Bulk import of references from papers into your Zotero Library

I always wanted to be able to import the entire reference section of a paper into my own Zotero library. I finally found the best tool to do just that:

https://anystyle.io/

Step (1): Copy the list of references from a PDF paper and past these into Notepad
Step (2): Remove spurious newline characters until you have all references on one line like this with references separated by empty lines. Example: (of cause this website brakes the lines again in display, but you get the idea.

Thompson, D.; Sagar, P. (1999). Are our estuaries wader-friendly? Water & Atmosphere 7(3): 10–12.

Schwarz, A.-M.; Parker, S.; Grose, M. (2005). Estuary monitoring by communities. NIWA Information Series No. 56. 8 p.

Step(3): Copy the whole reference section that you now have in txt format into the website: https://anystyle.io/

It will parse the references into what you expect. Download the result as BibTex file on the site.

Import into Zotero, Presto!!!

When will Zotero get its own text reference parser that would permit this directly???
  • surprised no-one cared to say thanks so far, as I'm sure many have benefited from this post - so I will be the first :) worked great for me.
  • (AnyStyle is our standard recommendation on Importing Formatted Bibliographies. It was created by a Zotero developer.)
  • edited August 12, 2021
    This is incredible. Thank you to nameless Zotero developer who put it together and to @tare2 for sharing this.
  • edited August 13, 2021
    Somewhat off-topic but concerning AnyStyle:

    Originally posted here but feels better as a new question:
    https://forums.zotero.org/discussion/91211/anystyle-source-document-of-poor-quality?new=1
  • Just wanted to say a HUGE thank you for sharing this. Saved me many hours of worry!
  • I cannot agree more. This is really incredible! Thank you so much. It really helps me so much.
  • Thanks a lot.
    It would be perfect for me if it would complete the references by searching on the Web and completing each paper with the information that the browsers' Zotero import plugins retrieve. The search feature would prefer the published version of the paper to the preprint.
  • I know this wasn`t available back then but you can easily do that with chat-gpt(4, 4o) now. it will write your references in bibtex format and give the file to import to zotero
  • As suggested by @kimarq , here's an example of the steps in doing that with ChatGPT ...
    https://x.com/MushtaqBilalPhD/status/1753785908266963066

    For doing that on a more regular basis from the reference lists of PDFs you *already* have in Zotero, you can use the zotero reference plugin:
    https://github.com/MuiseDestiny/zotero-reference
  • @tim820
    I would find it very helpful if you could link directly to the Bibliography ChatGPT utility without going through X (Twitter). For the past couple of days, (since they forwarded twitter.com to X.com) I've not been able to sign in by ether URL. I keep receiving a retry button.

    How does this utility differ from the very-useful AnyStyle? Thanks
  • Unfortunately a few things seem to have broken in that transfer from twitter.com to x.com.

    But it was just a user posting instructions on twitter/x for using ChatGPT in that way. It's not actually a utility for that specific purpose. Someone submitted the thread to threadreaderapp so it should also still be accessible here:
    https://threadreaderapp.com/thread/1753785908266963066.html?utm_campaign=topunroll

    Someone else in the thread reported that after copying the converted references from ChatGPT, you can just go to Zotero and choose "Import from clipboard". Avoiding the need for an import file. But copy from the clipboard apparently causes Zotero to put all items in your library, whereas the file-based approach puts them in a new collection if you want.

    Any advantage over anystyle might be in the number of steps required, any manual assistance with parsing needed, cleanup required, etc. Probably case-specific. Whether ChatGPT-based approaches are more powerful in the long run will await more user experience with them.
  • Thanks @tare2
    What i used before was a custom script getting and parsing refs from https://api.semanticscholar.org/v1/paper/DOI , then using doi2bib to generate a .bib file for import in Zotero
Sign In or Register to comment.