279 lines
11 KiB
Python
279 lines
11 KiB
Python
# -*- coding: utf-8 -*-
|
|
"""
|
|
Created on Tue Sep 2 19:26:15 2014
|
|
|
|
@author: Anton Grigoryev
|
|
@author: Sammy Pfeiffer
|
|
"""
|
|
from binascii import crc32 as originalcrc32
|
|
def crc32(data):
|
|
return originalcrc32(data) & 0xffffffff
|
|
from datetime import datetime
|
|
from time import time
|
|
import io
|
|
import os.path
|
|
import json
|
|
import socket
|
|
import struct
|
|
|
|
def vis(bs):
|
|
"""
|
|
Function to visualize byte streams. Split into bytes, print to console.
|
|
:param bs: BYTE STRING
|
|
"""
|
|
bs = bytearray(bs)
|
|
symbols_in_one_line = 8
|
|
n = len(bs) // symbols_in_one_line
|
|
for i in range(n):
|
|
print(str(i*symbols_in_one_line)+" | "+" ".join(["%02X" % b for b in bs[i*symbols_in_one_line:(i+1)*symbols_in_one_line]])) # for every 8 symbols line
|
|
if not len(bs) % symbols_in_one_line == 0:
|
|
print(str((i+1)*symbols_in_one_line)+" | "+" ".join(["%02X" % b for b in bs[(i+1)*symbols_in_one_line:]])+"\n") # for last line
|
|
|
|
|
|
class TlConstructor:
|
|
def __init__(self, json_dict):
|
|
self.id = int(json_dict['id'])
|
|
self.type = json_dict['type']
|
|
self.predicate = json_dict['predicate']
|
|
self.params = []
|
|
# case of vector
|
|
for param in json_dict['params']:
|
|
if param['type'] == "Vector<long>":
|
|
param['type'] = "Vector t"
|
|
param['subtype'] = "long"
|
|
else:
|
|
param['subtype'] = None
|
|
self.params.append(param)
|
|
|
|
class TlMethod:
|
|
def __init__(self, json_dict):
|
|
self.id = int(json_dict['id'])
|
|
self.type = json_dict['type']
|
|
self.method = json_dict['method']
|
|
self.params = json_dict['params']
|
|
|
|
|
|
class TL:
|
|
def __init__(self, filename):
|
|
with open(filename, 'r') as f:
|
|
TL_dict = json.load(f)
|
|
|
|
# Read constructors
|
|
|
|
self.constructors = TL_dict['constructors']
|
|
self.constructor_id = {}
|
|
self.constructor_type = {}
|
|
for elem in self.constructors:
|
|
z = TlConstructor(elem)
|
|
self.constructor_id[z.id] = z
|
|
self.constructor_type[z.predicate] = z
|
|
|
|
self.methods = TL_dict['methods']
|
|
self.method_id = {}
|
|
self.method_name = {}
|
|
for elem in self.methods:
|
|
z = TlMethod(elem)
|
|
self.method_id[z.id] = z
|
|
self.method_name[z.method] = z
|
|
|
|
|
|
## Loading TL_schema (should be placed in the same directory as mtproto.py
|
|
tl = TL(os.path.join(os.path.dirname(__file__), "TL_schema.JSON"))
|
|
|
|
|
|
def serialize_obj(bytes_io, type_, **kwargs):
|
|
try:
|
|
tl_constructor = tl.constructor_type[type_]
|
|
except KeyError:
|
|
raise Exception("Could not extract type: %s" % type_)
|
|
bytes_io.write(struct.pack('<i', tl_constructor.id))
|
|
for arg in tl_constructor.params:
|
|
serialize_param(bytes_io, type_=arg['type'], value=kwargs[arg['name']])
|
|
|
|
|
|
def serialize_method(bytes_io, type_, **kwargs):
|
|
try:
|
|
tl_method = tl.method_name[type_]
|
|
except KeyError:
|
|
raise Exception("Could not extract type: %s" % type_)
|
|
bytes_io.write(struct.pack('<i', tl_method.id))
|
|
for arg in tl_method.params:
|
|
serialize_param(bytes_io, type_=arg['type'], value=kwargs[arg['name']])
|
|
|
|
|
|
def serialize_param(bytes_io, type_, value):
|
|
if type_ == "int":
|
|
assert isinstance(value, int)
|
|
bytes_io.write(struct.pack('<i', value))
|
|
elif type_ == "long":
|
|
assert isinstance(value, int)
|
|
bytes_io.write(struct.pack('<q', value))
|
|
elif type_ in ["int128", "int256"]:
|
|
assert isinstance(value, bytes)
|
|
bytes_io.write(value)
|
|
elif type_ == 'string' or 'bytes':
|
|
l = len(value)
|
|
if l < 254: # short string format
|
|
bytes_io.write(struct.pack('<b', l)) # 1 byte of string
|
|
#bytes_io.write(int.to_bytes(l, 1, 'little')) # 1 byte of string
|
|
bytes_io.write(value) # string
|
|
bytes_io.write(b'\x00'*((-l-1) % 4)) # padding bytes
|
|
else:
|
|
bytes_io.write(b'\xfe') # byte 254
|
|
bytes_io.write(struct.pack('<i', l)[:3]) # 3 bytes of string
|
|
#bytes_io.write(int.to_bytes(l, 3, 'little')) # 3 bytes of string
|
|
bytes_io.write(value) # string
|
|
bytes_io.write(b'\x00'*(-l % 4)) # padding bytes
|
|
|
|
def deserialize(bytes_io, type_=None, subtype=None):
|
|
"""
|
|
:type bytes_io: io.BytesIO object
|
|
"""
|
|
assert isinstance(bytes_io, io.BytesIO)
|
|
|
|
# Built-in bare types
|
|
if type_ == 'int': x = struct.unpack('<i', bytes_io.read(4))[0]
|
|
elif type_ == '#': x = struct.unpack('<I', bytes_io.read(4))[0]
|
|
elif type_ == 'long': x = struct.unpack('<q', bytes_io.read(8))[0]
|
|
elif type_ == 'double': x = struct.unpack('<d', bytes_io.read(8))[0]
|
|
elif type_ == 'int128': x = bytes_io.read(16)
|
|
elif type_ == 'int256': x = bytes_io.read(32)
|
|
elif type_ == 'string' or type_ == 'bytes':
|
|
l = struct.unpack('<B', bytes_io.read(1))[0]
|
|
assert l <= 254 # In general, 0xFF byte is not allowed here
|
|
if l == 254:
|
|
# We have a long string
|
|
long_len = struct.unpack('<I', bytes_io.read(3)+b'\x00')[0]
|
|
x = bytes_io.read(long_len)
|
|
bytes_io.read(-long_len % 4) # skip padding bytes
|
|
else:
|
|
# We have a short string
|
|
x = bytes_io.read(l)
|
|
bytes_io.read(-(l+1) % 4) # skip padding bytes
|
|
assert isinstance(x, bytes)
|
|
elif type_ == 'vector':
|
|
assert subtype is not None
|
|
count = struct.unpack('<l', bytes_io.read(4))[0]
|
|
print("count is: " + str(count))
|
|
x = [deserialize(bytes_io, type_=subtype) for i in range(count)]
|
|
else:
|
|
# Boxed types
|
|
i = struct.unpack('<i', bytes_io.read(4))[0] # read type ID
|
|
try:
|
|
tl_elem = tl.constructor_id[i]
|
|
except KeyError:
|
|
raise Exception("Could not extract type: %s" % type_)
|
|
base_boxed_types = ["Vector t", "Int", "Long", "Double", "String", "Int128", "Int256"]
|
|
if tl_elem.type in base_boxed_types:
|
|
x = deserialize(bytes_io, type_=tl_elem.predicate, subtype=subtype)
|
|
else: # other types
|
|
x = {}
|
|
for arg in tl_elem.params:
|
|
x[arg['name']] = deserialize(bytes_io, type_=arg['type'], subtype=arg['subtype'])
|
|
return x
|
|
|
|
class Session:
|
|
""" Manages TCP Transport. encryption and message frames """
|
|
def __init__(self, ip, port):
|
|
# creating socket
|
|
self.sock = socket.socket()
|
|
self.sock.connect((ip, port))
|
|
self.auth_key_id = None
|
|
self.number = 0
|
|
|
|
def __del__(self):
|
|
# closing socket when session object is deleted
|
|
self.sock.close()
|
|
|
|
@staticmethod
|
|
def header_unencrypted(message):
|
|
"""
|
|
Creating header for the unencrypted message:
|
|
:param message: byte string to send
|
|
"""
|
|
# Basic instructions: https://core.telegram.org/mtproto/description#unencrypted-message
|
|
|
|
# Message id: https://core.telegram.org/mtproto/description#message-identifier-msg-id
|
|
# http://stackoverflow.com/questions/8777753/converting-datetime-date-to-utc-timestamp-in-python
|
|
# to make it work in py2 and py3 (py3 has the timestamp() method but py2 doesnt)
|
|
curr_timestamp = (datetime.utcfromtimestamp(time()) - datetime(1970, 1, 1)).total_seconds()
|
|
msg_id = int(curr_timestamp*2**30)*4
|
|
#msg_id = int(datetime.utcnow().timestamp()*2**30)*4
|
|
|
|
return (b'\x00\x00\x00\x00\x00\x00\x00\x00' +
|
|
struct.pack('<Q', msg_id) +
|
|
struct.pack('<L', len(message)))
|
|
|
|
# TCP Transport
|
|
|
|
# Instructions may be found here: https://core.telegram.org/mtproto#tcp-transport
|
|
# If a payload (packet) needs to be transmitted from server to client or from client to server,
|
|
# it is encapsulated as follows: 4 length bytes are added at the front (to include the length,
|
|
# the sequence number, and CRC32; always divisible by 4) and 4 bytes with the packet sequence number
|
|
# within this TCP connection (the first packet sent is numbered 0, the next one 1, etc.),
|
|
# and 4 CRC32 bytes at the end (length, sequence number, and payload together).
|
|
|
|
def send_message(self, message):
|
|
"""
|
|
Forming the message frame and sending message to server
|
|
:param message: byte string to send
|
|
"""
|
|
|
|
|
|
data = self.header_unencrypted(message) + message
|
|
step1 = struct.pack('<LL', len(data)+12, self.number) + data
|
|
step2 = step1 + struct.pack('<L', crc32(step1))
|
|
self.sock.send(step2)
|
|
self.number += 1
|
|
# Sending message visualisation to console
|
|
print('>>')
|
|
vis(step2)
|
|
|
|
def recv_message(self):
|
|
"""
|
|
Reading socket and receiving message from server. Check the CRC32.
|
|
"""
|
|
packet_length_data = self.sock.recv(4) # reads how many bytes to read
|
|
|
|
if len(packet_length_data) > 0: # if we have smth. in the socket
|
|
packet_length = struct.unpack("<L", packet_length_data)[0]
|
|
packet = self.sock.recv(packet_length - 4) # read the rest of bytes from socket
|
|
|
|
# Received message visualisation to console
|
|
print('<<')
|
|
vis(packet_length_data+packet)
|
|
|
|
(x, auth_key_id, message_id, message_length)= struct.unpack("<L8s8sI", packet[0:24])
|
|
data = packet[24:24+message_length]
|
|
crc = packet[-4:]
|
|
# print("crc is: " + str(crc))
|
|
# print("type of crc: " + str(type(crc)))
|
|
# print("crc.__repr__(): " + crc.__repr__())
|
|
# print("struct.unpack('<L', crc): (next line)")
|
|
# print(struct.unpack('<L', crc))
|
|
# print("crc32(packet_length_data + packet[0:-4]): " + str(crc32(packet_length_data + packet[0:-4])))
|
|
# print("crc32(packet_length_data + packet[0:-4]).__repr__(): " + crc32(packet_length_data + packet[0:-4]).__repr__())
|
|
|
|
# Checking the CRC32 correctness of received data
|
|
if crc32(packet_length_data + packet[0:-4]) == struct.unpack('<L', crc)[0]:
|
|
|
|
return data
|
|
else:
|
|
raise Exception("CRC32 was not correct!")
|
|
else:
|
|
raise Exception("Nothing in the socket!")
|
|
|
|
def method_call(self, method, **kwargs):
|
|
z=io.BytesIO()
|
|
serialize_method(z, method, **kwargs)
|
|
# z.getvalue() on py2.7 returns str, which means bytes
|
|
# on py3.4 returns bytes
|
|
# bytearray is closer to the same data type to be shared
|
|
z_val = bytearray(z.getvalue())
|
|
# print("z_val: " + z_val.__repr__())
|
|
# print("z_val type: " + str(type(z_val)))
|
|
# print("len of z_val: " + str(len(z_val)))
|
|
self.send_message(z_val)
|
|
server_answer = self.recv_message()
|
|
return deserialize(io.BytesIO(server_answer))
|