Add refinediag preanalysis tests

[singhd789]

Describe your changes

• added dummy preanalysis and refinediag scripts
• added preanalysis and refinediag settings in yaml config (for both local and remote)
• updated test_cloud_runner pipeline for tasks that needed to be checked
• added sed line in runscript to check a line was changed (not really needed though, more for my own sanity)
• updated flow.cylc for preanalysis and refinediag tasks (missing some lines/needed to be updated)

Issue ticket number and link (if applicable)

Fixes #216

Checklist before requesting a review

• I ran my code
• I tried to make my code readable
• I tried to comment my code
• I wrote a new test, if applicable
• I wrote new instructions/documentation, if applicable
• I ran pytest and inspected it's output
• I ran pylint and attempted to implement some of it's feedback
• No print statements; all user-facing info uses logging module

Manual Pipeline Run Details

Was the manual pipeline (test_cloud_runner) triggered for this PR?

• Yes

• No

• Manual pipeline was run but changes made in pipeline not seen until merged in (I don't think)

• pipeline results show preanalysis and refinediag scripts run BUT there is a mask-atmos-plevel_atmos_month_cmip error 😢 (for workflow artifact files: https://github.com/NOAA-GFDL/fre-workflows/actions/runs/25129758975/job/73656944006)

(Paste relevant logs, output, or a link to the workflow run here)

How to trigger the manual pipeline:

The test_cloud_runner pipeline is not automatically associated as a required check with the PR; it must be triggered to test changes in a full post-processing run.

To trigger the manual pipeline:

1. Follow the link to the test_cloud_runner actions tab here

   • you should see "This workflow has a workflow_dispatch event trigger"

2. Click the dropdown "Run workflow":

   a. If trying to merge from a branch on fre-workflows: choose branch from the first drop down, leave the next 2 inputs blank, and choose the fre-cli branch to test

   b. If trying to merge from a fre-workflows fork: can skip first branch selection, input the fork name (ex: [user]/fre-workflows), input the fork's branch name, and choose the fre-cli branch to test

3. Click "Run workflow"

Note: you may need to reload the page to see your running workflow.

[ceblanton]

How about /work/$USER/... instead so it will work for everyone?

[singhd789]

OH, good catch! I meant to change this

[singhd789]

Chris....you may have just really helped the problem I've had for like 2 weeks with this catch. (at least for the local test). I didn't realize the refineDiag post-script was looking somewhere else (with $USER) for files.

[singhd789]

Wait...uhg, nevermind, I see my same issue again. It seems when I update the refine_history path to something new, the refineDIag post-script works. However, if I don't update the path and run again, it does not somehow. Continuing to look into

[ceblanton]

Lying in bed last night, I thought that I should have suggested removing hsmget and hsmput entirely at the first glimpse that they were non-portable, and replaced them with humble tars.

But.. since we're this far, can we use hsm/1.4.0 instead of this sed?

[singhd789]

Yeah, luckily this line I think is only checking that that line was changed with the -n option - so it's not actually changing anything here. I can remove this line.

Now, unfortunately, in order to use the hsm1.4.0 package, I had to update the cylc-flow-tools.yaml , which meant the container needed to be rebuilt.....but when I built it again is when I realized fregrid was not available anymore 😢

[ceblanton]

Except for the ppan.cylc test changes, this is more than ready.

How are we going to get fre-nctools back into the cloud pipeline?

[ceblanton]

It looks perfect. Thank you so much for the this, which turned out to be so much harder than I could have imagined.

[singhd789]

It looks perfect. Thank you so much for the this, which turned out to be so much harder than I could have imagined.

I just don't know how it was working before 😆