Mac Firefox Connector diacritics

Thanks for any help someone can suggest on this to a Zotero novice.

Using the 5.0.61 Zotero Connector for Firefox (72.0.1) on a Mac (10.14.6), the diacritics (in this case, é è and ç) are coming through as question marks in my Zotero Standalone (5.0.66).

Is there a way to fix this?

Thanks!
  • That would be a problem with a specific site. What's the URL where you're seeing this?
  • Ah, ok! These were added from WorldCat (https://firstsearch-oclc-org.ezproxy.morris.umn.edu) through my university, @dstillman.
  • Is that something that can be fixed?
  • edited January 14, 2020
    On sites without static URLs, we need the exact steps to reproduce (e.g., search terms).
  • edited January 14, 2020
    @ecbruce: Never mind, I was able to find a record with diacritics in the title and reproduce this ("The Chrysanthème papers"), so don't worry about the example. We'll take a look.
  • @adamsmith: The Refer file being returned here is ISO-8859. Do you remember the background for removing the responseCharset argument in the FirstSearch translator? (That was back in 2016, so not sure if something has changed more recently.)
  • yeah, the background was that OCLC had implemented UTF-8 consistently for Firstsearch, making the awkward workaround of testing for the charset and switching around unnecessary. They went back to 8859? WHY???

    I also vaguely remember that there was a distinct downside to testing for the charset (maybe the test was ignored when saving via server?), which is necessary because e.g. Chinese references weren't imported when fixing the charset to 8859, but memory there is hazy.
  • Thanks for looking into this, @dstillman and @adamsmith!
  • edited January 15, 2020
    I have a work-around. Although it requires several extra steps. I copy text from WorldCat and paste it to a text editor (that I've set to use UTF-8) and then copy/paste to Zotero from the editor screen. I'm currently using BBEdit on Mac OS but I've had success with other editors on both Mac and Windows systems (before the problem was initially resolved a year or two ago).

    OCLC regressed away from UTF-8 a couple of months ago. I tried to reach my contact there but the had person resigned. My university library sent a technical services request but (without going into details) received only frustrating non-answer replies. I gather that the current tech support gatekeepers at OCLC don't understand the issue and simply dismiss the problem instead of referring it to someone who can solve it. This may be a time for people reading this forum to talk with their librarian and have her/him contact OCLC. Squeaky wheel...

    edits for clarity
Sign In or Register to comment.