I’m a learner in terms of APIs, especially in combination with KNIME AP. I know there is a Flickr API OAuth Node in Selenium/Palantir but this doesn’t return a Flickr-specific value which is needed in some cases. So I want to build the API interface on my own and without external scripting (Java, JS, Python, etc.) - the hard way. Especially as the Flickr API is known to be old-fashioned and poorly documented
But still the HMAC Hashing works in another way in KNIME than in PHP. And the GET in the end only delivers a 400 error (which is most likely connected). As I have already invested several hours in this workflow and also learned a lot, I would be happy if someone with more expirience here could help me in debugging the whole thing. And no, external scripting is still not desired, as long as there is a node which can do the job
I’m not an expert in this domain by any means, but it seems like the sticky bit is related to the hashing calculation. Since you are making use of the Palladian nodes (and since he has a really good grasp of web scraping) let’s tag @qqilihq and see if he can chime in?
looking back, I’ve seen that you reported a Flickr-related bug for the OAuth Connector node in Palladian here:
As I cannot reply to this thread any more, but I do not want to keep it an unfulfilled promise, I’ll put my feedback here: Unfortunately the bug reported cannot be fixed in Palladian, as Twitter insists on the https callback URL (which makes it impossible to acquire a token via http://localhost – which is a perfectly valid use case).
There’s two Flickr threads I noted back then, and both do not look like Flickr still cares too much about their API: