I read somewhere in the documentation that it is considered best practice to register DOIs not just for articles, but for a nested structure like journal > issue > articles. However, I get seemingly mutually contradictory error message when trying to register issues. If I register issues and articles in the same batch, nesting the articles inside the issues like follows, then I get the error “Invalid content was found starting with element ‘journal_article’. No child element is expected at this point.“.
<journal_issue>
<journal_article>
</journal_article>
</journal_issue>
If I try registering them separately, without this nested structure, I get errors like: Invalid content was found starting with element ‘journal_issue’. One of ‘{“http://www.crossref.org/schema/5.3.1”:journal_article}’ is expected.
I have followed the examples given in documentation, as far as they go, but they don’t completely spell out this case.
Ideally, we would register articles and issues separately, in separate batches, as it would make for cleaner code on our side.
Our official best practice recommendation is to register DOIs for all journal articles and one DOI for the journal as a whole. But, it is not best practice to register DOIs for issues or volumes.
Crossref DOIs are intended as citation identifiers, so the general rule of thumb is that if something is likely to get cited, then it should be registered. Generally journal issues and volumes are not cited, so there’s really no benefit to registering DOIs for them, though you can if you really want to have those identifiers. (journals as a whole are also not typically cited, so there’s not really much practical benefit to registering journal-level DOIs either, but our schema has to support them for situations where there are no ISSNs)
In terms of why your xml isn’t validating, it sounds like some of the tagging is out of order, but we’d need to look at the file itself to know for sure. The community forum doesn’t allow file uploads, so we’ll have to take that question elsewhere. Could you send an email to support@crossref.org with that xml file attached?
Ah thanks. I can’t remember where I got the thing about registering issues too; perhaps something misremembered. Good that the recommendation is to not do this, as it seemed a bit odd to me anyway, and it will make it easier to handle the registration with just articles and the journal itself. There should be no problem with the XML if so.
I’m having similar issues - it sounds like including volume or issue DOIs in the same batch is not recommended. Our journal seems a little odd with regard to volumes and issues:
They do one volume per year. The volumes are numbered, but the numbers are not the same as they years (but close) - Volume 24 is for 2026.
They also do Special Volumes - those are sequentially numbered.
The archives at Digital Commons have the special volumes as [url]/vol0/issN, where the regular volumes are [url]/volN where N is the volume number.
I’m not totally sure how to deal with the above.
But given what you are saying, do we actually register DOIs for volumes and issues? Or do we just register the articles and include the volume &/or issue # in the article data? Or just skip trying to deal with the volumes/issues since the purpose is citation.
In our case, the answer was to stop trying to register DOIs for issues, volumes etc themselves. We only have DOIs for the individual articles (as well as top-level ones for journals themselves–not their individual issues).
That’s kind of what I was coming to, and was able to get almost the whole XML file to work without errors. A bit more tomorrow and I should be good to go…
Thanks for your thoughts…