Okay, I did some more simplification. Should be fine from my side. However, I think that I have only have very limited access to Gale DB.

I added a fix for multiples.

Thank you for the fix! I couldn't test multiples because I don't have much access to the Gale DBs.

@dstillman -- are you able to test multiples successfully? I consistently get an error message (entry is null), presumably because the page isn't fully loaded when processSingleEntry fires. Happens in both Firefox and Scaffold.

I tried now with this URL in Gale Primary Sources - British Library Newspapers (searched for "save dog") and it works fine in Scaffold as well as in the browser. The new added monitorDOMChanges should be taking care about the dynamic loading of the website. However, in Scaffold this shouldn't even be needed because you can simply wait until the page is loaded and only afterwards click on the detectWeb button. Which URL resp. search did you try?

It might have to do with the proxy? When I try to follow the link given by in the zotero-label section, e.g. https://link-gale-com.libezproxy2.syr.edu/apps/doc/A609314409/AONE?u=nysl_ce_syr&sid=zotero&xid=ff03afdf it takes me to https://go-gale-com.libezproxy2.syr.edu/ps/i.do?p=AONE&userGroupName=nysl_ce_syr and "document could not be retrieved.

I just tried going through the regular article link instead (i.e. just h3 a) and that's working. That's on Academic one file, but might be more fragile since you're relying on identical page structure.

Just tried again and this isn't due to the proxy. https://link.gale.com/apps/doc/A602462094/PROF?u=nysl_ce_syr&sid=zotero&xid=18c0fad7 fails, too

@adam3smith I think the issue is specific to your library's EZProxy. I will investigate further and get back. Thanks.

Thanks @elamgit for looking into this. Let us know if this is fixable on your end. Otherwise I think we can find other ways to get the multiples, but the dedicated Zotero class is particularly robust, so it'd be nice to keep.

@adam3smith, Can you try now? I made a configuration change at Gale. You should not be getting the "document could not be retrieved" error message anymore.

Thanks @elamgit -- initial tests looking good. I'll run through a couple more pages, but I think we're good to merge here. Any remaining concerns from your end?

@adam3smith, I agree. No concerns from me and everything is looking good. Thanks.

Thanks everyone! Looking great now.
I'm leaving the two other Gale translators in place for now. It looks like the resources they point to still exist, though they're gradually being phased out (e.g. ECCO is also available through Primary Sources, which is covered by this translator). Since the other translators detect for different URLs, there's no harm in keeping them, at least for now.