Inventory of configuration files requiring parser and updater support

[minghangli-uni]

To ensure the experiment generator supports all relevant configuration files across all model types (ACCESS-OM2, ACCESS-OM3, ESM1.5/1.6), we need a definitive inventory of files that require:

• Parser - the ability to read/interpret/write the file format
• Updater - the ability to modify and write back to the file while preserving format and structure

Filename / File Format	OM2	OM3	ESM1.5/1.6	Parser	Updater	Notes
YAML	📍	📍	📍	✅	✅	Payu configuration etc
_in and .nml	📍	📍	📍	✅	✅	Fortran namelists
MOM_input	–	📍	–	✅	✅	Special file for MOM6 only
nuopc.runseq	–	📍	–	✅	✅	Special file for NUOPC coupler only
nuopc.runconfig	–	📍	–	✅	✅	Special file for NUOPC coupler only
.xml	–	📍	–	❌	❌	data forcing configuration file CDEPS
forcing.json	📍	–	–	✅	✅	Special forcing file for OM2 only: #20
field_table	📍	📍	📍	✅	✅	declares what variables exist and how the model should handle them internally (Ocean model)
data_table	📍	📍	📍	❌	❌	a forcing-field look up table - tells the model where to get external data for certain fields (Ocean model)
MOM_override	–	📍	–	❌	❌	Special file for MOM6 only

📍 Required for the given model
– Not applicable
✅ Implemented
❌ Not implemented
🛠️ Work in progress in a related PR

---

Parsers are currently located in src/experiment_generator/tmp_parser. Once ACCESS-NRI/access-parsers is ready, these temporary implementations will be replaced.

---

Several configuration files are still missing from the table above. Could you @chrisb13 @anton-seaice @dougiesquire @blimlim @Whyborn @ccarouge @penguian @aekiss please add them and update the table directly using the same merged-emoji format? I think most of the missing files are from esm1.5/1.6 configs? Also, please tag anyone else who might have relevant input or should be aware of this repo. Thanks!

I’d also greatly appreciate it if you could share some use cases as new GitHub issues in this repo — I’ll review them and see what we can do :)

[micaeljtoliveira]

@minghangli-uni Would be good to have a similar issue opened in ACCESS-NRI/access-parsers.

[minghangli-uni]

Sure I'll do that

[anton-seaice]

In theory, there is the NAMCOUPLE file for oasis based models. I'm not sure any of us are particularly familiar with it, so i wouldn't worry until requested.

I vaguely recall discussions of options for a yaml for the mom "_table" files? Maybe that's easier than implementing a special parser for them ?

[minghangli-uni]

I vaguely recall discussions of options for a yaml for the mom "_table" files? Maybe that's easier than implementing a special parser for them ?

Since the current tool already supports yaml, so the yaml input for the diag_table should work as well. I’ll try to add an example under the examples/ directory. One thing to note, before running the configs using access-experiment-runner, you would need to check out each branch, run something like python3 make_diag_table.py, and then relink the correct diag_table.

This isn’t a big deal for the control experiment (done once), but for perturbatios it may be inefficient.

[dougiesquire]

I vaguely recall discussions of options for a yaml for the mom "_table" files? Maybe that's easier than implementing a special parser for them ?

Yaml support is available in recent versions of FMS (e.g. used by ACCESS-OM3), but it is not available in the older versions of FMS used in our MOM5-era models. I think it would be big undertaking to update FMS for these models or to backport the yaml file support.