Rework master_list* files etc.

[samsrabin]

Description of changes

• Renames master_list*.rst files to history_fields*.rst.
• Updates those files with latest output lists and descriptions.
• Removes line numbers from those files.
• Renames various variables, changes comments, etc. to avoid use of "master" (in references to these files only).

Will unblock #2074.

Specific notes

Contributors other than yourself, if any: None

CTSM Issues Fixed:
Resolves #2083.
Some work on #1127

Are answers expected to change (and if so in what way)? No

Any User Interface Changes (namelist or namelist defaults changes)? Yes: hist_master_list_file parameter changed to hist_fields_list_file.

Testing performed, if any: Running with hist_fields_list_file = .true. produces history_fields_nofates.rst, which looks fine.

[samsrabin]

@ekluzek What do you think of adding this to your upcoming simple bit-for-bit tag? It'd be nice to get it in as it's blocking #2074.

[ekluzek]

@samsrabin yes, I'll add it to the list.

[samsrabin]

• Resolve issue where history field list file is saved without rst extension.

[samsrabin]

@ekluzek I'm requesting a re-review because I discovered some more places to change variable names and references. I also added a testmod SaveHistFieldList.

Finally, I updated the .rst files that we ship, using the outputs of the following tests:

SMS_Ld1.f10_f10_mg37.I1850Clm50BgcCrop.cheyenne_intel.clm-SaveHistFieldList
SMS_Ld1.f10_f10_mg37.I2000Clm50FatesCru.cheyenne_intel.clm-SaveHistFieldList

These tests use the default output variables (i.e., without specifying any additional outputs in user_nl_clm). The .rst files thus indicate "Active?" with T/F based on the default list. I felt this would be the most useful for people referring to those files.

However, it differs from the description @slevis-lmwg wrote in README_master_list_files (original text below; diff with my updated version here):

2021/9/8 slevis

The files master_list_nofates.rst and master_list_fates.rst each contain a table of the history fields, active and inactive, available in the CTSM cases that get generated by these tests:
ERP_P36x2_D_Ld3.f10_f10_mg37.I1850Clm50BgcCrop.cheyenne_gnu.clm-extra_outputs
ERS_Ld9.f10_f10_mg37.I2000Clm50FatesCru.cheyenne_intel.clm-FatesColdDefCH4

To reproduce these .rst files, run the above tests and the files will appear in the corresponding run directories.

What do y'all think of my change to indicate "Active?" based on the default settings?

[slevis-lmwg]

What do y'all think of my change to indicate "Active?" based on the default settings?

If I'm understanding correctly that this will indicate active/inactive for each field, yet all the fields will still get listed, then I like that.

[samsrabin]

Yep, that's right!

[samsrabin]

• Add vertical dimension (e.g., nlevsoi) as a column to the .rst files. If none, use - to match ncdump convention.
• Merge in changes from history write optimization #2115 once that's complete.
• Do a final update to the .rst files.

[billsacks]

@samsrabin - thanks for your recent commit adding the level dimension to this table!

[ekluzek]

This is great. The change to remove the field number from the files will really help prevent git conflicts in those files. I also like your removal of master for the field lists. You probably didn't even know, but that is work towards #1127 which is really awesome. There are some technical terms in use that have negative historical connotations and we should remove them from our use. So it's good to be making some progress in that regard.

[ekluzek]

Oh, something we should check if this interfere's with some of @johnpaulalex PR's that are still outstanding. So I'll try a merge and see what I see...

[ekluzek]

It turns out @samsrabin had already checked @johnpaulalex PR's, but I ended up rechecking and came to the same conclusion. There's one PR that has a conflict, but it won't be hard to resolve. But, it might have been good for me to check to know what to do to resolve it.