origin/next for the formatting. Is there another command I sue use for clang-tidy?

[github-actions]

clang-tidy made some suggestions

There were too many comments to post at once. Showing the first 25 out of 39. Check the log or trigger a new build to see more.

[dschwoerer]

I have this line here:

BOUT-dev/src/mesh/boundary_standard.cxx

Line 154 in a960447

throw BoutException("dirichlet_o1 BC is not implementated for staggered grids.");

That is sufficient. However, it would be better to check whether the fields location is at CELL_CENTRE.
Running with staggered grids is not an issue for this operator, as long as it is only used for fields at CELL_CENTRE.

Also, I used: git clang-format origin/next for the formatting. Is there another command I sue use for clang-tidy?

Formatting is now fine. For clang-tidy you see the comments that it posts. Some of them would require a refactoring, which is probably not in scope for this PR, like this:

warning: returning a newly created resource of type 'BoundaryOp *' or 'gsl::owner<>' from a function whose return type is not 'gsl::owner<>' [cppcoreguidelines-owning-memory]

[github-actions]

clang-tidy made some suggestions

[github-actions]

clang-tidy made some suggestions

[dschwoerer]

This looks decent. The only question is whether we want to add more stuff to boundary_standard