Zotero is open source and the code for various parts can be found at
https://github.com/zotero/. I think for the extraction the Translation
Server is the key. It can be found at
https://github.com/zotero/translation-server. Also the Zotero page has
its own forum (https://forums.zotero.org/discussions) and I think they
are usually quick to respond.

Greetings
Tobias

PS: There is also the little less known zoterobib (https://zbib.org/)
that might be interesting for some use cases (frontend and backend are
also at Github).

PPS: The browser plugin is at
https://github.com/zotero/zotero-connectors (but I still think the
Translation Server is what does the hard work. But I very well might be
wrong :))

On 2020-04-14 02:20, Mark Weiler wrote:
> The Zotero browser plugin seems to have the ability to extract metadata (e.g., author name) and PDF from a publisher's page. 
>
> From a screen reader perspective this is helpful as it avoids having to learn how to navigate publisher's pages.
>
> Anyone have time to talk about how this done via phone?
> mark