Unneeded comma in Japanese names (new bug in 5.0.88?)

nathan.hopson

Zotero 5.0.88
macOS 10.15.5

An unneeded comma is being inserted in Japanese names for sources with the language set to "Japanese."

Examples:
Unno, Hiroshi. Hyakkaten no hakubutsushi. Ātsu ando Kurafutsu, 2003.
Yamamoto, Yūzō. Kindai Nihon no shisō zendō to kokumin tōgō. Azekura Shobō, 2011.

I believe this is an artifact of either
1. the new update
2. a difference in the way "Japanese" and "ja" or "ja-JP" are processed as language tags
3. both

For sources set to "ja," no comma is inserted:
Nakata Kyōko. “Kyōdō kondate zairyō haikyū ni tsuite: Ichi jukyūsha no kansō.” Eiyō to ryōri 7, no. 11 (November 1941): 62–65.
Inoue Kaneo. “Iwashi wa atama ga daiji.” Yomiuri Shinbun, March 13, 1942.

If there's not a quick hot fix coming, is there a way to globally replace "Japanese" with "ja" for the correct references?

EDIT: I know there's a way to search for all the offending items. I just don't see a non-manual way to replace once I've done that.

Thanks as always!

dstillman

@fbennett: Is this a regression in citeproc-js? 5.0.88 included an update from citeproc-js 1.2.18 to 1.3.22.

fbennett

@dstillman: I'll test this out, but I don't think it's a regression. As far as I remember, nothing has changed around name joins in years.

Hi @nathan.hopson! Feel free to contact me direct on multilingual issues if you like (as we're neighbors). Will post back here after looking into the issue.

fbennett

@nathan.hopson: Oh, wait. Are the entries that are adding the comma set to "Japanese" (the word) in the Language field? If that's the case, it should be a BCP47 language code (either "ja" or "ja-JP" should work). We fudge "English" by lowercasing and trimming to the first two characters to read it as "en." We did that because there are bazillion records set to English; but we can't do that for all languages, because many language codes are not the first two characters of the (English) name of the language.

The right thing to do, eventually, will be to set the Language field from a controlled list of the world's languages, but that won't happen anytime soon. Assuming that this is the cause (language name in the field), send me mail and we can connect over video. I can walk you through the steps to do a bulk find-and-replace.