Previously, I’ve been using a proxy address provided by my old institution (LSE) to access materials easily through Zotero Connector on Firefox (Mac). The string required was '%h.gate3.library.lse.ac.uk/%p'. This continues to successfully redirect my searches (e.g. on Google Scholar) through my old university’s SSO system and gives me quick access to files and PDFs. So far, so good – but I'm about to lose access to my LSE account.

This is because I've now moved to Oxford University, who appear to use EZProxy for redirections. The system is called 'proxy by port' and it's been suggested to me that 'https://ezproxy-prd.bodleian.ox.ac.uk/login?url=' should work... but I can't get it to redirect at all with any combinations of %h and %p appended to the beginning or end of the address.

Frustratingly, it seems like it almost works but not quite! For instance, I get a 404 error if I use 'https://ezproxy-prd.bodleian.ox.ac.uk/login?url=%h./%p’ in Zotero’s preferences. The fullstop after the hostname string (%h) is clearly incorrect, I think, but after removing that fullstop Zotero no longer allows me to reload through a proxy at all, which is quite puzzling! When I right click the Zotero Connector there is simply no option to use the Oxford proxy, only my old LSE one.

What makes it particularly odd is that this naming scheme (excluding the fullstop) should make sense… e.g. If I copy paste any of the ezproxied links from this page (https://oxford.rl.talis.com/lists/D882A1AC-F76A-E2C9-CBB6-B11EB2D6FCED/bibliography.html?style=national-library-of-medicine-grant-proposals) into my browser, I get correctly rerouted through Oxford's SSO system and can login etc. It’s just Zotero that doesn’t manage to let me do that automatically with other pages! Trying strings like 'https://ezproxy-prd.bodleian.ox.ac.uk/login?url=http://%h/%p’ or 'https://ezproxy-prd.bodleian.ox.ac.uk/login?url=%h/%p’ for instance do not work.

I also tried '%h.ezproxy-prd.bodleian.ox.ac.uk/login?url=/%p' but that didn’t work either… e.g. it fails to load 'https://www-tandfonline-com.ezproxy-prd.bodleian.ox.ac.uk/login?url=/doi/abs/10.1080/14479338.2014.11081981’ (as an example file to access).

Interestingly, if I remove the ‘login’ portion, I get a message form Oxford’s system telling me to add it back into the URL, so it’s seemingly getting through to Oxford’s system but not working for some other reason.

Finally, through manually loading some pages, Zotero is indeed automatically picking up the correct proxies and storing them for the future. 'ezproxy-prd.bodleian.ox.ac.uk:2109/%p' now works for TandF online, for example, as does a similar one for ProQuest. Obviously I'd rather not manually enter a string for every single journal though, complete with customised port number.

So, is there some way to get Zotero to correctly route my proxy requests and automatically re-route any new searches through the proxy like it used to before I moved institutions? Any and all help would be greatly appreciated!