EBSCO host translators not working

Hi,

Last night I started having problems using the Zotero translators with EBSCOhost databases on one of my computers. The icons for downloading still appear in the address bar, and when I click the red-rimmed Zotero box appears below, but nothing downloads, and eventually the red box disappears.

I have updated translators, and I have tried disabling all other add-ons to see if there was an extension overlap. I hovered my cursor over the address-bar icon, and it reads "EBSCO." I am using Firefox 5.0.1 and Zotero 2.1.8 on a Mac running OS X 10.6.8.

I have tried other sites on this computer, and their translators work fine: NYTimes, JSTOR, Amazon.

I am NOT having the same problem on my other computer--the EBSCO translator works fine there.

Do you have any advice?

Thanks,
Michele
«13
  • Do you use a proxy to access EBSCO from that computer? There are some known issues with library proxies and EBSCOhost, but they can be resolved by the university changing some settings. EBSCO reps know of this and can help the university change its settings.
  • Yes, I do use a proxy - through Hampshire College library. Should I have the library IT folks contact Zotero?
  • I just want to clarify, though -- I am not having trouble with my other computer, which also uses that proxy. Nor did a colleague of mine have trouble this morning downloading citations from an EBSCO database.
  • First, go through http://www.zotero.org/support/troubleshooting_translator_issues and post an error ID if those steps don't address the issue. If it works for other people on the same proxy, it's probably not the proxy issue.
  • I did go through those steps.

    I don't get an error message, and when this happens I am not able to select "Report errors".

    It has only been happening since yesterday. I don't know how you could reproduce it, since it seems to be happening on just my computer. I'm going to restart the whole computer - I've already tried restarting Firefox - and see if that works...
  • Restarting the computer did not fix the problem.
  • EBSCO import now fails *silently*, both from result list and detailed record. A blank red box pops up.
  • sample URL?
  • Please provide the URL and say whether you are using a proxy, so we can see what might be going on in your case.
  • I've made assorted small changes to the EBSCO translator-- you can try this version:
    Please go to http://github.com/ajlyon/translators/raw/master/EBSCOhost.js and save the file to the translators directory of your Zotero data directory (http://www.zotero.org/support/zotero_data).

    I'm not sure whether this will address the problems you're seeing-- but this is certainly the latest version, so we can start with a known up-to-date and known working version.
  • edited July 29, 2011
    Thanks, ajlyon -- I saved the new JavaScript into the zotero /translators directory (double-checked and found the correct one). The error still occurs after restarting.

    I'm using the standalone version of Zotero, if that makes a difference.

    Just pick an item:

    http://web.ebscohost.com/ehost/detail?vid=3&hid=14&sid=8dacf034-3633-4627-8153-929afdf30cb2@sessionmgr13&bdata=JnNpdGU9ZWhvc3QtbGl2ZQ==#db=a9h&AN=60796974

    Permalink:
    http://search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=60796974&site=ehost-live

    Also straight search results like this:
    http://search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=60796974&site=ehost-live

    Sorry to take so long with this -- the one at home would have had the proxy server, which wouldn't have explained the problem very well!

    -Jodi
  • It's probably frustrating to hear this, but I'm disinclined to troubleshoot this for standalone, simply because the standalone version is pretty outdated at this point, and it's long overdue for an update (and that update will bring a whole bunch of new functionality, especially regarding translators). I hope that the team releases another standalone alpha soon-- but this isn't really worth diagnosing with Standalone now.

    If the same happens with Zotero-in-Firefox, let me know.
  • Dear ajlyon,

    I can confirm the EBSCO translator is not working. There is a workaround still, via Export Manager. A brief manual for anybody with the same problem as I have experienced:

    1. Click on "Export" in the menu on the right side of the EBSCOhost page
    2. Choose "Direct Export to EndNote, ProCite, CITAVI, or Reference Manager" (it is the first option in the list)
    3. The citation is retrieved along with two notes (accession number, abstract) and a "direct link".

    Anyway, I have not tried the newest version of the translator yet. I will do it and let you know about the results.

    Kudos to your work, in any case
    JR

    ---

    FF 5.0, Zotero 2.1.8, using ezproxy.is.cuni.cz (Charles University E-resources Portal) and ezproxy.lib.cas.cz (Academy of Sciences Library).
  • Ajlyon,

    Yes, within Firefox,

    * Saving from the results view is reasonable (see picky notes below)
    * From the item view, there's still a problem: I get "known translator error" when trying to save an item:
    http://search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=60796974&site=ehost-live

    A few picky points about saving from the results view, using this example:
    DOI:10.1061/(ASCE)EY.1943-7897.0000035

    --abstract doesn't save correctly, e.g. get "Copyright of Journal of Energy Engineering is the property of American Society of Civil Engineers and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)"
    --call number is saved as 0000
    --Asian names aren't parsed correctly as first/last
  • The abstract issue is fixed in an updated RIS import translator that I didn't remember point you to: http://github.com/ajlyon/translators/raw/master/RIS.js

    Install it the same as the EBSCO one.

    I'll look into the saving error, call number, and names.
  • Awesome, thanks Avram!

    Do you know if there's a way to set standalone to get translator updates?

    -Jodi
  • The saving error seems to be an issue on EBSCO's side-- I've notified the developer there who has been working on Zotero support.

    I don't see the call number issue here-- does that appear on every item? Or just specific ones? The translator doesn't specify a call number, so it's odd that you'd see anything.

    For names, it can be hard to make them work right, in large part because the incoming data often doesn't specify the name ordering. Still, can you provide example records? I'll see what we can do.

    As for standalone updates, they'll work when the next version comes out. In the meantime, I highly recommend using Zotero-in-Firefox with Chrome/Safari instead, as described here: http://ideophone.org/zotero-for-chrome-and-safari/
  • edited August 3, 2011
    Dear Avram,

    I have tested the latest version of the EBSCOhost translator as I had promised. Unfortunately, I still need to use the workaround I had described previously. The RIS translator works fine now, abstract is being filled in the proper field as it should be.

    The links I have tried:

    EBSCOhost Example 1;
    EBSCOhost Example 2;
    EBSCOhost Example 3;
    EBSCOhost Example 4;
    EBSCOhost Example 5;

    ---
    Edited: I have just realised sometimes the latest translator actually worked. See the following URLs:

    EBSCOhost Example 6.

    ---

    Thanks again for all your work.
    JR

    P.S.: I would just like to make sure of me updating the translators properly. I do it by opening the .js file in Notepad++ and copy-pasting the new code in it, overwriting the previous one. The reason is that when I just downloaded a file directly through FF, it became either a text file, or a html file. This was a nonsense, of course. I did not realise that one had to change the file type when saving from FF (to prevent .js from becoming .txt). Neither did I realise how easy it was to download everything from translators repository in one zip file. I have decided to write here about it to prevent somebody following my original, overcomplicated "method".
  • The saving issue is affecting me too. Hopefully the EBSCO developers will have an answer soon.
  • The EBSCO issue is apparently that the data it needs isn't available on all pages. Quoting the EBSCO developer:
    It is not an architecture change. Following a persistent link to the site without scope=site on the persistent link query string will result in the Add to Folder link containing the JSON not being there on a page. This link will work:
    http://search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=37589798&lang=cs&site=ehost-live&scope=site
    I'm not sure how you're getting the links you posted, but if you place &scope=site at the end of them, they'll work. Search results still work for you, correct?
  • Dear Avram,

    Thank you for your reply and all your endeavour. The links I have posted here were generated by EBSCOhost "Permalink" instrument (I have thought it was the way how to make them work for anybody, I was mistaken). You are absolutely right my links do not work (citations are not retrievable), your one does (citation is retrievable). Even the format of the entry is different, as autor, publication and other fields are clickable on the page linked by you.

    I am not sure what exactly you have meant by answering if "search results still worked for me". Anyway, I have tried to search the article linked by you again and to retrieve the citation, and - what a surprise - it worked. And so it was in case of the other articles, via both proxies I use. I do not understand what has changed, because yesterday I did the same as today and had to use the workaround. So, kudos to you!

    Anyway, there are still two "small details" I would beg for to be solved, some day, I hope. First, the date format, at present the retrieved entry is usually "SeasonYYYY YYYY" ("Winter1999 1999"), in other words, the year number is repeated. Secondly, the URL of the referred record is not retrieved to Item Info, which is quite incovenient when one needs to actually cite it (it is similar to what I wrote in another thread on ProQuest). Therefore, one has to manually fill it in. Ironically, when using "Export" the URL is retrieved into Item Info automatically. However, in such case you have to download manually both html and pdf attachments. So, it is a trade-off. Either you have attachment, or you have the URL. In both cases, one has to intervene. Please, take these remarks as suggestions for future improvements, I understand everything cannot be perfect instantly. And I acknowledge the voluntary work you do for the community.

    Regards
    JR
  • The double date issue is fixed in the latest version of the EBSCO translator (posted Aug 1 16:11:49 2011 +0400). Install again from Github -- go to http://github.com/ajlyon/translators/raw/master/EBSCOhost.js and save the file to the translators directory of your Zotero data directory (http://www.zotero.org/support/zotero_data).

    The URL issue is exactly what I said for ProQuest-- as it's not a canonical URL, we don't save it in the URL field.

    Rád jsem vám pomoc. Doufám, že EBSCO-vý tlumočník pro vás tady dost dobře funguje.
  • edited August 2, 2011
    Please go to http://github.com/ajlyon/translators/raw/master/EBSCOhost.js and save the file to the translators directory of your Zotero data directory (http://www.zotero.org/support/zotero_data).

    This version should work with all the permalinks given above. Kudos to Josh at EBSCO for finding a way to support pages of this sort as well.
  • Done, tested, flawless. Kudos again, both to you and Josh. Thanks a million! I think I am not the only one who will benefit from this.

    As for "the URL thing", I agree. Partially. As I have written in the "ProQuest thread", in order to respect the prevalent ISO 690 understanding in my country, I would again vote for the possibility of the EBSCOhost Accession number being retrieved and filled in the "Call Number" field. Nevertheless, I respect the fact my view may be in minority.

    Anyway, it was quite funny to read the words "translated" into Czech. A bit more Czech version would be: "Rád jsem Vám pomohl. Doufám, že překladač EBSCO je z Vašeho pohledu dostatečně funkční." (meaning "It was a pleasure to help you. I hope the EBSCO translator works fine for you.").

    Regards
    JR
  • Oh -- the translator thing was my weak attempt at a pun in Czech. Apparently my Czech is a little rustier than I thought -- I was at Charles University last summer, but haven't had much practice since.

    I've posted an updated version to GitHub that now includes the accession number.
  • ajlyon, your Czech puns are the main reason I visit this forum. :)
  • edited August 3, 2011
    Dear Avram,

    Thanks for all the improvements. I very much appreciate and admire you are so committed to making translators perfect. I have tested the latest version of the EBSCOhost translator, and, as for AN, it works fine. However, there is a problem with attachments.

    I have tested these ANs: 52515109, 4370815, 2237262, 9704044807. In the first case, the "EBSCO Full Text" and "EBSCO Record" attachments were downloaded. When I clicked on the first one there was a new tab opened with a short message:
    Sorry, we are unable to retrieve the document you requested.
    When I clicked on the second one there was a new tab opened with EBSCOhost login page (which was problematic as well, see below).

    After testing the other ANs, I realised the "EBSCO Full Text" attachment was not downloaded properly in case there was no pdf to retrieve (quite logical - a solution might be to make the translator not to force download when there is no pdf). In case there was a pdf, it was downloaded properly.

    Nevertheless, with the "EBSCO Record" attachment there was no change. It was never retrieved as a snapshot, no matter if there was only an abstract or the whole fulltext in html. Moreover, the link retrieved along with the "Record" attachment was strange. When I clicked on it, the EBSCO login page was opened (which I regard as a problem of its own). When I tried to get to the EBSCOhost internal record by manually adding the proxy name into the link above in the address line, it did not allow me into the database, but directed me to a page with this error message:
    No results were found.
    EBSCO Support Site
    New: EBSCOhost iPhone/iPod Touch Application
    Privacy Policy
    Terms of Use
    Copyright
    © 2011 EBSCO Industries, Inc. All rights reserved.
    Then I tried to change the EBSCOhost language settings from Czech to English (do not ask me why, at first I did not realise at all it might be somehow related to the problem). And? Yes, the link then worked, in case I manually added the proxy name. However, the "Record" still was without any abstract or html fultext retrieved (although there was an abstract and/or html fultext in the respective record).

    I am very sorry my report is negative again. However, there is a positive side of it, because it helped me to discover there actually was a problem already with the previous version of the translator ("EBSCO Record" attachment with html full text or abstract page not retrieved properly), but I overlooked it. When writing this report, I went through the items from the last two days, and indeed, the "EBSCO Record" attachment was just a link, moreover, corrupted due to language settings as described above, and perhaps due to other reasons, as several links I went through included the very same AN 2237262, although they were not related to the same EBSCOhost internal record. Next time I have to do more thorough testing. I apologise.

    Regards
    JR


    Offtopic Addendum: It is great to read you have visited my alma mater! Should you come to Prague again, do not hesitate to contact me. In case you would like to continue learning Czech, I am ready to help as well.

    ---
    Edit: Clickable links added.
  • It sounds like perhaps your proxy settings need to be checked. Can you post the host and scheme for your proxy, from the Proxies tab of the Zotero preferences?
  • edited August 4, 2011
    Dear Avram,

    Thank you for your response. I do not have the proxy redirection enabled. Do you think it still could be the source of the problem? I used redirection in the past, but then I realised it led to complications, so I disabled it (sometimes I needed to log in the same database through a different proxy, the redirection forced me to use the proxy I did not want to, so I had no other option than to disable the redirection completely, and do everything manually). However, I kept the schemes in the settings, so perhaps I should try to delete them all. I will report on this in about twelve hours. Now I really need to have a rest.

    Regards
    JR
  • I'm having similar problems with ebsco - I tried updating my translators, and then using the newer version of the translator, and got a 'silent' fail - eg the empty red box and nothing entered in my list of references in zotero. Wiley works fine, amazon is fine too.

    I do use a proxy, but have since using Zotero, and have had no problems.

    So, there seems still to be an issue with the EBSCO translator
Sign In or Register to comment.