Many publishers tried to prepare JATS XML files but have problems with a different versions of the software or do not know how to do that. What are your experiences with creating JATS XML? Did you use some of the software, what are the advantages or disadvantages of using the same…? Any opinion, suggestion, links, or tutorials are welcome.
For JOSS, the Journal for Open Source Software, we have developed a
publishing pipeline based on pandoc. It currently generates PDF, JATS,
and Crossref XML output.
We have talked about the approach on the last JATS-Con (slides); the article is here:
TL;DR: The approach works really well for papers that don’t have “special” elements, as is the case with typical JOSS submissions, and it could be modified to work with simple LaTeX or Word/Docx input. However, elements like complex tables, figures with subfigures, etc. are problematic and can usually not be handled by the pipeline.