[feat]: rerender_skeleton()

[Schiano-NOAA]

What is the feature?

• Separate rerender_skeleton() and create_template()

How have you implemented the solution?

• Removed all code associated with the argument rerender_skeleton within create_template and made it it's own function
• repetitive code was made into their own functions within the utils.R file
• functionality is exactly the same for users -- aka
  create_template(rerender_skeleton = TRUE, ...) == rerender_skeleton(...)-- arguments are the same

Does the PR impact any other area of the project, maybe another repo?

• no

Note: this will be NOT be backwards compatible, so once this is merged, we will need to increase to asar v3.0.0

[sbreitbart-NOAA]

Before I review the code, a couple thoughts:

1. This needs to be announced far and wide. Could you please update the documentation (vignettes, tutorial, etc.) to highlight this change, and explain how to alter one's workflow in accordance?
2. Can this feature be released through a deprecation cycle?

[Schiano-NOAA]

@sbreitbart-NOAA all good points! On second thought, hold off on the review. I will re-ping when it's done. I have to do a bunch more then to make this ready.

[Schiano-NOAA]

Before I review the code, a couple thoughts:

1. This needs to be announced far and wide. Could you please update the documentation (vignettes, tutorial, etc.) to highlight this change, and explain how to alter one's workflow in accordance?
2. Can this feature be released through a deprecation cycle?

I have initiated deprecation of the argument rerender_skeleton and updated documentation to reflect this. Next week, I will update the tutorial in a PR, but I feel we should merge this without waiting for the update. I don't think it will take me long but I don't want to put the review of this on hold until another PR is made.

Could you review this one and update_report and I will hold off on merge until the tutorial and workshop is updated? I don't think this is mentioned in any vignettes since I did not find any mention of rerender_skeleton in them.

[sbreitbart-NOAA]

👏 bravo!!
Even if this wasn't as difficult as you expected, I think this is a big achievement. I'm relieved that create_template() will be much easier to parse through/edit/test, and I'm hopeful that our code coverage will increase significantly since the main area missing tests was associated with rerendering a skeleton. I left a couple minor comments- two about removing what seems like outdated comments and one about having trouble updating the title with a new species. Once those are addressed, I'll check out the other stacked PR!

[Schiano-NOAA]

@sbreitbart-NOAA this should be good to go now!

[sbreitbart-NOAA]

The rerender_skeleton functionality is working very well! One minor error that doesn't prevent rendering, but might be problematic: the title's quotes disappear when the skeleton has been rerendered.
Also, when I run create_template(bib_file = FALSE), the qmd doesn't render because of issues with the journals style file. I'll make a separate issue for that.

[Schiano-NOAA]

@sbreitbart-NOAA sounds good! I will fix the issue with the title and open a new PR with the fix for the render issue. I think it will be easy and straightforward.

[Schiano-NOAA]

Note: legacy replacement does not work on rerender

it does not rename these documents -- only in the skeleton

[Schiano-NOAA]

@sbreitbart-NOAA as long as the checks pass, this should be good now. One issue is that the testing for the legacy figs and tabs do not pass and won't be id'd since they only run locally. The issue is that there is no code to rename the files themselves, only those in the skeleton which creates a mismatch.

I am not sure where to put this. I could just rename them but in other template types, the number of these documents varies. Any ideas? Could you make this adjustment since you handled the change from tables & figures to figures & tables?

[sbreitbart-NOAA]

@sbreitbart-NOAA as long as the checks pass, this should be good now. One issue is that the testing for the legacy figs and tabs do not pass and won't be id'd since they only run locally. The issue is that there is no code to rename the files themselves, only those in the skeleton which creates a mismatch.

I am not sure where to put this. I could just rename them but in other template types, the number of these documents varies. Any ideas? Could you make this adjustment since you handled the change from tables & figures to figures & tables?

If it's too complicated to apply to this situation, I think it's ok to make sure it doesn't run. Instead, the user can manually change the files if needed; it's straightforward and fast.

[Schiano-NOAA]

@sbreitbart-NOAA I'm not sure if it's too complicated but just saying it's not changing them in rerender_skeleton

[sbreitbart-NOAA]

@sbreitbart-NOAA I'm not sure if it's too complicated but just saying it's not changing them in rerender_skeleton

Oh, I understand now. I'd go with correctly ordering the docs in the skeleton. Then, we can add code that 1) IDs other files in the skeleton's folder and evaluates if Figures appears before Tables, and if not, 2) reorders the files by changing the filenames. How about you merge this into the update-report branch, then I'll work on the above feature and merge it into update-report before merging into main?