Alexander Winklerirrregular reminder that you can ask google to unlock books that are in the public domain but not openly available via the feedback form. It usually takes them one or two days to review the request.
Till Grallert@awinkler Great info! Do you also happen to know how to effectively query Google Books metadata? The options available in the interface and through the API seem far too limited to, for instance, return all Arabic periodicals in their holdings.
@tillgrallert Not really ... I just know the api docs: https://developers.google.com/books/docs/v1/using
Here's an api call for 'magazines', can't get the langRestrict parameter to work though: https://www.googleapis.com/books/v1/volumes?q=a&printType=magazines&langRestrict=it
@awinkler same here, even though it seemingly works for other queries. But I also seemingly always get in excess of 1 mio positive hits.
@awinkler This is one example for the >1 mio. hits searching for some Arabic keywords in titles. Note that most if not all magazines are classified as books.
@tillgrallert bottomline: it's a mess. Btw, I've already hit the rate limit, so the API seem pretty useless for data analysis.
Libraries partnering with google books should have published corresponding metadata separately. Ages ago, i tried to get a list of google books ids with corresponding ids from state library in Munich: https://github.com/alexander-winkler/gb2b3kat/tree/main But that's painful and hacky and doesn't obviously scale.
@awinkler absolute mess indeed. To be fair though, a lot of this data is available through Hathitrust and their open APIs albeit without the corresponding Google Book ID. Libraries also share the data with OCLC, who then lock it away from our prying eyes.