tools/msnchat

   1 #!/usr/bin/env python
   2 """
   3 Process packet capture files and produce a nice HTML
   4 report of MSN Chat sessions.
   5
   6 Copyright (c) 2003 by Gilbert Ramirez <gram@alumni.rice.edu>
   7
   8 This program is free software; you can redistribute it and/or
   9 modify it under the terms of the GNU General Public License
  10 as published by the Free Software Foundation; either version 2
  11 of the License, or (at your option) any later version.
  12
  13 This program is distributed in the hope that it will be useful,
  14 but WITHOUT ANY WARRANTY; without even the implied warranty of
  15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 GNU General Public License for more details.
  17
  18 You should have received a copy of the GNU General Public License
  19 along with this program; if not, write to the Free Software
  20 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  21 """
  22
  23 import os
  24 import re
  25 import sys
  26 import array
  27 import string
  28 import WiresharkXML
  29 import getopt
  30
  31 # By default we output the HTML to stdout
  32 out_fh = sys.stdout
  33
  34 class MSNMessage:
  35     pass
  36
  37 class MSN_MSG(MSNMessage):
  38     def __init__(self, timestamp, user, message):
  39         self.timestamp = timestamp
  40         self.user = user
  41         self.message = message
  42
  43
  44 class Conversation:
  45     """Keeps track of a single MSN chat session"""
  46
  47     re_MSG_out = re.compile("MSG (?P<TrID>\d+) (?P<ACKTYPE>[UNA]) (?P<len>\d+)")
  48     re_MSG_in  = re.compile("MSG (?P<user>\S+)@(?P<domain>\S+) (?P<alias>\S+) (?P<len>\d+)")
  49
  50     USER_NOT_FOUND = -1
  51     DEFAULT_USER = None
  52
  53
  54     DEFAULT_USER_COLOR = "#0000ff"
  55     USER_COLORS = [ "#ff0000", "#00ff00",
  56             "#800000", "#008000", "#000080" ]
  57
  58     DEFAULT_USER_TEXT_COLOR = "#000000"
  59     USER_TEXT_COLOR = "#000080"
  60
  61     def __init__(self):
  62         self.packets = []
  63         self.messages = []
  64
  65     def AddPacket(self, packet):
  66         self.packets.append(packet)
  67
  68     def Summarize(self):
  69         for packet in self.packets:
  70             msg = self.CreateMSNMessage(packet)
  71             if msg:
  72                 self.messages.append(msg)
  73             else:
  74                  #XXX
  75                  pass
  76
  77
  78     def CreateMSNMessage(self, packet):
  79         msnms = packet.get_items("msnms")[0]
  80
  81         # Check the first line in the msnms transmission for the user
  82         child = msnms.children[0]
  83         user = self.USER_NOT_FOUND
  84
  85         m = self.re_MSG_out.search(child.show)
  86         if m:
  87             user = self.DEFAULT_USER
  88
  89         else:
  90             m = self.re_MSG_in.search(child.show)
  91             if m:
  92                 user = m.group("alias")
  93
  94         if user == self.USER_NOT_FOUND:
  95             print >> sys.stderr, "No match for", child.show
  96             sys.exit(1)
  97             return None
  98
  99         msg = ""
 100
 101         i = 5
 102         check_trailing = 0
 103         if len(msnms.children) > 5:
 104             check_trailing = 1
 105
 106         while i < len(msnms.children):
 107             msg += msnms.children[i].show
 108             if check_trailing:
 109                 j = msg.find("MSG ")
 110                 if j >= 0:
 111                     msg = msg[:j]
 112                     i += 5
 113                 else:
 114                     i += 6
 115             else:
 116                 i += 6
 117
 118         timestamp = packet.get_items("frame.time")[0].get_show()
 119         i = timestamp.rfind(".")
 120         timestamp = timestamp[:i]
 121
 122         return MSN_MSG(timestamp, user, msg)
 123
 124     def MsgToHTML(self, text):
 125         bytes = array.array("B")
 126
 127         new_string = text
 128         i = new_string.find("\\")
 129
 130         while i > -1:
 131             # At the end?
 132             if i == len(new_string) - 1:
 133                 # Just let the default action
 134                 # copy everything to 'bytes'
 135                 break
 136
 137             if new_string[i+1] in string.digits:
 138                 left = new_string[:i]
 139                 bytes.fromstring(left)
 140
 141                 right = new_string[i+4:]
 142
 143                 oct_string = new_string[i+1:i+4]
 144                 char = int(oct_string, 8)
 145                 bytes.append(char)
 146
 147                 new_string = right
 148
 149             # ignore \r and \n
 150             elif new_string[i+1] in "rn":
 151                 copy_these = new_string[:i]
 152                 bytes.fromstring(copy_these)
 153                 new_string = new_string[i+2:]
 154
 155             else:
 156                 copy_these = new_string[:i+2]
 157                 bytes.fromstring(copy_these)
 158                 new_string = new_string[i+2:]
 159
 160             i = new_string.find("\\")
 161
 162
 163         bytes.fromstring(new_string)
 164
 165         return bytes
 166
 167     def CreateHTML(self, default_user):
 168         if not self.messages:
 169             return
 170
 171         print >> out_fh, """
 172 <HR><BR><H3 Align=Center> ---- New Conversation @ %s ----</H3><BR>""" \
 173             % (self.messages[0].timestamp)
 174
 175         user_color_assignments = {}
 176
 177         for msg in self.messages:
 178             # Calculate 'user' and 'user_color' and 'user_text_color'
 179             if msg.user == self.DEFAULT_USER:
 180                 user = default_user
 181                 user_color = self.DEFAULT_USER_COLOR
 182                 user_text_color = self.DEFAULT_USER_TEXT_COLOR
 183             else:
 184                 user = msg.user
 185                 user_text_color = self.USER_TEXT_COLOR
 186                 if user_color_assignments.has_key(user):
 187                     user_color = user_color_assignments[user]
 188                 else:
 189                     num_assigned = len(user_color_assignments.keys())
 190                     user_color = self.USER_COLORS[num_assigned]
 191                     user_color_assignments[user] = user_color
 192
 193             # "Oct  6, 2003 21:45:25"  --> "21:45:25"
 194             timestamp = msg.timestamp.split()[-1]
 195
 196             htmlmsg = self.MsgToHTML(msg.message)
 197
 198             print >> out_fh, """
 199 <FONT COLOR="%s"><FONT SIZE="2">(%s) </FONT><B>%s:</B></FONT> <FONT COLOR="%s">""" \
 200                 % (user_color, timestamp, user, user_text_color)
 201
 202             htmlmsg.tofile(out_fh)
 203
 204             print >> out_fh, "</FONT><BR>"
 205
 206
 207 class CaptureFile:
 208     """Parses a single a capture file and keeps track of
 209     all chat sessions in the file."""
 210
 211     def __init__(self, capture_filename, tshark):
 212         """Run tshark on the capture file and parse
 213         the data."""
 214         self.conversations = []
 215         self.conversations_map = {}
 216
 217         pipe = os.popen(tshark + " -Tpdml -n -R "
 218             "'msnms contains \"X-MMS-IM-Format\"' "
 219             "-r " + capture_filename, "r")
 220
 221         WiresharkXML.parse_fh(pipe, self.collect_packets)
 222
 223         for conv in self.conversations:
 224             conv.Summarize()
 225
 226     def collect_packets(self, packet):
 227         """Collect the packets passed back from WiresharkXML.
 228         Sort them by TCP/IP conversation, as there could be multiple
 229         clients per machine."""
 230         # Just in case we're looking at tunnelling protocols where
 231         # more than one IP or TCP header exists, look at the last one,
 232         # which would be the one inside the tunnel.
 233         src_ip = packet.get_items("ip.src")[-1].get_show()
 234         dst_ip = packet.get_items("ip.dst")[-1].get_show()
 235         src_tcp = packet.get_items("tcp.srcport")[-1].get_show()
 236         dst_tcp = packet.get_items("tcp.dstport")[-1].get_show()
 237
 238         key_params = [src_ip, dst_ip, src_tcp, dst_tcp]
 239         key_params.sort()
 240         key = '|'.join(key_params)
 241
 242         if not self.conversations_map.has_key(key):
 243             conv = self.conversations_map[key] = Conversation()
 244             self.conversations.append(conv)
 245         else:
 246             conv = self.conversations_map[key]
 247
 248         conv.AddPacket(packet)
 249
 250
 251     def CreateHTML(self, default_user):
 252         if not self.conversations:
 253             return
 254
 255         for conv in self.conversations:
 256             conv.CreateHTML(default_user)
 257
 258
 259 def run_filename(filename, default_user, tshark):
 260     """Process one capture file."""
 261
 262     capture = CaptureFile(filename, tshark)
 263     capture.CreateHTML(default_user)
 264
 265
 266 def run(filenames, default_user, tshark):
 267     # HTML Header
 268     print >> out_fh, """
 269 <HTML><TITLE>MSN Conversation</TITLE>
 270 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
 271 <BODY>
 272 """
 273     for filename in filenames:
 274         run_filename(filename, default_user, tshark)
 275
 276     # HTML Footer
 277     print >> out_fh, """
 278 <HR>
 279 </BODY>
 280 </HTML>
 281 """
 282
 283
 284 def usage():
 285     print >> sys.stderr, "msnchat [OPTIONS] CAPTURE_FILE [...]"
 286     print >> sys.stderr, "  -o FILE       name of output file"
 287     print >> sys.stderr, "  -t TSHARK  location of tshark binary"
 288     print >> sys.stderr, "  -u USER       name for unknown user"
 289     sys.exit(1)
 290
 291 def main():
 292     default_user = "Unknown"
 293     tshark = "tshark"
 294
 295     optstring = "ho:t:u:"
 296     longopts = ["help"]
 297
 298     try:
 299         opts, args = getopt.getopt(sys.argv[1:], optstring, longopts)
 300     except getopt.GetoptError:
 301         usage()
 302
 303     for opt, arg in opts:
 304         if opt == "-h" or opt == "--help":
 305             usage()
 306
 307         elif opt == "-o":
 308             filename = arg
 309             global out_fh
 310             try:
 311                 out_fh = open(filename, "w")
 312             except IOError:
 313                 sys.exit("Could not open %s for writing." % (filename,))
 314
 315         elif opt == "-u":
 316             default_user = arg
 317
 318         elif opt == "-t":
 319             tshark = arg
 320
 321         else:
 322             sys.exit("Unhandled command-line option: " + opt)
 323
 324     run(args, default_user, tshark)
 325
 326 if __name__ == '__main__':
 327     main()