# -*- coding: utf-8 -*-
"""
This module contains all logic necessary to decipher the signature.
YouTube's strategy to restrict downloading videos is to send a ciphered version
of the signature to the client, along with the decryption algorithm obfuscated
in JavaScript. For the clients to play the videos, JavaScript must take the
ciphered version, cycle it through a series of "transform functions," and then
signs the media URL with the output.
This module is responsible for (1) finding and extracting those "transform
functions" (2) maps them to Python equivalents and (3) taking the ciphered
signature and decoding it.
"""
import logging
import re
from itertools import chain
from typing import List, Tuple, Dict, Callable, Any, Optional
from pytube.exceptions import RegexMatchError
from pytube.helpers import regex_search, cache
logger = logging.getLogger(__name__)
class Cipher:
def __init__(self, js: str):
self.transform_plan: List[str] = get_transform_plan(js)
var, _ = self.transform_plan[0].split(".")
self.transform_map = get_transform_map(js, var)
self.js_func_regex = re.compile(r"\w+\.(\w+)\(\w,(\d+)\)")
def get_signature(self, ciphered_signature: str) -> str:
"""Decipher the signature.
Taking the ciphered signature, applies the transform functions.
:param str ciphered_signature:
The ciphered signature sent in the ``player_config``.
:rtype: str
:returns:
Decrypted signature required to download the media content.
"""
signature = list(ciphered_signature)
for js_func in self.transform_plan:
name, argument = self.parse_function(js_func) # type: ignore
signature = self.transform_map[name](signature, argument)
logger.debug(
"applied transform function\n"
"output: %s\n"
"js_function: %s\n"
"argument: %d\n"
"function: %s",
"".join(signature),
name,
argument,
self.transform_map[name],
)
return "".join(signature)
@cache
def parse_function(self, js_func: str) -> Tuple[str, int]:
"""Parse the Javascript transform function.
Break a JavaScript transform function down into a two element ``tuple``
containing the function name and some integer-based argument.
:param str js_func:
The JavaScript version of the transform function.
:rtype: tuple
:returns:
two element tuple containing the function name and an argument.
**Example**:
>>> parse_function('DE.AJ(a,15)')
('AJ', 15)
"""
logger.debug("parsing transform function")
parse_match = self.js_func_regex.search(js_func)
if not parse_match:
raise RegexMatchError(caller="parse_function", pattern="js_func_regex")
fn_name, fn_arg = parse_match.groups()
return fn_name, int(fn_arg)
[docs]def get_initial_function_name(js: str) -> str:
"""Extract the name of the function responsible for computing the signature.
:param str js:
The contents of the base.js asset file.
:rtype: str
:returns:
Function name from regex match
"""
function_patterns = [
r"\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
r"\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*encodeURIComponent\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
r'\b(?P<sig>[a-zA-Z0-9$]{2})\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)', # noqa: E501
r'(?P<sig>[a-zA-Z0-9$]+)\s*=\s*function\(\s*a\s*\)\s*{\s*a\s*=\s*a\.split\(\s*""\s*\)', # noqa: E501
r'(["\'])signature\1\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(',
r"\.sig\|\|(?P<sig>[a-zA-Z0-9$]+)\(",
r"yt\.akamaized\.net/\)\s*\|\|\s*.*?\s*[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?:encodeURIComponent\s*\()?\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
r"\b[cs]\s*&&\s*[adf]\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
r"\b[a-zA-Z0-9]+\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
r"\bc\s*&&\s*a\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
r"\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
r"\bc\s*&&\s*[a-zA-Z0-9]+\.set\([^,]+\s*,\s*\([^)]*\)\s*\(\s*(?P<sig>[a-zA-Z0-9$]+)\(", # noqa: E501
]
logger.debug("finding initial function name")
for pattern in function_patterns:
regex = re.compile(pattern)
function_match = regex.search(js)
if function_match:
logger.debug("finished regex search, matched: %s", pattern)
return function_match.group(1)
raise RegexMatchError(caller="get_initial_function_name", pattern="multiple")
[docs]def reverse(arr: List, _: Optional[Any]):
"""Reverse elements in a list.
This function is equivalent to:
.. code-block:: javascript
function(a, b) { a.reverse() }
This method takes an unused ``b`` variable as their transform functions
universally sent two arguments.
**Example**:
>>> reverse([1, 2, 3, 4])
[4, 3, 2, 1]
"""
return arr[::-1]
[docs]def splice(arr: List, b: int):
"""Add/remove items to/from a list.
This function is equivalent to:
.. code-block:: javascript
function(a, b) { a.splice(0, b) }
**Example**:
>>> splice([1, 2, 3, 4], 2)
[1, 2]
"""
return arr[:b] + arr[b * 2 :]
[docs]def swap(arr: List, b: int):
"""Swap positions at b modulus the list length.
This function is equivalent to:
.. code-block:: javascript
function(a, b) { var c=a[0];a[0]=a[b%a.length];a[b]=c }
**Example**:
>>> swap([1, 2, 3, 4], 2)
[3, 2, 1, 4]
"""
r = b % len(arr)
return list(chain([arr[r]], arr[1:r], [arr[0]], arr[r + 1 :]))
[docs]def map_functions(js_func: str) -> Callable:
"""For a given JavaScript transform function, return the Python equivalent.
:param str js_func:
The JavaScript version of the transform function.
"""
mapper = (
# function(a){a.reverse()}
(r"{\w\.reverse\(\)}", reverse),
# function(a,b){a.splice(0,b)}
(r"{\w\.splice\(0,\w\)}", splice),
# function(a,b){var c=a[0];a[0]=a[b%a.length];a[b]=c}
(r"{var\s\w=\w\[0\];\w\[0\]=\w\[\w\%\w.length\];\w\[\w\]=\w}", swap),
# function(a,b){var c=a[0];a[0]=a[b%a.length];a[b%a.length]=c}
(
r"{var\s\w=\w\[0\];\w\[0\]=\w\[\w\%\w.length\];\w\[\w\%\w.length\]=\w}",
swap,
),
)
for pattern, fn in mapper:
if re.search(pattern, js_func):
return fn
raise RegexMatchError(caller="map_functions", pattern="multiple")