Utility.py 14 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391
  1. import calendar as cal
  2. import datetime as dt
  3. import ipaddress
  4. import os
  5. import random as rnd
  6. import lea
  7. import xdg.BaseDirectory as BaseDir
  8. import scapy.layers.inet as inet
  9. import scipy.stats as stats
  10. import pytz as pytz
  11. CACHE_DIR = os.path.join(BaseDir.xdg_cache_home, 'id2t')
  12. CODE_DIR = os.path.dirname(os.path.abspath(__file__)) + "/../"
  13. ROOT_DIR = CODE_DIR + "../"
  14. RESOURCE_DIR = ROOT_DIR + "resources/"
  15. TEST_DIR = RESOURCE_DIR + "test/"
  16. OUT_DIR = None
  17. # List of common operation systems
  18. platforms = {"win7", "win10", "winxp", "win8.1", "macos", "linux", "win8", "winvista", "winnt", "win2000"}
  19. # Distribution of common operation systems
  20. platform_probability = {"win7": 48.43, "win10": 27.99, "winxp": 6.07, "win8.1": 6.07, "macos": 5.94, "linux": 3.38,
  21. "win8": 1.35, "winvista": 0.46, "winnt": 0.31}
  22. # List of no-ops
  23. x86_nops = {b'\x90', b'\xfc', b'\xfd', b'\xf8', b'\xf9', b'\xf5', b'\x9b'}
  24. # List of pseudo no-ops (includes ops which won't change the state e.g. read access)
  25. x86_pseudo_nops = {b'\x97', b'\x96', b'\x95', b'\x93', b'\x92', b'\x91', b'\x99', b'\x4d', b'\x48', b'\x47', b'\x4f',
  26. b'\x40', b'\x41', b'\x37', b'\x3f', b'\x27', b'\x2f', b'\x46', b'\x4e', b'\x98', b'\x9f', b'\x4a',
  27. b'\x44', b'\x42', b'\x43', b'\x49', b'\x4b', b'\x45', b'\x4c', b'\x60', b'\x0e', b'\x1e', b'\x50',
  28. b'\x55', b'\x53', b'\x51', b'\x57', b'\x52', b'\x06', b'\x56', b'\x54', b'\x16', b'\x58', b'\x5d',
  29. b'\x5b', b'\x59', b'\x5f', b'\x5a', b'\x5e', b'\xd6'}
  30. # Characters which result in operational behaviour (e.g. FTPWinaXeExploit.py)
  31. forbidden_chars = [b'\x00', b'\x0a', b'\x0d']
  32. # Used in get_attacker_config
  33. attacker_port_mapping = {}
  34. # Used in get_attacker_config
  35. attacker_ttl_mapping = {}
  36. # Identifier for attacks
  37. generic_attack_names = {"attack", "exploit"}
  38. def update_timestamp(timestamp: float, pps: float, delay: float=0, inj_pps: float=0, inj_timestamp: float=0):
  39. """
  40. Calculates the next timestamp to be used based on the packet per second rate (pps) and the maximum delay.
  41. :return: Timestamp to be used for the next packet.
  42. """
  43. # FIXME: throw Exception if pps==0
  44. second = 0
  45. packets_this_second = 0
  46. if inj_pps != 0 and inj_timestamp != 0:
  47. time = timestamp - inj_timestamp
  48. packets_so_far = time / inj_pps
  49. packets_this_second = packets_so_far % inj_pps
  50. else:
  51. inj_pps = 0
  52. if delay == 0:
  53. # Calculate request timestamp
  54. # To imitate the bursty behavior of traffic
  55. random_delay = lea.Lea.fromValFreqsDict({1 / pps: 70, 2 / pps: 20, 5 / pps: 7, 10 / pps: 3})
  56. result_delay = rnd.uniform(1 / pps, random_delay.random())
  57. else:
  58. # Calculate reply timestamp
  59. random_delay = lea.Lea.fromValFreqsDict({delay / 2: 70, delay / 3: 20, delay / 5: 7, delay / 10: 3})
  60. result_delay = rnd.uniform(1 / pps + delay, 1 / pps + random_delay.random())
  61. result = timestamp + result_delay
  62. if inj_pps > packets_this_second and int(result) - int(timestamp) != 1:
  63. result = result + 1
  64. return result
  65. def get_timestamp_from_datetime_str(time: str):
  66. return pytz.timezone('UTC').localize(dt.datetime.strptime(time, "%Y-%m-%d %H:%M:%S.%f")).timestamp()
  67. def get_interval_pps(complement_interval_pps, timestamp):
  68. """
  69. Gets the packet rate (pps) for a specific time interval.
  70. :param complement_interval_pps: an array of tuples (the last timestamp in the interval, the packet rate in the
  71. corresponding interval).
  72. :param timestamp: the timestamp at which the packet rate is required.
  73. :return: the corresponding packet rate (pps) .
  74. """
  75. for row in complement_interval_pps:
  76. if timestamp <= row[0]:
  77. return row[1]
  78. return complement_interval_pps[-1][1] # in case the timestamp > capture max timestamp
  79. def get_nth_random_element(*element_list):
  80. """
  81. Returns the n-th element of every list from an arbitrary number of given lists.
  82. For example, list1 contains IP addresses, list 2 contains MAC addresses. Use of this function ensures that
  83. the n-th IP address uses always the n-th MAC address.
  84. :param element_list: An arbitrary number of lists.
  85. :return: A tuple of the n-th element of every list.
  86. """
  87. if len(element_list) <= 0:
  88. return None
  89. elif len(element_list) == 1 and len(element_list[0]) > 0:
  90. return rnd.choice(element_list[0])
  91. else:
  92. range_max = min([len(x) for x in element_list])
  93. if range_max > 0:
  94. range_max -= 1
  95. n = rnd.randint(0, range_max)
  96. return tuple(x[n] for x in element_list)
  97. else:
  98. return None
  99. def get_rnd_os():
  100. """
  101. Chooses random platform over an operating system probability distribution
  102. :return: random platform as string
  103. """
  104. os_dist = lea.Lea.fromValFreqsDict(platform_probability)
  105. return os_dist.random()
  106. def check_platform(platform: str) -> None:
  107. """
  108. Checks if the given platform is currently supported
  109. if not exits with error
  110. :param platform: the platform, which should be validated
  111. """
  112. if platform not in platforms:
  113. raise ValueError("ERROR: Invalid platform: " + platform + "." +
  114. "\n Please select one of the following platforms: " + ",".join(platforms))
  115. def get_ip_range(start_ip: str, end_ip: str):
  116. """
  117. Generates a list of IPs of a given range. If the start_ip is greater than the end_ip, the reverse range is generated
  118. :param start_ip: the start_ip of the desired IP-range
  119. :param end_ip: the end_ip of the desired IP-range
  120. :return: a list of all IPs in the desired IP-range, including start-/end_ip
  121. """
  122. start = ipaddress.ip_address(start_ip)
  123. end = ipaddress.ip_address(end_ip)
  124. ips = []
  125. if start < end:
  126. while start <= end:
  127. ips.append(start.exploded)
  128. start = start + 1
  129. elif start > end:
  130. while start >= end:
  131. ips.append(start.exploded)
  132. start = start - 1
  133. else:
  134. ips.append(start_ip)
  135. return ips
  136. def generate_source_port_from_platform(platform: str, previous_port=0):
  137. """
  138. Generates the next source port according to the TCP-port-selection strategy of the given platform
  139. :param platform: the platform for which to generate source ports
  140. :param previous_port: the previously used/generated source port. Must be 0 if no port was generated before
  141. :return: the next source port for the given platform
  142. """
  143. check_platform(platform)
  144. if platform in {"winnt", "winxp", "win2000"}:
  145. if (previous_port == 0) or (previous_port + 1 > 5000):
  146. return rnd.randint(1024, 5000)
  147. else:
  148. return previous_port + 1
  149. elif platform == "linux":
  150. return rnd.randint(32768, 61000)
  151. else:
  152. if (previous_port == 0) or (previous_port + 1 > 65535):
  153. return rnd.randint(49152, 65535)
  154. else:
  155. return previous_port + 1
  156. def get_filetime_format(timestamp):
  157. """
  158. Converts a timestamp into MS FILETIME format
  159. :param timestamp: a timestamp in seconds
  160. :return: MS FILETIME timestamp
  161. """
  162. boot_datetime = pytz.timezone('UTC').localize(dt.datetime.fromtimestamp(timestamp))
  163. boot_filetime = 116444736000000000 + (cal.timegm(boot_datetime.timetuple()) * 10000000)
  164. return boot_filetime + (boot_datetime.microsecond * 10)
  165. def get_rnd_boot_time(timestamp, platform="winxp"):
  166. """
  167. Generates a random boot time based on a given timestamp and operating system
  168. :param timestamp: a timestamp in seconds
  169. :param platform: a platform as string as specified in check_platform above. default is winxp. this param is optional
  170. :return: timestamp of random boot time in seconds since EPOCH
  171. """
  172. check_platform(platform)
  173. if platform is "linux":
  174. uptime_in_days = lea.Lea.fromValFreqsDict({3: 50, 7: 25, 14: 12.5, 31: 6.25, 92: 3.125, 183: 1.5625,
  175. 365: 0.78125, 1461: 0.390625, 2922: 0.390625})
  176. elif platform is "macos":
  177. uptime_in_days = lea.Lea.fromValFreqsDict({7: 50, 14: 25, 31: 12.5, 92: 6.25, 183: 3.125, 365: 3.076171875,
  178. 1461: 0.048828125})
  179. else:
  180. uptime_in_days = lea.Lea.fromValFreqsDict({3: 50, 7: 25, 14: 12.5, 31: 6.25, 92: 3.125, 183: 1.5625,
  181. 365: 0.78125, 1461: 0.78125})
  182. timestamp -= rnd.randint(0, uptime_in_days.random() * 86400)
  183. return timestamp
  184. def get_rnd_x86_nop(count=1, side_effect_free=False, char_filter=set()):
  185. """
  186. Generates a specified number of x86 single-byte (pseudo-)NOPs
  187. :param count: The number of bytes to generate
  188. :param side_effect_free: Determines whether NOPs with side-effects (to registers or the stack) are allowed
  189. :param char_filter: A set of bytes which are forbidden to generate
  190. :return: Random x86 NOP bytestring
  191. """
  192. result = b''
  193. nops = x86_nops.copy()
  194. if not side_effect_free:
  195. nops |= x86_pseudo_nops.copy()
  196. if not isinstance(char_filter, set):
  197. char_filter = set(char_filter)
  198. nops = list(nops - char_filter)
  199. for i in range(0, count):
  200. result += nops[rnd.randint(0, len(nops) - 1)]
  201. return result
  202. def get_rnd_bytes(count=1, ignore=None):
  203. """
  204. Generates a specified number of random bytes while excluding unwanted bytes
  205. :param count: Number of wanted bytes
  206. :param ignore: The bytes, which should be ignored, as an array
  207. :return: Random bytestring
  208. """
  209. if ignore is None:
  210. ignore = []
  211. result = b''
  212. for i in range(0, count):
  213. char = os.urandom(1)
  214. while char in ignore:
  215. char = os.urandom(1)
  216. result += char
  217. return result
  218. def check_payload_len(payload_len: int, limit: int) -> None:
  219. """
  220. Checks if the len of the payload exceeds a given limit
  221. :param payload_len: The length of the payload
  222. :param limit: The limit of the length of the payload which is allowed
  223. """
  224. if payload_len > limit:
  225. raise ValueError("Custom payload too long: " + str(payload_len) +
  226. " bytes. Should be a maximum of " + str(limit) + " bytes.")
  227. def get_bytes_from_file(filepath):
  228. """
  229. Converts the content of a file into its byte representation
  230. The content of the file can either be a string or hexadecimal numbers/bytes (e.g. shellcode)
  231. The file must have the keyword "str" or "hex" in its first line to specify the rest of the content
  232. If the content is hex, whitespaces, backslashes, "x", quotation marks and "+" are removed
  233. Example for a hexadecimal input file:
  234. hex
  235. "abcd ef \xff10\ff 'xaa' x \ ab"
  236. Output: b'\xab\xcd\xef\xff\x10\xff\xaa\xab'
  237. :param filepath: The path of the file from which to get the bytes
  238. :return: The bytes of the file (either a byte representation of a string or the bytes contained in the file)
  239. """
  240. try:
  241. file = open(filepath)
  242. result_bytes = b''
  243. header = file.readline().strip()
  244. content = file.read()
  245. if header == "hex":
  246. content = content.replace(" ", "").replace("\n", "").replace("\\", "").replace("x", "").replace("\"", "") \
  247. .replace("'", "").replace("+", "").replace("\r", "")
  248. try:
  249. result_bytes = bytes.fromhex(content)
  250. except ValueError:
  251. print("\nERROR: Content of file is not all hexadecimal.")
  252. file.close()
  253. exit(1)
  254. elif header == "str":
  255. result_bytes = content.strip().encode()
  256. else:
  257. print("\nERROR: Invalid header found: " + header + ". Try 'hex' or 'str' followed by endline instead.")
  258. file.close()
  259. exit(1)
  260. for forbidden_char in forbidden_chars:
  261. if forbidden_char in result_bytes:
  262. print("\nERROR: Forbidden character found in payload: ", forbidden_char)
  263. file.close()
  264. exit(1)
  265. file.close()
  266. return result_bytes
  267. except FileNotFoundError:
  268. print("\nERROR: File not found: ", filepath)
  269. exit(1)
  270. def handle_most_used_outputs(most_used_x):
  271. """
  272. :param most_used_x: Element or list (e.g. from SQL-query output) which should only be one element
  273. :return: most_used_x if it's not a list. The first element of most_used_x after being sorted if it's a list.
  274. None if that list is empty.
  275. """
  276. if isinstance(most_used_x, list):
  277. if len(most_used_x) == 0:
  278. return None
  279. most_used_x.sort()
  280. return most_used_x[0]
  281. else:
  282. return most_used_x
  283. def get_attacker_config(ip_source_list, ip_address: str):
  284. """
  285. Returns the attacker configuration depending on the IP address, this includes the port for the next
  286. attacking packet and the previously used (fixed) TTL value.
  287. :param ip_source_list: List of source IPs
  288. :param ip_address: The IP address of the attacker
  289. :return: A tuple consisting of (port, ttlValue)
  290. """
  291. # Gamma distribution parameters derived from MAWI 13.8G dataset
  292. alpha, loc, beta = (2.3261710235, -0.188306914406, 44.4853123884)
  293. gd = stats.gamma.rvs(alpha, loc=loc, scale=beta, size=len(ip_source_list))
  294. # Determine port
  295. port = attacker_port_mapping.get(ip_address)
  296. if port is not None: # use next port
  297. next_port = attacker_port_mapping.get(ip_address) + 1
  298. if next_port > (2 ** 16 - 1):
  299. next_port = 1
  300. else: # generate starting port
  301. next_port = inet.RandShort()
  302. attacker_port_mapping[ip_address] = next_port
  303. # Determine TTL value
  304. ttl = attacker_ttl_mapping.get(ip_address)
  305. if ttl is None: # determine TTL value
  306. is_invalid = True
  307. pos = ip_source_list.index(ip_address)
  308. pos_max = len(gd)
  309. while is_invalid:
  310. ttl = int(round(gd[pos]))
  311. if 0 < ttl < 256: # validity check
  312. is_invalid = False
  313. else:
  314. pos = (pos + 1) % pos_max
  315. attacker_ttl_mapping[ip_address] = ttl
  316. # return port and TTL
  317. return next_port, ttl
  318. def remove_generic_ending(string):
  319. """"
  320. Returns the input string with it's ending cut off, in case it was a generic one
  321. :param string: Input string
  322. :return: Input string with ending cut off
  323. """
  324. for end in generic_attack_names:
  325. if string.endswith(end):
  326. return string[:-len(end)]
  327. return string