HathiTrust API
HathiTrust partners (and any other interested parties) need a way to get information unique to HathiTrust (handle data, rights information) into their own systems. This documents begins to describe an HTTP interface that aims to provide:
- data about a small number of records at once (other avenues will be available for larger queries)
- only data unique to HathiTrust, along with enough unique identifiers to make sanity-checking possible
- an interface amenable for use within a browser environment
The initial target of the design will be a typical “Full Record” page in a web-based OPAC; other uses may reveal themselves and influence the design along the way. Something I’m going to want help with from all comers is how to deal with serials. They’re a mess.
Why not just use SRU/OpenURL/Whatever?
The short answer is that all those technologies are much, much heavier than what we need to get off the ground and all would be a huge pain to try to deal with in browser-embedded javascript. I’m not averse to adding interfaces as the need arises and time permits, of course.
Phased Development
Development will be segmented into phases, with ever-increasing functionality until we have…enough.
- Phase 0 (FINISHED) will focus on getting a single non-serial record by passing in an OCLC number and providing a javascript library to make it easy to work with the returned data in a browser.
- Phase 1 (FINISHED) will add to phase 0 the ability to pass multiple possible identifiers (e.g., an lccn and oclc number) and will provide server-side rank-ordering of any multiple results.
- Phase 2 will provide a second interface designed to allow multiple simultaneous queries. This will still be focused on a small number (n < 20ish) of records at once.

