AutoCrop (OpenWith solution for Briss)

[EnergyLens]

Here is a little script that can be called from the OpenWith plugin in Calibre to automate the use of Briss <http://sourceforge.net/projects/briss/> to Crop PDFs for easier reading on your eReader.

You need to make the file executable and edit this line in the file:

# Must specify the location of briss-0.9.jar on your system
path_to_briss = '/Applications/ebook tools/briss-0.9/briss-0.9.jar'

For those on Mac: I just updated to Yosemite and had to re-install Java to keep Briss working: <http://support.apple.com/kb/DL1572>

The program will make a copy of your original PDF and add an ORIGINAL_PDF filetype into the Calibre database (similar to ORIGINAL_EPUB). You can then try to teach your OS to open the ORIGINAL_PDF filetype with the PDF Reader of your choice (though Preview refuses... so I had to use Adobe on Mac).

[EnergyLens]

I also wanted to note that the inspiration for this was not primarily to make PDFs more readable on eReaders (which it does), but to streamline cleanup after OCR of PDF files. Briss does a great job of automatically cutting off page numbers and chapter names which occur on every page, thus making the resulting text much more readable.

I use PDFOCRx on Mac, which does a great job with two-column PDFs, and produces soft line-wraps. Sometimes OCR works even better (in terms of cleanup) than extracting the text from the native PDF.

I also have a custom column in my Calibre library which sorts PDFs into three type: Scanned, Scan with embedded OCR data, and Native. I populate this column when loading new PDFs into the library and then later the value of the column helps me to decide how to process the file when converting to other formats.