[4.0b2] Retrieve Metadata for PDF

adamsmith

First, let me say that I'm very happy about the improvements. The inclusion of ISBNs in the search works great for me (it picks up a lot of reports from international organization, e.g.) and the google scholar search also seems to have improved significantly.

One request and one error report:
1. (Request) I would request that, as for ISBNs by identifier, we query the LoC before querying Worldcat, as data is much better - it doesn't seem like that's currently the case.

2. (Error report):
this PDF:
http://www.oit.org.ar/WDMS/bib/publ/libros/estrategias_asociativas.pdf
finds a completely non-sensical result for me. The reason is that it searches for
http://scholar.google.com/scholar?q=%22_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%20_%22%20&hl=en&lr=&btnG=Search

i.e. a bunch of underscores with no text. We should make sure that retrievePDF uses a string with actual letters.