PDF downloading from PubMed

cmf5

In a previous discussion (http://forums.zotero.org/discussion/21990/pdf-not-automatically-downloading/), a Zotero developer noted that Zotero could not download pdfs directly from PubMed, only the databases that directly provide them (e.g. PubMed Central or journal publisher websites). While that is true, in practice this makes use of Zotero quite inconvenient compared to other bibliographic database managers. Prior to using Zotero (primarily due to its cross platform (Mac, Linux) compatibility), I used Bookends for the Mac. This program was able to search PubMed directly with search terms and return hits without using a browser. (This would be a nice feature for the Zotero standalone app.) You could also have it automatically download pdf's for any hits you imported into your database, whether they were in PubMed Central or a publisher's website (so long as you had access to the pdf), and the download nearly always worked flawlessly. This behavior was particularly impressive when your IP address on a university network was detected, which verified the university's status as paid subscriber if that was required for pdf download.

In contrast, when using Zotero with a PubMed search, you need to click on individual hits one-by-one, then follow the links on each hit's page to PubMed Central or the publisher's site to download the pdf. Even then, I notice that clicking the icon in the URL field of Firefox does not result in the pdf downloading. (Clicking the pdf link and selecting "Import into Zotero" on the resulting dialogue window does work though. The net result is that downloading citations with attached pdf's takes a lot more time.

Any way Zotero developers could smooth the flow of pdf importation from PubMed would be greatly appreciated.

adamsmith

Zotero is not going to support searching without a browser. (The reasons are mainly technical, but the answer is that it's not going to happen).

Note that pdf import does work from PubMed Central, just not from Pubmed. I dont' think that will change anytime soon, I see no reasonable way of making that work.

If a pdf doesn't import from the website of a commercial provider where you are able to access it manually, report that in a separate thread - that should work from all the major databases (Wiley, Springer, ScienceDirect, Sage), though it might be less reliable using standalone and connectors than using Zotero within Firefox.

I don't know if automatic pdf attachment for downloaded items will happen - it's tricky and I don't see this happening anytime soon, though I can see that it's at least theoretically possible - Bookends uses the DOI to locate the publisher and then gets the PDF from there - that doesn't strike me as impossible, but its nowhere close.

aurimas

I've recently started writing a paper and gathering sources mostly through pubmed. This same issue was slightly annoying to me, so I started thinking about ways to streamline pubmed translator. I think now that 3.0.4 came out and we can do cross domain processDocument calls, it might be possible to automate this. I'll have to try a couple things and get back to you on the details. Can't say how soon, since it's a little bit busy for me right now.

mugo

1. I tried downloading PDFs without success. First, I selected them (tick mark) and them clicked on the icon at top right hand corner of Chrome. When the dialogue box opened, I selected all. The PDFs did not download. I am missing something? I really need to download them.

NOTE: Under Zotero Preferences: I have already selected "Automatically attach associated PDFs and other files when saving"

2. On another observation, I noticed as if the PDF was downloaded but then it seemed to "somehow disappear". What is happening?

3. For some PDFs which I dragged into Zotero, I tried to "Retrieve metadata for PDF" but it not work; however, for some PDFs, it worked. What is the problem here? Please help.

Thank you in advance for your help.

adamsmith

1. You'll not get PDFs from pubmed.
Whether you get PDFs from other sites depends - you'd have to ask which ones, most supported sites should work.
2. If you see PDF briefly and then it vanishes, Zotero has determined that the file it downloaded as a PDF isn't one - that happens e.g. if you download from a site where you don't have full text access.

3. Retrieve Metadata sometimes fails - namely when Zotero can't find the metadata.

mfjackson32

Realize this is a somewhat dated thread, but wondering whether there has been any progress on implicating this feature request.

Automatic PDF downloads work great from Pubmed central, but the ability to do so from Pubmed would be much more helpful

adamsmith

no - this isn't really any closer than 18 months ago.