# encoding: utf-8

'''

firefox_session_textareas.py - Recover text area content from Firefox session files

Print out text area content cached by Firefox in the saved session file.

Sample usage:

python firefox_session_textareas.py "~/.firefox/Profiles/*/sessionstore.js"

    -- Print out all text area content found (including the corresponding page URL)

python firefox_session_textareas.py "~/.firefox/Profiles/*/sessionstore.js" spam

    -- Print out all only text areas found to contain the string "spam"

       (or whose corresponding page URL contains that string)

'''

import urllib

import sys

import re

TEXTAREA_PAT_3_0 = re.compile(r'{url:"([^"]*)"[^}]*}[^}]*text:"#editor-textarea=([^"]*)"')

TEXTAREA_PAT_3_5 = re.compile(r'{"url":"([^"]*)"[^}]*?,"title":"([^"]*)".*?,"formdata":({("([^"]*)":("((\\"|[^"])*)"|\d+|true|false),?)*})')

FORMDATA_PAT_3_5 = re.compile(r'("([^"]*)":("((\\"|[^"])*)"))')

HEADING_TPL = '------------ firefox_session_textareas.py - %s ------------'

def textarea_content(s):

    return urllib.unquote(s)

#FIXME: Use optparse, etc to clean up command line handling

sessioninfo = open(sys.argv[1]).read()

try:

    search = sys.argv[2]

except IndexError:

    search = None

#First check the Firefox 3 sessionstore.js format

for match in TEXTAREA_PAT_3_0.finditer(sessioninfo):

    if search is None or search in match.group(0):

        print HEADING_TPL%match.group(1)

        print

        print textarea_content(match.group(2))

        print

#Now check the Firefox 3.5 sessionstore.js format

for match in TEXTAREA_PAT_3_5.finditer(sessioninfo):

    #{"url":"http://trac.xml3k.org/newticket","title":"New Ticket – xml3k – Trac","ID":2969506219,"scroll":"0,121","formdata":{"#proj-search":"","#field-summary":"Fix-up of Amara bindery nodes after xml_namespace update","#field-reporter":"http://uche.myopenid.com/","#field-description":"Amara bindery nodes are bound to their parents by connecting (via a descriptor) a Python attribute name to a namespace/local name combination.\u000a\u000aGiven the following document:\u000a\u000a{{{\u000a<top xmlns:a=\"urn:bogus:a\">\u000a  <a:monty/>\u000a</top>\u000a}}}

    if search is None or search in match.group(0):

        #print match.group(0)

        print HEADING_TPL%match.group(1)

        print

        print match.group(2)

        for datamatch in FORMDATA_PAT_3_5.finditer(match.group(3)):

            #Were we to convert JS true to Python True, and prepend u to strings, we couild probabaly use eval

            print "Field name:", datamatch.group(2)

            try:

                print "Field value\n", eval('u'+datamatch.group(3)).encode('utf-8')

            except Exception, e:

                print "Unable to parse field value"

                print e

                print

                print

                print "Raw field value\n", datamatch.group(3)

            print