Can I prevent automatic child item creation for DOIs?

ndrwclrk

I sincerely apologize if this question has been asked and answered elsewhere. I couldn't find it if so. My zotero database is mostly populated by importing .ris files from places like sciencedirect, etc. There's a perfectly good spot within the database for the doi, and the import routine generally populates this field correctly. However, for some unknown reason, it also creates a child item with the doi. I wish it wouldn't do this. I like to reserve my child items for attaching links to PDFs of the complete article.

It's not a huge deal - I can continue as always just deleting the child item after every import. I'm curious about two things, though:
1) What was the logic behind creating the child item doi in the first place? Why not just store it in the main record?
2) Is there a setting that I'm missing, that would allow me to "turn off" the automatic child item creation for doi?

Any thoughts, pointers, suggestions are much appreciated.

adamsmith

Do you mean a note with the DOI? That's likely just due to a DOI listed in the note tag in the RIS -- RIS is often messy from publishers. Any reason you're not using the Save to Zotero icon? We try to clean up things like that.

ndrwclrk

Yes, I guess that's what I mean. I'm not exactly sure of the terminology. I use standalone zotero, and historically some of the features like "save to zotero" seemed to work only with the browser extension. Whether or not that's officially true, I don't know. It is just my impression. Anyway, I developed the habit of importing things via ris files because I found it to be more reliable. I've been using zotero since version 1.0, so sometimes things evolve, and I don't adopt them because I'm stuck in my ways.

I will examine some of the ris files I've imported recently and see if they have doi info in note tags. Those would be the N1, N2, etc tags?

adamsmith

It's N1 and RN. N2 would go into the abstract field if there's no AB and otherwise be discarded.

I'd very much recommend using the browser extensions and Zotero import is optimized for those and almost always supplements or correct metadata supplied by publishers or databases.

You're obviously free to use Zotero however you want, though, but you'll understand that we'll not spend a ton of time troubleshooting non-standard workflows.

bwiernik

Yes, those would be the tags. You would want to change those tags to DO. Generally, the Save to Zotero Button will almost always yield as good or better data than downloading and importing from RIS files, so I would recommend trying to switch to using that.

adamsmith

(in this case I'm right about the RIS tags for notes, but only because I just checked)

ndrwclrk

I did a little digging, and found out that it's mostly as you say - the DOIs are in either N1 or M3 tags within the ris files. So I guess my choices are to either: 1) try using the bookmarklet to add things directly on the back end; or 2) figure out a way to strip out those tags automatically myself.

Another way to look at this would be as a feature request. If there's already code to clean things up that works on the server side, maybe implement that same code in the routines that import references from files?

My current routine is as follows: 1) find an interesting article; 2) export the .ris file; 3) import the .ris file; 4) download the pdf; and 5) attach the pdf. I like to store my pdfs locally. I guess when I think about it, if I used the bookmarklet, I'm just replacing step (3) with a "sync zotero to server" step. I'll give it a try for a while and see what happens.

Thanks for your help.

adamsmith

(M3 should actually import as DOI correctly)

I think it may be worthwhile taking another look at how Zotero works-- I understand you've been using it for a long time, but you seem to have a number of misconceptions about it, most notably about the connection between browser extensions and cloud components (the two have almost no connection).

For your workflow, for example, you wouldn't use the bookmarklet. You'd use the browser extension, assuming you're able to use Firefox, Chrome, or Safari.
Then the workflow would be
1) find an interesting article
2) click the save to Zotero button which will import the article into Zotero and attach the PDF

this does not depend on using Zotero sync -- it all happens locally. You don't even need a Zotero account for this.

The issue with the clean-up code isn't that it's only run server side (it isn't) but that clean-up depends on the source of the metadata.

E.g. we wouldn't want to strip DOIs from all notes -- some people may store DOIs in notes. We strip them when importing from publishers where we know that they put DOIs in note tags in RIS incorrectly.

You could modify the RIS importer to not include notes or check if they are DOIs befor importing, but it does seem like a waste of time to me. But FWIW: https://www.zotero.org/support/dev/translators/coding

ndrwclrk

You're right-on about the misconceptions and, yes, I should re-evaluate my current workflow. Another way to look at this is, "wow, Zotero has come a long way in 11+ years!" Thanks for all the excellent work you guys put into this amazing collection of free software!

bwiernik

Here are probably the best two pages for brushing up on where Zotero is currently at:
https://www.zotero.org/support/getting_stuff_into_your_library
https://www.zotero.org/support/quick_start_guide

ndrwclrk

Thank you both for your prompt and helpful responses. I'm going to check out the links you suggested, re-install the chrome extension, and give the new workflow a spin.

ndrwclrk

FYI, two years and a bit later, I have totally switched my workflow around, and the browser extensions are working great for me in Chrome and Firefox. Thanks again!