Work in progress tool for extracting highlights from Kindle ebooks.

kindle-highlight-extractor.py

import binascii
import re

# highlights are stored in *.mbp files corresponding to the filename of the book
#    in ~/Library/Application Support/Kindle/My Kindle Content/

# EA 44 41 54 41 - beginning of highlight
# 44 41 54 41 - end of highlight (except last highlight which *sometimes* ends with 42 4B 4D 4B)

fh = '0131177052.WrkngEffLegCode.mbp'

with open(fh, 'rb') as f:
	content = f.read()

data = binascii.hexlify(content)

# Regex doesn't work reliably yet
matches = re.findall('ea4441544100000[1.](.*?)44415441', data, re.S)
for match in matches:
	line = binascii.unhexlify(match)
	print line[1:]
	
print
print str(len(matches)) + ' highlights found'

After an hour or so of messing around with this and learning a lot about how the Kindle app writes annotations I came across a guy's documentation from his own reversing project.

MBP file format info: http://www.angelfire.com/ego2/idleloop/archives/mbp_file_format.txt
Perl tool for extracting annotations: http://www.angelfire.com/ego2/idleloop/mbp_reader.html