Error when capturing book details on Primo (Report ID: 1159069997)
The follow error occurs when capturing book details on Primo using Zotero Chrome connector:
An error occurred saving with Primo. Attempting to save using Embedded Metadata instead.
An error occurred saving with Primo. Attempting to save using Embedded Metadata instead.
http://prts-primo.hosted.exlibrisgroup.com/primo_library/libweb/action/display.do?tabs=detailsTab&ct=display&fn=search&doc=01PRTS_IIIocn997536196&indx=2&recIds=01PRTS_IIIocn997536196&recIdxs=1&elementId=1&renderMode=poppedOut&displayMode=full&frbrVersion=&frbg=&&dscnt=0&scp.scps=scope:(01PRTS_III),primo_central_multiple_fe&mode=Basic&vid=01PRTS&srt=rank&tab=default_tab&vl(freeText0)=companion%20historiography&dum=true&dstmp=1536153359781&gathStatIcon=true
We are also having this same problem.
We have noticed recently that Zotero no longer works for Ex Libris Primo sites that are still running on the 'classic' UI. When trying to save a record we get the following message:
An error occurred while saving this item. See Troubleshooting Translator Issues for more information.
Or sometimes:
An error occurred saving with Primo. Attempting to save using Embedded Metadata instead.
See for example:
http://virtuose.uqam.ca/
https://primo.tug-libraries.on.ca/primo_library/libweb/action/search.do?vid=WATERLOO
http://decouverte.uquebec.ca/primo_library/libweb/action/search.do?vid=UQTR
http://primo-standard-lb.hosted.exlibrisgroup.com/primo_library/libweb/action/search.do?vid=UOTTAWA
Searching for records on these sites and then attempting to save via Zotero results in an error (using up to date versions of Chrome or Firefox).
Curiously, for the University of Vienna, Zotero still works with their classic interface:
http://usearch.univie.ac.at/
Using the Chrome network inspector we can see that both urls have status 200 for Vienna:
https://usearch.univie.ac.at/primo_library/libweb/showPNX.jsp?id=TN_doaj_soai_doaj_org_article_47f5128d766d48f78d1765125b3e7d70
https://usearch.univie.ac.at/primo_library/libweb/showPNX.jsp?id=2
In the non-functionning sites, there is only one URL, using the docID rather than the session ID, and it gets a 404 or 500 error:
https://primo.tug-libraries.on.ca/primo_library/libweb/showPNX.jsp?id=vtug2256046 (404)
http://primo-standard-lb.hosted.exlibrisgroup.com/primo_library/libweb/showPNX.jsp?id=TN_jstor_archive_1341886962 (404)
http://virtuose.uqam.ca/primo_library/libweb/showPNX.jsp?id=UQAM_BIB001313503 (500)
There was a recent update to the Zotero Connector :
Version 5.0.41, Updated September 4, 2018
This may or may not be relevant but it seems the problems started around this time.
Is there a way to diagnose and fix this problem? I do not think the source of the problem is with Primo, since this problem is occuring in both our test and production environments, and those are running on different versions (test is on the most recent August 2018 release, while production lags behind on the August 2017 release. Zotero previously worked properly in both environments).
Please let me know if any additional information is necessary.
Thanks,
Sacha
Here is a follow up from Ex Libris.
Sacha
-- cf. --
Thank you for reporting this. I was able to verify that Zotero is not working properly in the Classic UI. When going to the newUI:
http://dam.bib.uqam.ca:1701/primo-explore/search?query=any,contains,space&tab=default_tab&search_scope=upac&vid=UQAM&offset=0
The behavior was better.
We have reported this as a defect to be handled by the development team.
We will continue to update this case as new information becomes available.
As to why the Primo translator has broken, it appears to be related to showPNX.jsp. TUG was returning 404 for showPNX.jsp because at some point it got removed (maybe in a Primo upgrade?). After we re-added it, based on the version available in the Developer's Network [2], now we're seeing 500, and the output doesn't look like showPNX.jsp is getting called at all, but rather that the regular Primo page generator is getting called with parameters it doesn't understand and is failing. Maybe some Tomcat configuration was required on the FEs to get that URL to be loaded directly?
[1] https://en.wikipedia.org/wiki/COinS
[2] https://developers.exlibrisgroup.com/blog/showPNX_Revisited
http://www.lib.ntu.edu.tw/
https://forums.zotero.org/discussion/comment/319017/#Comment_319017
Someone should probably let Ex Libris know this is resolved.