I’m trying to find instances of a specific word in comments across all iNat observations, but I’d like to exclude a certain commenter from the search results. Is there a way to do that with the API?
That was a hypothetical link to show what kind of syntax I would be looking for in the URL. The actual use I have is searching for the word “corrupt” without commenter ID “tonyrobelo”. I want to find corrupt image files by searching for conversation about corruption, but I noticed that user uses a copy-pasted reply in their curation activities that includes the word “corrupt” in hundreds (possibly thousands) of instances. That user is saying “corrupt” in a different sense, so I’d like to filter them out.
As a kludge that would catch many instances of what you’re looking for, Alex, you could search for strings such as “corrupts” (without quotation marks), “corruption”, “corrupted”, “corrupt%20t”, “corrupt%20a”, and so on, all of which would exclude the unwanted string “corrupt%20your%20coordinates”.
Yep! That’s what I was playing with, until I gave up trying to deal with duplicate hits and that user’s other formulae of responses that include other forms and pluralities of the root word “corrupt”. Your suggestion is an effective approach, though. Maybe it is best to treat it as a feature request.
I think for my purposes, curation is not exactly the category. I’m looking to collect corrupt image files rather than flag them for removal. I think they are interesting and valuable artifacts.