RIS Files: Why “é” Becomes “é”

Geremia

My RIS file from WorldCat has strange characters, like "é" instead of "é" etc.

Here's what worked very well for me for fixing these files, courtesy this excellent description of character encoding and how to convert between different encodings:

Suppose you want to fix these strange characters in your "WorldCat_2893657.ris" RIS file. Issue this command on Linux/Unix, which will make a corrected RIS file, named "WorldCat_2893657_2.ris" in this case:

iconv -f utf8 -t iso-8859-1 WorldCat_2893657.ris > WorldCat_2893657_2.ris

Essentially, you need to convert from UTF-8 to ISO-8859-1.

adamsmith

(note that we automatically detect and fix this if you use the URL-bar icon to import from WorldCat)

Geremia

@adamsmith: Yes, that's true, but it doesn't put into the notes field all the extra tags for which there are no corresponding Zotero fields, nor does it seems to recognize the difference between books, theses, articles, etc. I've found RIS importing much better than the the URL-bar "Open WorldCat" (which rarely works for me) and "COinS" (which works and indeed does get the correct character encoding, but is limited like I said). I have 3.0.14 on FireFox 19.0.2 on a Linux system. Perhaps there's a bug in my URL-bar importing?

adamsmith

that'd be odd since we use the same RIS in the translator (i.e. the URL bar icon) - we just get it automatically. What would be an example?

Geremia

I forgot to mention that sometimes when I use the URL-bar import, the first item in my WorldCat list only includes the title with no author, and there's an HTML entry of my WorldCat list webpage attached to that item. Oddly, I only see this behavior sporadically.

Also, when I go to one of my WorldCat lists and choose "Save to Zotero using 'Open WorldCat'", the dialog box asking me which entries to choose either never shows up or takes a very long time to show up. However, the dialog box for "Save to Zotero using 'COinS'" opens immediately every time.

And the import speed is incredibly slow for my ~100 item WorldCat lists; Zotero/Firefox uses nearly 100% of my CPU for a few minutes. I have no idea why. This doesn't happen when importing a RIS file.

adamsmith

that list imports fine (and quickly) for me from the URL bar using Open WorldCat with all information that's in the RIS.
What's your exact set-up? FF version, which linux, which Zotero version?

Geremia

Zotero 3.0.14 on FireFox 19.0.2 on a Linux system

Also, for large lists, I get the same issue with at least one item (the first?) having no author when importing from RIS, too. (For the list I linked to above, this was for the entry "Philosophy of Nature a Guide to the New Essentialism".)

thanks

adamsmith

the first item being this?
http://www.worldcat.org/title/physical-system-of-st-thomas/oclc/667874089?referer=list_view

adamsmith

ah OK - there is no author in the data for the Philosophy of Nature book.
Generally, Worldcat's data isn't all that great, unfortunately. That's particularly true for ebooks, which you seem to be saving a lot. The regular entry for the Ellis book should import fine with author:
http://www.worldcat.org/title/philosophy-of-nature-a-guide-to-the-new-essentialism/oclc/248451033?referer=di&ht=edition

Geremia

Oh, so it was a WorldCat issue! Thanks

Also, can you please explain to me the difference between "Open WorldCat" and "COinS" importing? Thanks

adamsmith

Open WorldCat is Zotero's dedicated translator for WorldCat. It automatically downloads and imports the RIS entry for the selected WorldCat entry(s). In most cases that should be the best available data.

COinS - Context Data in Spans - is bibliographic data embedded by WorldCat (and many other library catalogs and some other sites) into the webpage that Zotero can pick up and read.