Users reported that Corriere della Serra was showing a paywall after visiting the first (free) article.
We contacted them and they replied:
I see the use of a proxy, so you should put also the following files in your proxy:
*.tinypass.com
*.piano.io
*.pn.corriere.it
Once you use the files in your proxy the paywall will disappear.
Quick question, @Samwalton9: do you think the Corriere della Sera archives also needs those filed added?
I've just clicked around the archive and its content a bunch and not run into any issues, so I don't think so.
Pushed to prod! Users who still are experiencing issues should delete their cookies to solve it.
This is looking good to me - clicking around news stories on the site doesn't seem to trigger a paywall.
We're going to set them up with temporary access to the library to test from their end, like we did with Newspapers.com
We received a response from their testing:
I have verified calls made to our paywall and not all calls are mapped by your proxy.
These calls should also be proxed:
api-esp.piano.io (es. https://api-esp.piano.io/public/sdk/v04/sdk.js?v=vz.1.103.11-53a5e49)
c2.piano.io (es. https://c2.piano.io/xbuilder/experience/execute?aid=aG4QZwTKpu)
buy.tinypass.com (es. https://buy.tinypass.com/api/v3/anon/template/loadTemplateContext?aid=aG4QZwTKpu)
I have all of those endpoints configured for multiple styles of rewriting to include them in the proxy, but they are still getting missed. There is a ton of dom manipulation happening on this site, and the things that we're missing appear in iframes inserted via js. I have been unable to track down where those values are getting created, because if they were just in the response body for the page, they would be getting rewritten by now. If our contact could help us track down how those uri's are ending up in the page in the first place, it would help me create the rules for rewriting them.
https://github.com/WikipediaLibrary/twlight_ezproxy/pull/38