Add contributor roles

[cthoyt]

Closes #73. Supercedes #75

This PR enables annotating author roles using the Contribution Role Taxonomy (CRediT). It also enables (optionally) annotating the contribution degree, as suggested by JATS.

It does the following:

• Add contribution list to end of default.latex (after reference list)
• Add contribution list to end of preprint.latex (after reference list)

Demo

Here's what this looks like at the end of the main PDF build:

Future Work

Future work for future PRs:

1. Work with Crossref to get a reasonable model for capturing this. Crossref is currently discussing supporting this.
2. Upstream JATS generation in pandoc, started in Add CRediT roles to JATS jgm/pandoc#10153
3. See how this fits in the ConTeXt template, if this gets picked up by Open Journals

[cthoyt]

@tarleb I adopted the same kind of error handling you added to adding contributor information - now it will print the errors during build but then move on without causing a failure

[cthoyt]

This is ready for review @arfon @xuanxu @sneakers-the-rat

[sneakers-the-rat]

i love this, thanks for your work.

some minor comments, mostly thinking about future support of jats and other formats.

if the work is intended to have multiple stages, i wonder if we could PR these into a credit branch and merge them to main all together? atm we don't have a good way to regenerate papers from older submissions (i think?), so if people start using it before we have it finished then we would have some incomplete papers that have the LaTeX version but don't have the JATS version.

[tarleb]

@cthoyt How do you feel about putting this filter into a separate repository from which the filter is fetched during installation? This would give you more options to test the filter, would be more maintainable, and would make it easier to re-use the filter outside of Inara.

The perevir tool has been developed specifically for testing transformations like this, it might come in handy. See e.g. querverweis for a filter that uses it for tests. (Aside: I also plan to use querverweis in Inara to replace some hacky filters but, you probably guessed it, I didn't get around yet to making the switch.)

[cthoyt]

@tarleb are there examples of filters that have been split into their own repositories and use those tools? I would be hesitant otherwise, since I'm guessing there's going to be a learning curve both for using that tooling and also for getting it working inside Inara

alternatively, if this is looking like complete enough for you, we could merge it in Inara then work together to externalize some of these tools in a more systematic way (and write contribution guides)

[cthoyt]

@tarleb any thoughts on this? Would you consider merging this PR as is?

I don't think it's a great use of my time to learn new packaging and tooling without a lot of support

[tarleb]

I think we should hold off on this until we have fixed a few pressing issues, like getting JOSS into PUBMED (openjournals/joss#153). Sorry for the wait.