The CMAKE_GUARD for multiple guarding variables requires a
porper boolean statement.

Robert K authored 2 years ago and Andreas Dedner committed 2 years ago

use using and SizeType to size_type to adhere to istl conventions

Notice that this also fixes an incorrect `MultiTypeBlockMatrix::field_type`
in a special setting: If you have a nested matrix
`MultiTypeBlockMatrix< MultiTypeBlockVector<MultiTypeBlockMatrix<...>>>`,
the outer `MultiTypeBlockMatrix::field_type` forwards to the `MultiTypeBlockVector::field_type`
which forwards to the inner `FieldTraits<MultiTypeBlockMatrix>::field_type`.

And fix all spelling errors that codespell currently finds.

... because we also specialize `std::tuple_element` and
it derives from `std::tuple` after all.

Corner case that sometimes happens and produced a segmentation fault.

Unused variable warning indicated this oversight

See https://reuse.software/ for a description.

Nested sparse matrices were not allowed to have empty rows/cols. By choosing the minimum offset of other blocks, we get an educated guess for these cases.

Before this commit, the pattern was deallocated with the wrong size that created a huge memory leak.

The current behavior (allocating each row separately ) is rather
surprising and unnecessary as the matrix market file stores the
number of nonzeros anyway. Hence we now pass this number to
BCRSMatrix::SetSize to get the most efficient behavior

If one lane has zero residual (e.g. the RHS and initial guess is zero) the division
in the factor for the minimizing and orthogonalization is handled explicitly to
avoid NaNs.

add test for block size test for spqr

deactivate std::complex tests of spqr