Generic enrichment process

[reuterbal]

Another lift of functionality from the Scheduler overhaul.

This one adds a more generic enrichment process. It is anchored on the following concepts:

1. For anything that is imported via a module import (i.e. USE), it is sufficient to look at the import nodes and update the symbol table entries of the symbols listed there (or for all symbols from the remote module if no only list is given). This applies to both, modules and subroutines (we haven't done enrichment for the first before) and also gives us typedef and global variable enrichment.
2. For anything that inherits its type from a parent (e.g., import in the parent module, call in the module procedure), we can compare symbol table entries with the parent's symbol table
3. For Subroutines (the only ones where CallStatements can appear), we additionally apply the previous enrichment via call statements to
   a. retain the active property and
   b. enrich procedure symbols that are made available without an import (i.e., via explicit interface or non-inlined interface block include)

[github-actions]

Documentation for this branch can be viewed at https://sites.ecmwf.int/docs/loki/189/index.html

[codecov]

Codecov Report

Attention: 3 lines in your changes are missing coverage. Please review.

Comparison is base (a851213) 92.16% compared to head (be09a3f) 92.21%.
Report is 4 commits behind head on main.

Files	Patch %	Lines
loki/program_unit.py	95.00%	2 Missing ⚠️
loki/bulk/scheduler.py	93.75%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #189      +/-   ##
==========================================
+ Coverage   92.16%   92.21%   +0.04%     
==========================================
  Files          93       93              
  Lines       16792    16835      +43     
==========================================
+ Hits        15477    15524      +47     
+ Misses       1315     1311       -4     

Flag	Coverage Δ
lint_rules	96.22% <ø> (ø)
loki	92.18% <96.34%> (+0.05%)	⬆️
transformations	91.43% <ø> (ø)

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.

[mlange05]

Very neat and very useful. It also seems to "fix" some odd performance regression in large batch processing scenarios (ec-physics regression test). I left one small request to increase the log-level to capture this in the future, but otherwise this looks great and is GTG from my side. 👏 🙏

[mlange05]

During regression testing I noticed that this has previously become somewhat expensive, but this PR seems to remedy this. Could you please increase the log level here to info, so that any eventual runtime increase gets flagged by default?

[reuterbal]

Thanks, good point. Fixed.