Metadata for multiple pdf files

debchatterjee · April 14, 2013

I have copied close to 300 pdfs to my zotero library recently and was trying to get their metadata. I was locked out of Google Scholar after the first 30-40 files (predictably, as I found later). Searching around a bit, I realized that this a common problem, reflected even on this forum. The solution suggested is to do this a little at a time, say one or two files, give a minute or two in between, and then again a bit, and so on. Now this is fine if we are already into zotero, and once in a while getting papers into the library. But at least for the first time, or when shifting contents of a folder into zotero (as in my case), this doesn't work. So was wondering if any work around is possible to speed up the process. Thanks.

adamsmith · April 14, 2013

which version of Zotero? 4.0.4 has some additional provisions to prevent that, but it's not going to be perfect either way.
The best bet would be to experiment how fast you can go.
Also, as I do always with the retrieve feature a word of warning. While google scholar data has gotten better, it till has mistakes and missing fields - most importantly the DOI field, which is required in an increasing number of citation styles (notably, but not only, APA).

debchatterjee · April 15, 2013

I have 4.0.4 version. So far I think about 5 at a time certainly works. and yes, we should always check if all the fields are complete. But I was wondering if these retrieval queries could be directed to the zotero database itself. I would expect that a lot of the metadata would already have been captured,with so many members using the service for the last few years. The ones that are not available could then be obtained from Google Scholar. Just a suggestion, though I certainly wouldn't be sure of the technical issues and permission issues involved. Thanks.

adamsmith · April 15, 2013

But I was wondering if these retrieval queries could be directed to the zotero database itself.

both legal and technical issues make that unlikely any time soon, though I see the intuitive appeal of the idea.

dstillman · April 15, 2013

We're actually considering adding such functionality, though, yes, there are some issues to consider. But between file hashes (now) and content queries (with full-text content syncing in 4.1), there are multiple ways we could do this.