Python script to generate a MOM5 or MOM6 diag_table, for example using a systematic one-file-per-variable naming convention.
Edit diag_table_source.yaml and run make_diag_table.py to create a diag_table. Warning: this will overwrite any existing diag_table.
Adding or removing diagnostics is as simple as adding or removing their names in the appropriate fields section in the diag_table part of diag_table_source.yaml.
make_diag_table.py is very general, so you should be able to generate any diag_table you like, with any output file naming convention, entirely by editing diag_table_source.yaml. See the comments in diag_table_source.yaml for usage instructions.
Note that each of the top-level categories in the diag_table part of diag_table_source.yaml can have only one instance of each field name, so if you need multiple outputs of the same field (e.g. as both averages and snapshots), you'll need to make additional categories.
An (incomplete) list of MOM5 diagnostics is available here, obtained by this method.
MOM6 diagnostics are listed in available_diags.* following a run of the model.
Also see these discussions on standardized ACCESS-OM2 and ACCESS-OM3 filenames.