import from endnote/ Umlaute

Gitty K · January 24, 2008

Hello there,

I have tried to import my Endnote 6 library but no matter what format I use German "Umlaute" (ö,ä,ü) are not converted properly. I have exported my endnote libary into several formats but the problem persists with all of them. Bibtex has been mentioned in earlier discussions here as a solution but doesn't work for me either. Is there any solution to this?

Thanks,

Jochen

jgr · January 25, 2008

I had the same problem using Endnote 5. Probably the problem is that Zotero uses Unicode whereas Endnote 5 doesn't. My workaround was
- replace every Umlaut with some unique letter combination (eg. "u#" for "ü") in Endnote
- export
- open the export file with an editor
- change to Unicode
- re-replace the letter combinations
I guess one could write a macro doing that. I can't.
Quite complicated, maybe someone comes up with a better idea.

Gitty K · January 26, 2008

I have exported everything into a .txt-file with endnote. All Umlaute are correct in the .txt file. To give it a try I have replaced them with Umlaute from the special character table (which as I gather should be unicode then?), but that didn't help either.
Plus, before I replace them in Endnote, I can just as well correct them manually in Zotero. Isn't there a better solution to this?

dstillman · January 31, 2008

Unicode support for RIS has been discussed quite a bit before on these forums, and there's a ticket to improve the translator. (Summary: no RIS specification that we're aware of actually allows use of Unicode, but many programs emit it anyway and we'll likely support it soon.)

At the moment you might try converting the file to the IBM850 character set in a text editor and importing that.

Gitty K · January 31, 2008

thanks for this. How do I do that?

I am a bit surprised that this seems so difficult as the website claims that Zotero fully supports all languages (btw. French accents don't work either)

I also wonder if I need to change some Firefox settings, although in almost all cases FF displays "Umlaute" correctly (rare exceptions not withstanding)

Maybe it helps if I post some example references here:

Bibtex:
@book{
Author = {Glinka, Hans-Jürgen},
Title = {Das narrative Interview. Eine Einführung für Sozialpädagogen},
Publisher = {Juventa},
Address = {Weinheim, München},
Year = {2003} }

and RIS:
TY - BOOK
AU - Amendt, G.
PY - 1989
BT - Natürlich Anders. Zur Homosexualitätsdiskussion in der DDR
CY - Köln
PB - Pahl-Rugenstein
ID - 60
ER -

TY - JOUR
AU - Beaudin, C.L.
AU - Chambré, S.M.
PY - 1996
TI - HIV/AIDS as a chronic disease: emergence from the plague model
SP - 864-706
JF - American Behavioral Scientist
VL - 39
IS - 6
ID - 274
ER -

TY - JOUR
AU - Braun, S.
PY - 2001
TI - Bürgerschaftliches Engagement - Konjunktur und Ambivalenz einer gesellschaftspolitischen Debatte
JF - Leviathan
VL - 29
IS - 1
ID - 98
ER -

dstillman · January 31, 2008

How you change the character set of the text file depends on the text editor you're using. You might look in the Save As dialog.

Zotero supports Unicode internally, but for importing and exporting, different formats require different handling. Not to rehash the issue, but we originally coded RIS support to the RIS specification on the RefMan site, which only allows "IBM Extended Character Set," not Unicode.

Other import/export formats in Zotero support Unicode characters just fine, and you can also enter such characters directly into any field in Zotero. Firefox settings are irrelevant.

dstillman · January 31, 2008

For Unicode BibTeX support, if you're using 1.0.2 you can get the version of the translator that support it by doing a Reset Translators and Styles in the Advanced pane of the Zotero prefs. (We can't distribute it automatically to Zotero 1.0.* clients via the repository because of its size, but this will be fixed in Zotero 1.5.) There's a problem importing 'F' characters in that version, however, so you might want to wait for 1.0.3, which should be out very shortly, and use Reset Translators and Styles after installing that.

Gitty K · January 31, 2008

I've tried what you said, reset tranlstors and styles, restarted the FF, saved the txt-file (in word) in unicode (utf-7), (which was the only one that produced a readable text), but Zotero says it can't find a translator. Before it could and it seemed to be in ANSI (then the import worked...)

I guess I'll have to retype all Umlaute in my 500 or so references :-(((

dstillman · January 31, 2008

Again, you need to save the RIS file as IBM850 for the current translator. If you want to do this, you'll have to find a text editor that supports that character set or wait for an update to the RIS translator.

I know that jEdit can do what you want, though it's a bit overkill for the job and might be a bit hard to figure out. (After opening the file in jEdit, you double-click on the character set down in the status bar, change the setting, and resave.)

Gitty K · February 1, 2008

Thank you very much for this last tip with jEdit!! I finally got it working!!

(in my previous post I talked about trying it with a Bibtex file. I had understood that that would support Unicode, but there were still problems, but it doesn't matter anymore now that it worked out... Sorry for being a bit complicated...)

Hucklenbruch · September 22, 2009

I do realize that this is an old thread. However, it might be helpful for some MAC-Users that are confronted with the same problem. What worked for me is the following:
Open the RIS-file in Smultron (a free, versatile text editor, http://www.macupdate.com/info.php/id/15114/), change the text encoding to: "Western (Windows Latin 1)". Save the file. Import to Zotero. Done.

dstillman · September 22, 2009

Just to be clear, the information from 2008 above regarding IBM850 is long outdated. Zotero now supports UTF-8 RIS import, so you should be able to just save RIS files as UTF-8 (with a BOM, if possible) and Zotero should handle them fine.