Hello OJS team,
I'm working on converting some paper journals to the OJS using the XML import feature, and had a few comments about the import process that might be helpful:
- The DTD isn't valid. The <section>, <number>, and <title> tags each have two separate definitions -- particularly <section>, which is once defined as a #PCDATA and once as being able to contain tags.
- The DTD could match the OJS's requirements more accurately. For example (correct me if I'm wrong) the DTD specifies that the author email is a required tag, while the OJS doesn't seem to require it. This applies to many of the required tags -- <date_submitted>, <date_published>, <abstract>, etc., all of which are nullable in the OJS database.
- In your XML example, one tag contains the text "N/A", but I think it would be much more conceptually clean if this tag were just omitted entirely. (I might have misunderstood why N/A is used.)
- The XML document structure would benefit from the use of attributes rather than just tags with #PCDATA contents. For example, <section order="2">...</section> rather than <section><section_order>2</section_order>...</section>.
alec (lives at) smecher (a dot) bc (another dot) ca