Issues Importing BibTeX File from Sente

skamen

I would like to migrate my library from Sente for Mac to Zotero and have run into a few issues. As suggested by others on this forum who have used Sente, I exported my Sente library as a BibTeX file and imported that file into Zotero. Importing that file successfully added the references from my Sente library and the PDF attachments those references included to my Zotero library. However, many of the fields in those references did not import correctly, or as I would like them to be imported. Many of the item types did not import correctly (such as newspaper article being classified as books) and newspapers and magazine references, in particular, using the publication title as the article title. Does anyone have any suggestions as to how I might change import settings in Zotero or export the file different from Sente (as a different type of file perhaps?) in order to fix the problems problems with those references?

adamsmith

I don't know anything about Sente, but if you post individual bibtex entries (open the file with a text editor like TextEdit) that don't import correctly here, I'm happy to take a look.

Does Sente export RIS? That's more standardized and might work better.

skamen

I just checked and, unfortunately, Sente does not export RIS. Here's the text for the BibTeX file for one reference:

@misc{Galbraith:BostonGlobe:2012,
author = {Galbraith, John Kenneth},
title = {Boston Globe},
uuid = {BBF016EC-B252-48B9-BECB-68392A3EEFA3},
pages = {1-1},
month = {8},
year = {1969},
sentelink = {file:///Users/skamen/Documents/Sente%20Libraries%20/Research%20Library%20.sente6lib/Contents/Attachments/Galbraith%2C%20JK/2012/How%20To%20Control%20the%20Military%20II%20%232.pdf,Papers,PDF Full Text},
url = {http://search.proquest.com.proxy-um.researchport.umd.edu/docview/506273116/139DABED93C48941619/21?accountid=14696},
keywords = {Read;Primary Source},
Citation_identifier = {Galbraith 2012p},
publicationStatus = {Published}
}


I don't know why Sente is listing "Boston Globe" as the title rather than as the publication. I also don't know why the day of the month for the publication date ("5") is not included.

Thanks!

adamsmith

I'm really sorry, but I don't see how we can do much with this on the Zotero side. You note yourself the wrong title/publication title and missing day. Also @misc is the most generic item type in bibtex, so we can't extract any information on the item type.

No other export formats from Sente?

You could try their support or maybe posting in one of the old threads on this here on the forum. Maybe one of the other people who have successfully migrated has any ideas. Sorry I can't be of more help.

skamen

I tried exporting from Sente as an EndNote XML and the fields carried over to Zotero much more accurately than they did as a BibTeX file. Has anyone else used an EndNote XML export from Sente to migrate to Zotero? If so, would you recommend going that route?

adamsmith

Yeah, Endnote XML is a good transfer format, fairly standardized and all, so generally speaking I'd recommend that where available. Any problems?

erikmarsh

I have also been trying to move my library from Sente to Zotero - there are four available export formats. The Endnote X format works better than BibTeX. Zotero does not recognize the SenteXML format. The Refer format works but it does not include anything but the basic reference info (no notes, etc).

Sente's notes have titles, page numbers, and text. In both the BibTeX and Endnote X formats, the the titles are lost and the page numbers and text are exported as one long line of text (which Zotero does import). Only the SenteXML format preserves titles and page numbers, but Zotero doesn't import this format.

I'd also like to import the collections, but Sente does not seem to export this information (in any format).

Some of my references had secondary titles instead of titles, which Zotero did not recognize. I fixed this by replacing "secondary-title" with "title" in a text editor.

Any suggestions on how to import the notes or collections?

karnesky

If it isn't in the export file, we can't import it. It sounds like you need to be talking to Sente's support.

A common work-around for collections is to export the collections one-by-one.

adamsmith

@karnesky -- Sente is deprecating/closing, that's why we're seeing all these Sente posts. I doubt their support will be of much help.

It would probably be possible to write a SenteXML importer that could bring over more information (though it appears not the collections), but this type of stuff is a ton of work -- if someone wants to work on this I'm happy to give pointers and invest some time in helping this,
Someone with some interest&time, a little bit of javascript skill and a basic understanding of XML/xpaths could do this.

Given that this is only going to be used to get people out of Sente now that it's deprecated and then will be useless, I'm not going to invest the ~6hs it'd probably take me to write and test something production ready.

For collections -- could you duplicate them as tags in Sente and export those and use them to re-create collections?

erikmarsh

Sente has become abandon-ware. There hasn't been support (or any explanations about what happened) for at least a year. More recently, their website and cloud servers went offline. So you will probably getting a lot more Sente posts! Maybe it's not worth writing a SenteXML importer (though 6 hours sounds like some fast coding to me!) I wish I could help but I don't have any experience with XML.

Thanks for the suggestions on exporting collections. I will see what works best. I have also noticed inconsistencies in the Sente export > Zotero import by reference type. I can recommend to others: 1) convert any custom reference types to the most similar standard ones 2) export them one type at a time

adamsmith

If you'd be willing to post some sample Sente XML somewhere (either on dropbox or even better just pasted to gist.github.com ) and linked to from here that would make it easier to get a sense of what we're looking at. Ideally would be the whole set of available item types.

erikmarsh

Sure, no problem! Here are examples for 7 of the most common reference types. There are a total of 42 types (I don't use most of them). Take a look; if it's useful I can add the other types. https://gist.github.com/anonymous/a35edf9860c7ac39226404451cee8f2a

adamsmith

Yeah, so this is very clean, wouldn't be hard to do at all, but the way I'm seeing this it seems like the only additional data this would bring in are the titles and pages for notes, which seems like a lot of work for very limited pay-off. Indeed no sign of collections in there, unfortunately.

erikmarsh

Well thanks for taking a look. Yes, you're right about it being a limited pay off, because I can already do a mostly-complete import with the Endnote X export format. The other problems I've having are on my end: custom reference types and poor library organization. I appreciate the quick responses! Thanks