Expressions: Handle intrinsic function calls

[mlange05]

This PR adds handling of intrinsic function calls by adding an is_intrinsic property to ProcedureType and adding the procedure type to the parsing scope in frontends before creating the respective ProcedureSymbol.

We also add the new is_intrinsic flag to the symbol attributes table for quick lookup in scheduler items, as we never want to chase intrinsics.

EDIT: This also needs the fix from #439 and needs a rebase once that is merged.

[github-actions]

Documentation for this branch can be viewed at https://sites.ecmwf.int/docs/loki/416/index.html

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.32%. Comparing base (ebc7b69) to head (bd243ac).
Report is 10 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #416   +/-   ##
=======================================
  Coverage   93.31%   93.32%           
=======================================
  Files         212      212           
  Lines       40728    40769   +41     
=======================================
+ Hits        38007    38047   +40     
- Misses       2721     2722    +1     

Flag	Coverage Δ
lint_rules	96.39% <ø> (ø)
loki	93.28% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

---

🚨 Try these New Features:

• Flaky Tests Detection - Detect and resolve failed and flaky tests

[reuterbal]

Thanks, this looks great. One thing I'm missing here is the situation when intrinsic functions are hidden because of user-defined functions with the same name. This is unfortunately allowed by the standard and used in practice...

[reuterbal]

Do we somehow verify here that this is indeed the intrinsic function that is used, and not a locally defined function that happens to be called the same as an intrinsic function?

A concrete example is DOT_PRODUCT, where an implementation in IFS shadows the intrinsic Fortran function. In that case, the procedure symbol should likely not be classified as intrinsic.

It would probably be good to have a test that covers such shadowing.

[mlange05]

Hi @reuterbal , apologies, I rebased over main to sanity reasons. However, the response to your comment is in the second to last commit. With the top commit I piggy-backed a mitigation fix for ab upstream version incompatibility with last night's pydantic release. Please let me know if you'd like this to be a separate PR.

[reuterbal]

Many thanks, good to know about the limitation with shadowing in OMNI!
Thanks also for the pydantic hotfix.