Andy Porter

The benefit is that it captures everything within the `type` which is important because the new metadata handling (that Rupert developed) works by reading and writing the `type` and nothing...

Hugo has also hit this when we have an array specified using unsupported dimension expressions, e.g.: CODE_UNSUPPORTED_FTYPE_SCALAR="""subroutine foo() real*4 :: a(-1:10, 12), cff1 a(1) = cff1 end""" Issue #2138 will...

One of the reasons the DA will refuse to parallelise a loop is if it writes to a scalar. We could use that information to generate the appropriate clauses (possibly)?

Yes, I think so. Could be a part of #2690?

Thanks @JulienRemy. I've set the integration tests going...

Unfortunately both the LFRic integration tests and the compilation tests failed. You should be able to reproduce the latter by running pytest with the `--compile` flag. The error is: ...

@hiker and @sergisiso - we could do with your opinion on how to handle declarations of type `class(*)` - these are "unlimited polymorphic type". My thought is to have these...

Ready for a first look. Probably one for @sergisiso as it adds some SymbolTable functionality, as well as fixing some bugs. I've triggered the integration tests. Note that the sphinx-link-check...

Taking this back as the LFRic and NEMO integration tests failed :-(

Allowing for intrinsics has fixed the LFRic integration tests. However, the NEMO MO passthrough still fails: Executing:psyclone --limit output -api nemo -oalg /dev/null -opsy diatmb.f90 cfgs/MO_GO8_gnu/BLD/ppsrc/nemo/diatmb.f90 ... psyclone.psyir.symbols.symbol.SymbolError: PSyclone SymbolTable...