From browser to bookends

A place for users to ask each other questions, make suggestions, and discuss Bookends.
Post Reply
ghostwheel
Posts: 16
Joined: Sat Aug 15, 2015 5:30 am

From browser to bookends

Post by ghostwheel »

I'm trying to figure out a good workflow to get papers from my browser to bookends. I'm using Safari on OSX.
Here's the paper I'm trying:
https://www.medrxiv.org/content/10.1101 ... 22277306v1
Now, I have the "open in bookends" link as a bookmark. If I click on it when I see the web page, I can then click on the downarrow, then select "import pdf". A new entry is generated, with the right title and authors, but no pdf. If I first download the pdf, and then click "open in bookends", I'll get a reference with the attachment, but no other info.
Now I have a reference with the attachment. The paper has a DOI. I put that in the DOI field, and now I can right click on the paper DOI->lookup articles from DOI->Bookends browser. The same site as above is opened, and I'm at the starting point, I can either generate a reference with the info and no attachment or one with info but no pdf.
What am I doing wrong? There must be an easy way to import a pdf into bookends.... I think it used to be easier years ago.
I also think it is easier on the iPad. Oh, I just checked. On the iPad, starting from pdf autofill by DOI also doesn't find any info.
Jon
Site Admin
Posts: 10048
Joined: Tue Jul 13, 2004 6:27 pm
Location: Bethesda, MD
Contact:

Re: From browser to bookends

Post by Jon »

When you have Bookends to "attempt" to get the PDF automatically sometimes it works, sometimes it doesn't. If you don't have access privileges, it won't. And even if you do, which is the case here with MedRxiv,success depends on how the journal and published links to the article. Bookends has to have specific handlers for specific sources, and of course it can't cover all. And even the ones it covers can break Bookends by changing their HTML.

But if you can get the PDF yourself, and it has a DOI embedded in it, you can drop the PDF onto Bookends and check "Make new reference". That will attach the PDF and import the metadata. You probably didn't check "Make new reference" when you "imported" the PDF.

For PDFs with DOIs, you can attach the PDF to a new empty reference and then do an Autocomplete Paper, choosing a site that's likely to have the metadata (PubMed, Google Scholar, JSTOR, etc.). This is what you should have done when you had the PDF but no metadata.

Having said all that, I think the PDF should have been downloaded. I'll take a look and see what happened.

Jon
Sonny Software
ghostwheel
Posts: 16
Joined: Sat Aug 15, 2015 5:30 am

Re: From browser to bookends

Post by ghostwheel »

Thanks for the quick answer!
Can you try the paper I gave above? It is open access, since it is medrxiv, so there's no access problem.
When I download the pdf and drop into bookends (I drag the pdf onto the bookends icon in the dock) I get the
"Attach to Imported Reference", and there I can select "Attach to a new reference and autocomplete". Is that what you meant with the "Make new reference" checkmark? This is a bit more successful. I get the title correctly, but no authors and no DOI.
The problem is that this is a new madrxiv entry, so none of my search engines can autocomplete. Google scholar still doesn't show the paper. It is just in medrxiv.
Notice that if I open the web page (https://www.medrxiv.org/content/10.1101 ... 22277306v1) in bookends, and then click down arrow, I get title and authors (still not the DOI). On iPad, DOI is recognized from the PDF, though.

Isn't it cool that clicking "citation tools" in medRxiv gives the option to download for Bookends? Then I can open the downloaded file, and get all the info.
Jon
Site Admin
Posts: 10048
Joined: Tue Jul 13, 2004 6:27 pm
Location: Bethesda, MD
Contact:

Re: From browser to bookends

Post by Jon »

I looked and I think the problem is with Crossref. Bookends gets the DOI from the paper and tries to get the metadata from Crossref via their API (not the usual Crossref search) and the response is -- no matching reference.

Therefore, Bookends falls back on COinS data embedded in the HTML, which is something but, as you saw, not complete.

We'll have to wait and see if this is something that is temporarily broken on Crossref, or if their API doesn't support medrxiv DOIs (although their web interface does). This used to work, I believe, so I'm thinking the former.

Jon
Sonny Software
ghostwheel
Posts: 16
Joined: Sat Aug 15, 2015 5:30 am

Re: From browser to bookends

Post by ghostwheel »

Thanks!

OK, so I found a workflow for medRxiv. For now.

Download citation in bookends format.
open the RIS file in bookends.
Select the reference, right click, and select "GoTo URL".
Open pdf
"Open in Bookends",
add pdf to current reference.
Jon
Site Admin
Posts: 10048
Joined: Tue Jul 13, 2004 6:27 pm
Location: Bethesda, MD
Contact:

Re: From browser to bookends

Post by Jon »

It seems I was right about a problem at Crossref. I just tried again with your medRxiv link and the full metadata and PDF were imported from Bookends Browser. Please try it yourself.

Jon
Sonny Software
ghostwheel
Posts: 16
Joined: Sat Aug 15, 2015 5:30 am

Re: From browser to bookends

Post by ghostwheel »

Yes, it works now. Thanks. It must be an issue for articles which are just posted on axiv
iandol
Posts: 465
Joined: Fri Jan 25, 2008 2:31 pm

Re: From browser to bookends

Post by iandol »

Crossref does have some delay for new articles. This drives me crazy, and I end up with just adding these shiny baby DOIs to a temporary note and every so often importing them after a week or two.
Post Reply