Extraction from PDF

Libraries allowing to access the content of the PDF document:

• Poppler http://poppler.freedesktop.org/ - Part of the freedesktop.org initiative
  • weakly documented
  • supports Python
  • poppler tools, like pdfimages seem to fail on most of the scientific articles
• gnupdf http://gnupdf.org/
  • Is supposed to fill all the missing fetures in poppler (like forms/multimedia support, which might be relevant)
  • seems not to have any Python bindings ( But might be possible to using it after generating a python library automatically)

A web page describing many useful PDF tools and libraries: http://pdf-house.blogspot.com/ Problems

• sometimes ( like 1007.0043), images are included in the form of PDF files

-- PiotrPraczyk - 07-Sep-2010