Ever tired of this?

1. Find some scientific article on DBLP.

2. Click "electronic edition", be forced to navigate some publisher's site while all you want is the pdf.

3. Finally find the pdf, with a publisher generated crazy filename.

4. Be forced to go back to DBLP to download the matching bibtex file.

5. Change the pdf/bib filenames to something reasonable.

Paperget automates all of this.

You give paperget fragments of a title, authors, etc., and it figures out how to download the pdf and bib file and saves both with reasonable filenames matching the bibtex key.

$ ./paperget.py reingold connectivity log space
        authors: Omer Reingold
             ee: http://doi.acm.org/10.1145/1391289.1391291
          title: Undirected connectivity in log-space.
          venue: J. ACM
           year: 2008
Is this the paper you are looking for? [Yn]
Use the bibtex key 'reingold2008'? [Yn]
Following http://doi.acm.org/10.1145/1391289.1391291
Found acceptable destination: http://dl.acm.org/citation.cfm?doid=1391289.1391291
Using module ACM to download pdf...
Successfully downloaded ~/papers/reingold2008.pdf
No suitable bib module for: http://dl.acm.org/citation.cfm?doid=1391289.1391291
Falling back to DBLP bibtex information.
Successfully downloaded ~/papers/reingold2008.bib

The path ~/papers/ and the bibtex key schema are hardcoded but can easily be changed. TODO: Make this configurable.

Paperget does not break paywalls. Downloading pdfs from a paywalled site only works if you have access, for example via your academic institution. Having said this, if paperget hits a paywall, it will ask google for the title of a paper. Often a paper turns out to be available for free on the author's website.

Paperget does not support batched downloads.