RESTAPI - Data model gaps for 'sub-works' (eg book-chapter)?

I am new to the RESTAPI, but have a fair amount of experience with the unixref and unixsd xml models.

When using the rest api ‘works’ endpoint for a book-chapter doi (as an example) very limited information about the ‘container’ (book) is given. Using

api.crossref.org/works/10.1093/acprof:oso/9780199689804.003.0020

as an example, there is no information on the Editors, or container subtitle.

If we use the XML API

doi.crossref.org/servlet/query?pid=tolan.abbott@shabash.net&format=unixref&id=10.1093/acprof:oso/9780199689804.003.0020

We get a full section on the container/book as well as on the chaper/content_item

The rest api therefore does not seem to return enough data to validate/generate a typical citation to a book chapter (at least for an edited book). Nor does it seem to include the DOI of the ‘container’ to make further queries to fetch this data.

Am I missing something?

1 Like

Hello, and thanks for your question.

You’re not missing anything. This is a known issue that has been present since we built the REST API - it was essentially a flaw in modeling the work types and their relationships to one another. We unfortunately just haven’t had the resources to do the kind of comprehensive work necessary to get it corrected yet.

That said, if you need to generate a formatted citation based on a DOI’s metadata, you can use Content Negotiation. Even though that primarily relies on the data from the REST API, citations for book chapters and other ‘child’ items do include the container/parent information.

For example

$ curl -LH "Accept: text/x-bibliography; style=apa" https://doi.org/10.1093/acprof:oso/9780199689804.003.0020

returns

James, H., Borscheid, P., Gugerli, D., & Straumann, T. (2013). Growth and Cultural Change. The Value of Risk, 318–335. https://doi.org/10.1093/acprof:oso/9780199689804.003.0020

which has both the chapter title “Growth and Cultural Change” and the book title “The Value of Risk”.

Please let me know if you have any further questions.

Many thanks for the reply, that is a shame, but thanks for the suggested work around.

If we do go via the content negotiation route we are hitting doi.org directly. Is there still a requirement / desire to identify ourselves (ala polite pool) and if so what is the correct way to do so?

many thanks

No, that’s not necessary for content negotiation. It’s centrally managed by the DOI Foundation, because it supports getting citations/metadata for DOIs from several different registration agencies (Crossref, DataCite, and mEDRA). The requests that pertain to Crossref DOIs are directed from there to our servers, but the public/polite distinction isn’t maintained.

1 Like