Can't sync due to fields being too long

H_Koechlin

Hi, I keep receiving the same error message, saying that my library can't be synced because the creator name in one of my items is too long. It advises me to shorten the field and sync again, which I do, but then it will just come up again. I also tried deleting and re-adding the respective item, but it doesn't work either. Every time I try there is a new item that has a field too long - I've been using Zotero for a few years and I never had this problem before. I'm a little worried, as I haven't been able to sync in a while now...
Anyone ever had the same problem?

dstillman

The message really means what it says, and the same length restrictions have been in place for years. Note that if you delete an item instead of simply correcting it, you need to empty the trash as well. But there's no need to delete an item if you want it — you just need to fix the field that's too long.

dstillman

Basically, though, you can see the data in the item it identifies, so there shouldn't be any mystery as to what's causing this. If all the affected items are from the same site, there might be a problem with a translator pulling in incorrect data, and if you saved a lot of items from that site you'd have a bunch of items that you'd need to correct. If that's the case, provide an example URL so we can take a look.

H_Koechlin

Thank you for your answer. Most often it is the Editors field that is too long, but I don't find it to be the same site everytime... One of them has been from this site: https://www.journals.elsevier.com/journal-of-affective-disorders
If I correct / shorten the field I think is too long, I get the same message again for a different item.
Thanks a lot for your help.

dstillman

We would need a specific URL that you saved from, which should appear in the URL field for the affected item. (A DOI night also suffice.)

adamsmith

Could you say exactly how you're importing? If I just import articles from that journal, it doesn't import any editors (nor should it, I think) and everything looks correct.

H_Koechlin

Thank you for your comments and please excuse my late reply. Here is the DOI of the latest item that is too long: 10.1016/j.chiabu.2017.07.021, it doesn't have an URL.
I always import using the Zotero icon in the Firefox browser.

H_Koechlin

Hi there, I keep having the same problem. Any suggestions on how to solve it? I'm afraid my library won't sync correctly and as it is very important and valuable to me, I would rather avoid stuff being lost. Thanks a lot

adamsmith

I still don't know how you're importing that. Both by going to the page on Sciencedirect via the DOI and by importing the DOI in Zotero via add-by-identifier, I don't get any editors nor any other string that'd be too long for syncing.

If you get overly long author strings somehow, you'd just delete or reformat them.

Edit: and if you keep getting that message, I'm afraid you'll have to fix the length of fields for all items until you stop getting it. There's not quick fix here.

drameadows

I am having the same problem. I think it has arisen due to incorrect import from mass results files identified in a systematic review and exported as .bib or .ris files for import into Zotero. It is creating an unnecessary Editor field for journal articles and populating it with every name possibly in the issue? references? i don't know. dozens and dozens of names. in some cases, it is also putting much of this stuff into the Issue field (for journal articles). It also sometimes happens for book chapters, where there *would* be an editor, but the fields are enormous with loads of alphabetised names, again, possibly every name associated with the book, e.g. other authors?

I have a feeling this is from Web of Science, but I'm not sure how to tell what source these references came from, as I searched multiple databases and imported results, before de-duplicating. No URL is available in the records.

Original poster, the reason you are getting a new one when you fix an existing problem is because the problem has occurred across numerous records. I have over 600 that need fixing and doing it piecemeal by hand is extremely time consuming!

adamsmith

This is very likely from mass import of a broken import file. We've e.g. seen exports that put all authors/editors in a single line in RIS, which would cause this.

I'm afraid I don't have a quick solution. Do you have a list of databases you used? I'd be surprised if this was WOS, which typically does a good job with standardized export. The worst offender we have seen is Ovid, which does this sort of thing _all_ the time.

@zuphilip -- wondering if we should build in some sanity check into RIS export that treats AU - as a list of authors if there are more than 2(?) commas or semicolons.

drameadows

I did use Ovid. Over 25000 results (in about 80 .ris files). S***

zuphilip

@adamsmith Usually the comma is used in the AU/ED field for splitting the lastname from the first name. However, when there more than one comma in these fields then we could follow your suggestion to treat this as a list of creators. That sounds okay to do for me. Do you want to create an issue about it?

adamsmith

Yes, exactly what I had in mind. Ticket here: https://github.com/zotero/translators/issues/2006