[dustinswales]

Script (from Claude) to parse the Capgen generated datatable (.xml).

The script can provide the following information:

• List every subroutine call that touches a standard_name, in file order
• Trace a standard_name in actual execution order for one suite
• Trace local (Fortran) variable name used somewhere -- it will be resolved to the standard_name(s) that share it
• List available suites/groups/schemes

User interface changes? No

Addresses #784

Testing:
test removed:
unit tests:
system tests:
manual testing:

[climbfuji]

Quick first review, didn't try it yet.

A more general question I have is if the variable tracker should be "part" of capgen v1, i.e. live in scripts and utilize available tools to access the datatable (piggy-backing on the datatable query script). It would be a fourth entry point.

Right now, it is more like an external tool with its own datatable parser and its own directory (tools).

I am leaning towards a tigher integration, given that the script comes with the CCPP capgen v1 code.

Thoughts?

We also need CI testing for the script.

[dustinswales]

Quick first review, didn't try it yet.

A more general question I have is if the variable tracker should be "part" of capgen v1, i.e. live in scripts and utilize available tools to access the datatable (piggy-backing on the datatable query script). It would be a fourth entry point.

Right now, it is more like an external tool with its own datatable parser and its own directory (tools).

I am leaning towards a tigher integration, given that the script comes with the CCPP capgen v1 code.

Thoughts?

We also need CI testing for the script.

I like the idea of this script being an entrypoint.
I will move this to DRAFT and work on it tomorrow, along with some CI.