Item Type error - everything imports as web page when they are articles

Stan

In the past few weeks when I import journal articles into Zotero, they end up imported as web pages and then I need to manually search and fill in volume, issue, page number etc

I think it happens on both EBSCO and ProQuest. Neither of those two will import directly to Zotero for me so I need to first export them to a folder and then from folder view export them as a Reference citation. That always worked in the past - ie a journal article was captured as such but now they are always captured as web pages for some reason.

noksagt

If the URLs generated on failing sites can be used by other people, please include them. If the URLs are session-dependent, please provide a link to the page where you start your search & your search terms, so that people could try to replicate this.

Stan

Not sure that I can. I access the databases from a VPN of my university - through the University through to the database supplier - each of whom has another level of authentication. So here is an example of a typical search but I doubt it helps. This problem is recent - but there have been updates to Firefox and Zotero. FYI I am not yet on Firefox 3.6 because I am not sure if all the issues have been ironed out with Zotero but should I in the first instance try that update?

https://extranet.cranfield.ac.uk/,DanaInfo=proquest.umi.com+pqdweb?RQT=512&querySyntax=PQ&searchInterface=1&SrchMode=2&TS=1285687722&moreOptState=CLOSED&clientId=3224&h_pub_title=journal+of+interactive+marketing&h_pub_title1=&h_pub_title2=&h_pub_title3=&h_pub_title4=&h_pub_title5=&h_pub_title6=&h_pub_title7=&h_pmid=&h_pmid1=&h_pmid2=&h_pmid3=&h_pmid4=&h_pmid5=&h_pmid6=&h_pmid7=&SQ=journal+of+interactive+marketing&FO=PUB&OP1=AND&SQ1=customer+segmentation&FO1=TEXT&OP2=AND&SQ2=&FO2=CITABS&OP3=AND&SQ3=&FO3=CITABS&OP4=AND&SQ4=&FO4=CITABS&OP5=AND&SQ5=&FO5=CITABS&OP6=AND&SQ6=&FO6=CITABS&DBId=3&date=ALL&onDate=&beforeDate=&afterDate=&fromDate=&toDate=&Oppubtitle=AND&pubtitle=&Opsubject=AND&subject=&Opcompany=AND&company=&Opname=AND&name=&Opgeo=AND&geo=&Opcc=AND&cc=&Opsic=AND&sic=&OpSF=AND&SF=any&OpAT=AND&AT=any&revType=review&revPos=all&STYPE=all&sortby=REVERSE_CHRON&PageSize=10

adamsmith

As noksagt said:

If the URLs are session-dependent, please provide a link to the page where you start your search & your search terms, so that people could try to replicate this.

i.e. which database, which search terms?

Edit: I don't think there ever were any issues with Zotero and Firefox 3.6 - on the contrary, there have been some issues with 3.5 in the past.

ajlyon

Are you using Zotero 2.0?

Stan

Let me try to respond:

1. The databases I use are EBSCO Business Source Premier and ABI / INFORM (ProQuest) and these are accessed via my University's network. I am not sure if there is a built in "Athens" identification to further complicate matters.
2. Search terms - almost any but most recently I was looked at articles using a variety of terms such as segmentation, customer selection, methods overall and in the Journal of Interactive Marketing for which I am preparing a manuscript.
3. OK then I will upgrade to 3.6; I always wait for it to be stable as the discussion fora do have threads of people with some issues whenever FF gets updated
4. I have the latest Zotero version and yes I do use the 2 series solution

Stan

New Firefox and now Zotero won't work, something about a python extension not working; that is why I hesitate with upgrades to firefox. I will look online to see what to do but if you are online now, it would be great if you point me to the right solution.

ajlyon

For PythonExt, see http://www.zotero.org/support/word_processor_plugin_troubleshooting#component_loading_error

Stan

Thanks I have updated now; I forgot to first uninstall the old and then to remember that I needed to download within Firefox not Safari. I got there in the end.

However I still have the problem. I tested with a download off ProQuest an article by E Malthouse titled: Ridge regression and direct marketing scoring models and it came up as a web page still

noksagt

I cannot replicate this. Your page of search results:
http://proquest.umi.com/pqdweb?RQT=512&querySyntax=PQ&searchInterface=1&SrchMode=2&TS=1285687722&moreOptState=CLOSED&clientId=3224&h_pub_title=journal+of+interactive+marketing&h_pub_title1=&h_pub_title2=&h_pub_title3=&h_pub_title4=&h_pub_title5=&h_pub_title6=&h_pub_title7=&h_pmid=&h_pmid1=&h_pmid2=&h_pmid3=&h_pmid4=&h_pmid5=&h_pmid6=&h_pmid7=&SQ=journal+of+interactive+marketing&FO=PUB&OP1=AND&SQ1=customer+segmentation&FO1=TEXT&OP2=AND&SQ2=&FO2=CITABS&OP3=AND&SQ3=&FO3=CITABS&OP4=AND&SQ4=&FO4=CITABS&OP5=AND&SQ5=&FO5=CITABS&OP6=AND&SQ6=&FO6=CITABS&DBId=3&date=ALL&onDate=&beforeDate=&afterDate=&fromDate=&toDate=&Oppubtitle=AND&pubtitle=&Opsubject=AND&subject=&Opcompany=AND&company=&Opname=AND&name=&Opgeo=AND&geo=&Opcc=AND&cc=&Opsic=AND&sic=&OpSF=AND&SF=any&OpAT=AND&AT=any&revType=review&revPos=all&STYPE=all&sortby=REVERSE_CHRON&PageSize=10
and the page for the article:
http://proquest.umi.com/pqdweb?SQ=Ridge+regression+and+direct+marketing+scoring+models&DBId=-1&date=ALL&onDate=&beforeDate=&afterDate=&fromDate=&toDate=&pubtitle=&author=&FT=0&LA=any&AT=any&revType=review&revPos=all&STYPE=all&sortby=REVERSE_CHRON&RQT=305&querySyntax=PQ&searchInterface=1&moreOptState=OPEN&TS=1285702634&h_pubtitle=&h_pmid=&clientId=30032&JSEnabled=1
contain multiple items. The translator in the addressbar (folder icon) reveals the ProQuest translator is being used. Clicking and choosing articles leads to items being imported as journal articles.

Stan

To complicate it more. Using the Ebsco database I do not have this problem. I can export a reference to folder and export again to Zotero. Today the function to export to Zotero from the icon in the addressbar works too. Normally it does not. With ProQuest I cannot export from addressbar (never could) because of a known translator error. If I move the reference to a folder and then export to Zotero it still comes up as a web page.

I will use EBSCO as my first choice database! But still annoying.

adamsmith

I can confirm that Proquests RIS output fails in the way Stan describes -

Specifically for you, Stan, I'd suggest you focus on figuring out why proquest isn't working for you - it does for me and noksagt. So try disabling other FF add-ons, restart FF and try again. If it still doesn't work, open a new thread with a Report ID for the failing Proquest translator, including a specific list of search terms you used.

For the rest of us - I think figuring out how the RIS translator fails would be good because it seems like potentially a more general problem - and I can't for the life of it track down what makes it fail. The RIS looks clean and even if I change some lines to make it look more like Zotero's own RIS it still imports as a website, although the entries clearly begin with
TY - JOUR

noksagt

The translator does not like the blankline that ProQuest has after the TY line. Editing the translator to trim whitespace of the TY-tagged values fixes the type detection.
https://www.zotero.org/trac/attachment/ticket/1727/

dstillman

The updated translator (with a small tweak) is now available. Your copy of Zotero should auto-update within 24 hours, or you can update manually by clicking Update Now in the General pane of the Zotero prefs.

Thanks to noksagt for the patch.

Stan

Thanks to everyone - albeit I can not really understand fully the conversation! But what I takeout is that the translator built into Zotero should now read the ProQuest output "better" and if I update my Zotero in 24 hours, I should have another go.

FYI I have no add ons, and the only non Zotero related extension is "screen grab"

Your help is much appreciated. After midnight now in London so I will close down.

dstillman

if I update my Zotero in 24 hours, I should have another go.

Reread my previous message. You can update manually whenever you want.

noksagt

And please note that this merely fixes the import of RIS files generated at ProQuest.

As adamsmith reminded you: there is a ProQuest-specific translator that works for him and works for me. Because it does not work for you, we should still troubleshoot that issue.

Stan

Updated manually and that seems to resolve the problem. In ProQuest, I exported a journal reference to the folder and from folder to Zotero and now, Zotero categorises it correctly as a journal article rather than a web page. Thanks.

I cannot export the reference directly to Zotero from the little icon on the right hand side of the addressbar - I get an error message that this is a known translator issue. I have always had this and, when I brought it up a year or two ago, Dan suggested it is due to the particular nature of the proxy I use from my University.

So we have restored the functioning to where it was and that is great. It is but a small inconvenience to have the added step of exporting references via a folder.

Stan

Interesting new twist on this. I did a search on EBSCO that generated two items, when I click on the icon in the address bar, it automatically downloads the citation of the first - no list appears allowing me to choose which.

The search was in EBSCO global, for author Mccarty and publication "Journal of Business Research"; that generates articles from 2007 and 2004