Primo issues

kevinwilson
Hi, we're just trialling our Primo catalogue, which we're going to roll out very soon. I wanted to check whether it'd be compatible with Zotero, but each time I try to save some results, I get an error message and am asked to check known translation issues.

Here's the link to the catalogue - http://primo-service.hosted.exlibrisgroup.com/primo_library/libweb/action/search.do?vid=44GOL_V1

I gather it's a constant issue for libraries, how do we get around it? We're just about to start promoting Zotero as the citation management software of choice!

Thanks
  • adamsmith
    Zotero relies on the pnx data from Primo for the translator to work - in typical primo installations that can be reached by adding &showPnx=true to the details.do URL
    I think the URLs are session dependent, but we're looking at something like: http://primo-service.hosted.exlibrisgroup.com/primo_library/libweb/action/display.do?tabs=detailsTab&gathStatTab=true&ct=display&fn=search&doc=TN_dawson9781441169099&indx=1&recIds=TN_dawson9781441169099&recIdxs=0&elementId=0&renderMode=poppedOut&displayMode=full&frbrVersion=3&fctN=facet_rtype&dscnt=0&rfnGrp=1&scp.scps=primo_central_multiple_fe&fctV=books&frbg=&tab=default_tab&dstmp=1363368248346&srt=rank&mode=Basic&dum=true&rfnGrpCounter=1&vl%28freeText0%29=borges&vid=44GOL_V1&showPnx=true
    for your catalog.

    You can talk to exlibris how to enable that - it works well e.g. for the British National Library catalog http://explore.bl.uk/
  • kevinwilson
    Hi Adam, thanks. The link didn't work. I tried &showPnx=true to the details.do URL at the end of the URL we have - to make http://primo-service.hosted.exlibrisgroup.com/primo_library/libweb/action/search.do?vid=44GOL_V1&showPnx=true to the details.do URL

    It's a link that works and allows me to search but I still can't save any results to Zotero. I still get the same message as before. I checked some of the potential solutions - making sure I'm using Zotero 3 and making sure I have the latest translators in the preferences tab. This doesn't solve the problem I have.

    Is there anything else anyone can suggest?
  • adamsmith
    you misunderstood what I was saying. Zotero takes advantage of a Primo feature that is not implemented in your version of the catalog.
    You can see this easily:
    1. Go to the BNL catalog at http://explore.bl.uk/
    2. Search the main catalog for anything - let's say "Borges"
    3. For the first (or any of the) search result right-click on "Details" and select "Copy Link Location" (this assumes FF, will work in any browser but names may differ).
    4. Open a new tab and paste that - very long - URL into it. At the end of it, add &showPnx=true
    5. Press Return and you'll see an XML display of the reference information of the item - that's what Zotero imports from Primo


    Now repeat these exact same steps for your catalog. In step 5, instead of the XML file, you'll see an error with Firefox complaining about an XML parsing error.

    That's the reason this isn't working and there's nothing we can do about it from the Zotero side - no need to run through any of the Zotero troubleshooting steps.
    Talk to exlibris about how to enable the pnx view of an item - that's possible for every Primo catalog I've looked at, so it should be possible for yours.
  • kevinwilson
    Thanks Adam, that makes sense. I'll ask our systems librarian to contact exlibris and ask them to look into it. Cheers.
  • kevinwilson
    Hi Adam, sorry to bring this back up but Ex-libris has replied to our systems librarian. Their response is below. Sorry, I'm not remotely aware of what any of it means, but will this sort out the problems we've had using Zotero on our Primo catalogue?

    Thanks for your help with this, I really appreciate it.

    Kevin

    Primo is not relevant here, the way it works is Zotero has Transaltors for different sites/application it works with.
    Zotero have a script for Primo in their repository, and they are in charge of maintaince of this script
    More specifically the problem can be fixed preety easily by Zotero by the following change in:
    Primo.js:

    in line 41:
    Change:
    else if (doc.evaluate('//div[@class="EXLContent EXLFullDisplay"]', doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext() ) {

    TO:

    else if (doc.evaluate('//div[@class="EXLCustomLayoutContainer EXLContent EXLFullDisplay"]', doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext() ) {


    OR better:

    else if (doc.evaluate('//div[contains(@class, "EXLFullDisplay")]', doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext() ) {


    and in line 33 :

    Change:

    else if (doc.evaluate('//div[@class="EXLContent EXLBriefDisplay"]', doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext() ) {

    TO:

    else if (doc.evaluate('//div[@class="EXLCustomLayoutContainer EXLContent EXLBriefDisplay"]', doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext() ) {

    OR better:

    else if (doc.evaluate('//div[contains(@class, "EXLBriefDisplay")]', doc, nsResolver, XPathResult.ANY_TYPE, null).iterateNext() ) {
  • adamsmith
    sorry, while the support person is right that this is handled by Zotero in the Primo.js file, none of the rest makes sense.

    Neither of the lines s/he mentions actually exist in Primo.js (and hasn't for months at least)
    see for yourself:
    https://github.com/zotero/translators/blob/master/Primo.js

    The lines that I believe s/he wants us to change only affect whether the book/folder icon appears - which they already do for your catalog.

    This is still the same issue: we need a way to the PNX. I don't really know what to tell you. Maybe try asking about this without mentioning Zotero, which apparently triggers some condescension reflex from their tech support? In other words, just ask something like "I would really like for our librarians to be able to see the PNX data that Primo uses for any item. How can I get to that?" (PNX stands for Primo Normalized XML).
    Exlibris even used to have a FF add-on for that
    http://www.exlibrisgroup.org/display/PrimoCC/Show+PNX+Firefox+add-on
    but that doesn't work anymore because it's not updated.
  • Lennaron
    I'm working with Primo and had a user request last week. I found out that this error could even occur when PNX-export is enabled, when trying to import records without PNX entry. This could happen if a Primo instance imports metadata from other sources and not generating a PNX record for these entries. Maybe it will be an improvement to be more specific in the error log of zotero. I would suggest something like "No data found for this entry!" instead of a very generic import error. It could be easily triggered by the result of the "showPnx=true"-request if the answer is "null".
  • aurimas
    I'm sorry, but you will have to be more clear about what's not working. What steps exactly are you taking to trigger the problem? What's the URL of the Primo catalog?
  • adamsmith
    I know what Lennaron means. Right now, when there is no PNX, Zotero simply fails with
    09:13:07 null
    09:13:07 Translation using Primo failed:
    fileName => chrome://zotero/content/xpcom/translation/translate.js
    lineNumber => 591
    columnNumber => 5
    string => Error: No title specified for item

    (and the first line is only in the debug).
    We could throw something more expressive, at least in the error report.
  • adamsmith
    (note, though, that we have the issue listed under "known translator issues", which is linked to from the generic error message)
  • Lennaron
    @aurimas: Sorry I will show you an example so it might be clear what I meant.

    I'm working at the library catalogue of Humboldt-Universty in Berlin. If you take an entry for our own books like

    http://primo.kobv.de/primo_library/libweb/action/display.do?doc=hub_aleph000201467&showPnx=true

    Zotero will work but if you use an imported entry from licenced publishers like Springer you'll get an error:

    http://primo.kobv.de/primo_library/libweb/action/display.do?doc=TN_springer_bk978-3-531-09026-9&showPnx=true

    Right now I don't if it is a configuration problem caused by the library or a problem by the Primo developers. But I think it's a configuration problem.

    If I know more about this problem I will report it in this thread.
  • adamsmith
    yep, we're aware of this. As I say it's listed under known translator issues:
    http://www.zotero.org/support/known_translator_issues?&#translators_with_major_issues
    If you're involved in the local maintenance of the catalog, do have a look at this:
    https://forums.zotero.org/discussion/31294/possible-solution-for-primo-bug/#Item_4
    which should reduce the error quite a bit once we have it implemented on the Zotero side.
  • Lennaron
    @adamsmith

    Yes it was for a better understanding for aurimas. ;-)

    Anyway it's good to know that there is a kind of fix for that. I will go trough this on monday and I also can get in touch with one of the universities listed in the other thread, cause we're working together on Primo issues, so possibly I can get some advice in setting up this mentioned jsp-file. I'll let you know if situation in our Primo changes on the other thread.
  • adamsmith
    Just to let you know that it looks like you have the .jsp implemented and the book you have at
    http://primo.kobv.de/primo_library/libweb/action/display.do?doc=TN_springer_bk978-3-531-09026-9
    is importing fine with the translator version we're currently reviewing. You can follow the pull request here https://github.com/zotero/translators/pull/617 once it's accepted this will work for you.
  • adamsmith
    ok, the translator fix is now up:
    Your version of Zotero will automatically update within 24hs, or you can update manually using the "Update Now" button in the "General" tab of the Zotero preferences.

    It looks like UCL also has PNX implemented now, so their catalog is working as well.
Sign In or Register to comment.