2013/6/12 David Cuenca dacuetu@gmail.com:
It is not a trivial matter. The best bet would be to take an existing pdf import tool for a word processor, and try to write a similar tool for wikitext.
There is the Oracle PDF Import Extension for Open Office, the code can be browsed, maybe it can give you some ideas http://extensions.services.openoffice.org/project/pdfimport
PDF scraping is a technique that's is gaining more and more attention since a lot of data on the web are hidden in PDF; so some libraries for this task are under development. My favorite language being Python I will suggest this blog post: http://blog.scraperwiki.com/2010/12/17/scraping-pdfs-now-26-less-unpleasant-...
C