Use GitHub action to update model files when new version of SS3 is available

[kellijohnson-NOAA]

Problem

The R script to update the model files must be run locally and files need to be committed by a human, which can lead to a delay in updating the files.

Solution

Use a GitHub action to run the script and update the files.

I am not sure if the GitHub action should be initiated here or from the stock-synthesis repo upon a release of SS3. If initiated here, then it would need to be run on a schedule to check and see if the current version of files are the same as the newest version of SS3. Where, if it is in stock-synthesis, it could run upon a new release. See documentation on actions that are triggered with tags.

Tasks

• push local changes to .R script that decreases number of steps
• fix script to get executable from artifacts
• add GitHub action that uses the script and commits the changes to the repository then opens a pull request
• add step to action that tags this repository with the version number of SS3 being used

[iantaylor-NOAA]

I think this is a good idea and would be fine to just have it triggered by manual action (as many of the actions already are) with no automatic trigger to run it. Then we could move it up in the checklist sequence, to try it once before release and then repeat after.

[kellijohnson-NOAA]

@iantaylor-NOAA also noted that it might be a good practice to run this before a release to look at the differences in the files, for example, for 3.30.20 it was ran after and led to noticing that ADMB version number had not been updated.

[e-perl-NOAA]

This is now completed by manually running the github action here and was successfully run for the v3.30.21 release. There might be ways to even further automate this upon new releases and automating the release process in general but that would be a much bigger lift.