131 lines
4.0 KiB
Python
131 lines
4.0 KiB
Python
from math import ceil
|
|
import base64
|
|
import io
|
|
|
|
# FROM https://github.com/user234683/youtube-local/blob/master/youtube/proto.py
|
|
|
|
def byte(n):
|
|
return bytes((n,))
|
|
|
|
|
|
def varint_encode(offset):
|
|
'''In this encoding system, for each 8-bit byte, the first bit is 1 if there are more bytes, and 0 is this is the last one.
|
|
The next 7 bits are data. These 7-bit sections represent the data in Little endian order. For example, suppose the data is
|
|
aaaaaaabbbbbbbccccccc (each of these sections is 7 bits). It will be encoded as:
|
|
1ccccccc 1bbbbbbb 0aaaaaaa
|
|
|
|
This encoding is used in youtube parameters to encode offsets and to encode the length for length-prefixed data.
|
|
See https://developers.google.com/protocol-buffers/docs/encoding#varints for more info.'''
|
|
needed_bytes = ceil(offset.bit_length()/7) or 1 # (0).bit_length() returns 0, but we need 1 in that case.
|
|
encoded_bytes = bytearray(needed_bytes)
|
|
for i in range(0, needed_bytes - 1):
|
|
encoded_bytes[i] = (offset & 127) | 128 # 7 least significant bits
|
|
offset = offset >> 7
|
|
encoded_bytes[-1] = offset & 127 # leave first bit as zero for last byte
|
|
|
|
return bytes(encoded_bytes)
|
|
|
|
|
|
def varint_decode(encoded):
|
|
decoded = 0
|
|
for i, byte in enumerate(encoded):
|
|
decoded |= (byte & 127) << 7*i
|
|
|
|
if not (byte & 128):
|
|
break
|
|
return decoded
|
|
|
|
|
|
def string(field_number, data):
|
|
data = as_bytes(data)
|
|
return _proto_field(2, field_number, varint_encode(len(data)) + data)
|
|
nested = string
|
|
|
|
def uint(field_number, value):
|
|
return _proto_field(0, field_number, varint_encode(value))
|
|
|
|
|
|
|
|
|
|
def _proto_field(wire_type, field_number, data):
|
|
''' See https://developers.google.com/protocol-buffers/docs/encoding#structure '''
|
|
return varint_encode( (field_number << 3) | wire_type) + data
|
|
|
|
|
|
|
|
def percent_b64encode(data):
|
|
return base64.urlsafe_b64encode(data).replace(b'=', b'%3D')
|
|
|
|
|
|
def unpadded_b64encode(data):
|
|
return base64.urlsafe_b64encode(data).replace(b'=', b'')
|
|
|
|
def as_bytes(value):
|
|
if isinstance(value, str):
|
|
return value.encode('utf-8')
|
|
return value
|
|
|
|
|
|
def read_varint(data):
|
|
result = 0
|
|
i = 0
|
|
while True:
|
|
try:
|
|
byte = data.read(1)[0]
|
|
except IndexError:
|
|
if i == 0:
|
|
raise EOFError()
|
|
raise Exception('Unterminated varint starting at ' + str(data.tell() - i))
|
|
result |= (byte & 127) << 7*i
|
|
if not byte & 128:
|
|
break
|
|
|
|
i += 1
|
|
return result
|
|
|
|
|
|
def read_group(data, end_sequence):
|
|
start = data.tell()
|
|
index = data.original.find(end_sequence, start)
|
|
if index == -1:
|
|
raise Exception('Unterminated group')
|
|
data.seek(index + len(end_sequence))
|
|
return data.original[start:index]
|
|
|
|
def read_protobuf(data):
|
|
data_original = data
|
|
data = io.BytesIO(data)
|
|
data.original = data_original
|
|
while True:
|
|
try:
|
|
tag = read_varint(data)
|
|
except EOFError:
|
|
break
|
|
wire_type = tag & 7
|
|
field_number = tag >> 3
|
|
|
|
if wire_type == 0:
|
|
value = read_varint(data)
|
|
elif wire_type == 1:
|
|
value = data.read(8)
|
|
elif wire_type == 2:
|
|
length = read_varint(data)
|
|
value = data.read(length)
|
|
elif wire_type == 3:
|
|
end_bytes = encode_varint((field_number << 3) | 4)
|
|
value = read_group(data, end_bytes)
|
|
elif wire_type == 5:
|
|
value = data.read(4)
|
|
else:
|
|
raise Exception("Unknown wire type: " + str(wire_type) + ", Tag: " + bytes_to_hex(succinct_encode(tag)) + ", at position " + str(data.tell()))
|
|
yield (wire_type, field_number, value)
|
|
|
|
def parse(data):
|
|
return {field_number: value for _, field_number, value in read_protobuf(data)}
|
|
|
|
def b64_to_bytes(data):
|
|
if isinstance(data, bytes):
|
|
data = data.decode('ascii')
|
|
data = data.replace("%3D", "=")
|
|
return base64.urlsafe_b64decode(data + "="*((4 - len(data)%4)%4) )
|