Open judell opened 8 years ago
Thanks for helping clarify @judell! I'm still a bit unclear, then, on how uri.parts
actually works. From what I can tell, it: splits out all word phrases, ignoring punctuation, and then searches for any of those phrases within the annotation uri? Is that a correct understanding?
"Is that a correct understanding?" I have not penetrated deeply into the mysteries of Elasticsearch but, insofar as I do understand what's happening, yes :-)
For example, to find all the annotations on a domain, query the uri.parts field (but don't include .com, .org, .edu and the like; this currently results in the API returning annotations sharing any of those TLDs.)
Unfortunately it doesn't work the way you expect. For example:
https://hypothes.is/api/search?uri.parts=unix
We will, but don't yet, have a site: facet that does what you expect.
PS: Thanks for doing this excellent work!