# coding: utf-8
+import collections
+import itertools
+import io
import json
import netrc
import re
import socket
-import itertools
+import string
+import struct
+import traceback
+import zlib
from .common import InfoExtractor, SearchInfoExtractor
-from .subtitles import SubtitlesIE
+from .subtitles import SubtitlesInfoExtractor
from ..utils import (
compat_http_client,
compat_parse_qs,
(
(?:https?://)? # http(s):// (optional)
(?:(?:(?:(?:\w+\.)?youtube(?:-nocookie)?\.com/|
- tube\.majestyc\.net/) # the various hostnames, with wildcard subdomains
+ tube\.majestyc\.net/|
+ youtube\.googleapis\.com/) # the various hostnames, with wildcard subdomains
(?:.*?\#/)? # handle anchor (#/) redirect urls
(?: # the various things that can precede the ID:
(?:(?:v|embed|e)/) # v/ or embed/ or e/
if YoutubePlaylistIE.suitable(url): return False
return re.match(cls._VALID_URL, url, re.VERBOSE) is not None
+ def __init__(self, *args, **kwargs):
+ super(YoutubeIE, self).__init__(*args, **kwargs)
+ self._player_cache = {}
+
def report_video_webpage_download(self, video_id):
"""Report attempt to download video webpage."""
self.to_screen(u'%s: Downloading video webpage' % video_id)
"""Indicate the download will use the RTMP protocol."""
self.to_screen(u'RTMP download detected')
- def _decrypt_signature(self, s):
+ def _extract_signature_function(self, video_id, player_url):
+ id_m = re.match(r'.*-(?P<id>[a-zA-Z0-9]+)\.(?P<ext>[a-z]+)$',
+ player_url)
+ player_type = id_m.group('ext')
+ player_id = id_m.group('id')
+
+ # TODO read from filesystem cache
+
+ if player_type == 'js':
+ code = self._download_webpage(
+ player_url, video_id,
+ note=u'Downloading %s player %s' % (player_type, player_id),
+ errnote=u'Download of %s failed' % player_url)
+ res = self._parse_sig_js(code)
+ elif player_tpye == 'swf':
+ urlh = self._request_webpage(
+ player_url, video_id,
+ note=u'Downloading %s player %s' % (player_type, player_id),
+ errnote=u'Download of %s failed' % player_url)
+ code = urlh.read()
+ res = self._parse_sig_swf(code)
+ else:
+ assert False, 'Invalid player type %r' % player_type
+
+ # TODO write cache
+
+ return res
+
+ def _parse_sig_js(self, jscode):
+ funcname = self._search_regex(
+ r'signature=([a-zA-Z]+)', jscode,
+ u'Initial JS player signature function name')
+
+ functions = {}
+
+ def argidx(varname):
+ return string.lowercase.index(varname)
+
+ def interpret_statement(stmt, local_vars, allow_recursion=20):
+ if allow_recursion < 0:
+ raise ExctractorError(u'Recursion limit reached')
+
+ if stmt.startswith(u'var '):
+ stmt = stmt[len(u'var '):]
+ ass_m = re.match(r'^(?P<out>[a-z]+)(?:\[(?P<index>[^\]]+)\])?' +
+ r'=(?P<expr>.*)$', stmt)
+ if ass_m:
+ if ass_m.groupdict().get('index'):
+ def assign(val):
+ lvar = local_vars[ass_m.group('out')]
+ idx = interpret_expression(ass_m.group('index'),
+ local_vars, allow_recursion)
+ assert isinstance(idx, int)
+ lvar[idx] = val
+ return val
+ expr = ass_m.group('expr')
+ else:
+ def assign(val):
+ local_vars[ass_m.group('out')] = val
+ return val
+ expr = ass_m.group('expr')
+ elif stmt.startswith(u'return '):
+ assign = lambda v: v
+ expr = stmt[len(u'return '):]
+ else:
+ raise ExtractorError(
+ u'Cannot determine left side of statement in %r' % stmt)
+
+ v = interpret_expression(expr, local_vars, allow_recursion)
+ return assign(v)
+
+ def interpret_expression(expr, local_vars, allow_recursion):
+ if expr.isdigit():
+ return int(expr)
+
+ if expr.isalpha():
+ return local_vars[expr]
+
+ m = re.match(r'^(?P<in>[a-z]+)\.(?P<member>.*)$', expr)
+ if m:
+ member = m.group('member')
+ val = local_vars[m.group('in')]
+ if member == 'split("")':
+ return list(val)
+ if member == 'join("")':
+ return u''.join(val)
+ if member == 'length':
+ return len(val)
+ if member == 'reverse()':
+ return val[::-1]
+ slice_m = re.match(r'slice\((?P<idx>.*)\)', member)
+ if slice_m:
+ idx = interpret_expression(
+ slice_m.group('idx'), local_vars, allow_recursion-1)
+ return val[idx:]
+
+ m = re.match(
+ r'^(?P<in>[a-z]+)\[(?P<idx>.+)\]$', expr)
+ if m:
+ val = local_vars[m.group('in')]
+ idx = interpret_expression(m.group('idx'), local_vars,
+ allow_recursion-1)
+ return val[idx]
+
+ m = re.match(r'^(?P<a>.+?)(?P<op>[%])(?P<b>.+?)$', expr)
+ if m:
+ a = interpret_expression(m.group('a'),
+ local_vars, allow_recursion)
+ b = interpret_expression(m.group('b'),
+ local_vars, allow_recursion)
+ return a % b
+
+ m = re.match(
+ r'^(?P<func>[a-zA-Z]+)\((?P<args>[a-z0-9,]+)\)$', expr)
+ if m:
+ fname = m.group('func')
+ if fname not in functions:
+ functions[fname] = extract_function(fname)
+ argvals = [int(v) if v.isdigit() else local_vars[v]
+ for v in m.group('args').split(',')]
+ return functions[fname](argvals)
+ raise ExtractorError(u'Unsupported JS expression %r' % expr)
+
+ def extract_function(funcname):
+ func_m = re.search(
+ r'function ' + re.escape(funcname) +
+ r'\((?P<args>[a-z,]+)\){(?P<code>[^}]+)}',
+ jscode)
+ argnames = func_m.group('args').split(',')
+
+ def resf(args):
+ local_vars = dict(zip(argnames, args))
+ for stmt in func_m.group('code').split(';'):
+ res = interpret_statement(stmt, local_vars)
+ return res
+ return resf
+
+ initial_function = extract_function(funcname)
+ return lambda s: initial_function([s])
+
+ def _parse_sig_swf(self, file_contents):
+ if file_contents[1:3] != b'WS':
+ raise ExtractorError(
+ u'Not an SWF file; header is %r' % file_contents[:3])
+ if file_contents[:1] == b'C':
+ content = zlib.decompress(file_contents[8:])
+ else:
+ raise NotImplementedError(u'Unsupported compression format %r' %
+ file_contents[:1])
+
+ def extract_tags(content):
+ pos = 0
+ while pos < len(content):
+ header16 = struct.unpack('<H', content[pos:pos+2])[0]
+ pos += 2
+ tag_code = header16 >> 6
+ tag_len = header16 & 0x3f
+ if tag_len == 0x3f:
+ tag_len = struct.unpack('<I', content[pos:pos+4])[0]
+ pos += 4
+ assert pos+tag_len <= len(content)
+ yield (tag_code, content[pos:pos+tag_len])
+ pos += tag_len
+
+ code_tag = next(tag
+ for tag_code, tag in extract_tags(content)
+ if tag_code == 82)
+ p = code_tag.index(b'\0', 4) + 1
+ code_reader = io.BytesIO(code_tag[p:])
+
+ # Parse ABC (AVM2 ByteCode)
+ def read_int(reader=None):
+ if reader is None:
+ reader = code_reader
+ res = 0
+ shift = 0
+ for _ in range(5):
+ buf = reader.read(1)
+ assert len(buf) == 1
+ b = struct.unpack('<B', buf)[0]
+ res = res | ((b & 0x7f) << shift)
+ if b & 0x80 == 0:
+ break
+ shift += 7
+ return res
+
+ def u30(reader=None):
+ res = read_int(reader)
+ assert res & 0xf0000000 == 0
+ return res
+ u32 = read_int
+
+ def s32(reader=None):
+ v = read_int(reader)
+ if v & 0x80000000 != 0:
+ v = - ((v ^ 0xffffffff) + 1)
+ return v
+
+ def string(reader=None):
+ if reader is None:
+ reader = code_reader
+ slen = u30(reader)
+ resb = reader.read(slen)
+ assert len(resb) == slen
+ return resb.decode('utf-8')
+
+ def read_bytes(count, reader=None):
+ if reader is None:
+ reader = code_reader
+ resb = reader.read(count)
+ assert len(resb) == count
+ return resb
+
+ def read_byte(reader=None):
+ resb = read_bytes(1, reader=reader)
+ res = struct.unpack('<B', resb)[0]
+ return res
+
+ # minor_version + major_version
+ _ = read_bytes(2 + 2)
+
+ # Constant pool
+ int_count = u30()
+ for _c in range(1, int_count):
+ _ = s32()
+ uint_count = u30()
+ for _c in range(1, uint_count):
+ _ = u32()
+ double_count = u30()
+ _ = read_bytes((double_count-1) * 8)
+ string_count = u30()
+ constant_strings = [u'']
+ for _c in range(1, string_count):
+ s = string()
+ constant_strings.append(s)
+ namespace_count = u30()
+ for _c in range(1, namespace_count):
+ _ = read_bytes(1) # kind
+ _ = u30() # name
+ ns_set_count = u30()
+ for _c in range(1, ns_set_count):
+ count = u30()
+ for _c2 in range(count):
+ _ = u30()
+ multiname_count = u30()
+ MULTINAME_SIZES = {
+ 0x07: 2, # QName
+ 0x0d: 2, # QNameA
+ 0x0f: 1, # RTQName
+ 0x10: 1, # RTQNameA
+ 0x11: 0, # RTQNameL
+ 0x12: 0, # RTQNameLA
+ 0x09: 2, # Multiname
+ 0x0e: 2, # MultinameA
+ 0x1b: 1, # MultinameL
+ 0x1c: 1, # MultinameLA
+ }
+ multinames = [u'']
+ for _c in range(1, multiname_count):
+ kind = u30()
+ assert kind in MULTINAME_SIZES, u'Invalid multiname kind %r' % kind
+ if kind == 0x07:
+ namespace_idx = u30()
+ name_idx = u30()
+ multinames.append(constant_strings[name_idx])
+ else:
+ multinames.append('[MULTINAME kind: %d]' % kind)
+ for _c2 in range(MULTINAME_SIZES[kind]):
+ _ = u30()
+
+ # Methods
+ method_count = u30()
+ MethodInfo = collections.namedtuple(
+ 'MethodInfo',
+ ['NEED_ARGUMENTS', 'NEED_REST'])
+ method_infos = []
+ for method_id in range(method_count):
+ param_count = u30()
+ _ = u30() # return type
+ for _ in range(param_count):
+ _ = u30() # param type
+ _ = u30() # name index (always 0 for youtube)
+ flags = read_byte()
+ if flags & 0x08 != 0:
+ # Options present
+ option_count = u30()
+ for c in range(option_count):
+ _ = u30() # val
+ _ = read_bytes(1) # kind
+ if flags & 0x80 != 0:
+ # Param names present
+ for _ in range(param_count):
+ _ = u30() # param name
+ mi = MethodInfo(flags & 0x01 != 0, flags & 0x04 != 0)
+ method_infos.append(mi)
+
+ # Metadata
+ metadata_count = u30()
+ for _c in range(metadata_count):
+ _ = u30() # name
+ item_count = u30()
+ for _c2 in range(item_count):
+ _ = u30() # key
+ _ = u30() # value
+
+ def parse_traits_info():
+ trait_name_idx = u30()
+ kind_full = read_byte()
+ kind = kind_full & 0x0f
+ attrs = kind_full >> 4
+ methods = {}
+ if kind in [0x00, 0x06]: # Slot or Const
+ _ = u30() # Slot id
+ type_name_idx = u30()
+ vindex = u30()
+ if vindex != 0:
+ _ = read_byte() # vkind
+ elif kind in [0x01, 0x02, 0x03]: # Method / Getter / Setter
+ _ = u30() # disp_id
+ method_idx = u30()
+ methods[multinames[trait_name_idx]] = method_idx
+ elif kind == 0x04: # Class
+ _ = u30() # slot_id
+ _ = u30() # classi
+ elif kind == 0x05: # Function
+ _ = u30() # slot_id
+ function_idx = u30()
+ methods[function_idx] = multinames[trait_name_idx]
+ else:
+ raise ExtractorError(u'Unsupported trait kind %d' % kind)
+
+ if attrs & 0x4 != 0: # Metadata present
+ metadata_count = u30()
+ for _c3 in range(metadata_count):
+ _ = u30()
+
+ return methods
+
+ # Classes
+ TARGET_CLASSNAME = u'SignatureDecipher'
+ searched_idx = multinames.index(TARGET_CLASSNAME)
+ searched_class_id = None
+ class_count = u30()
+ for class_id in range(class_count):
+ name_idx = u30()
+ if name_idx == searched_idx:
+ # We found the class we're looking for!
+ searched_class_id = class_id
+ _ = u30() # super_name idx
+ flags = read_byte()
+ if flags & 0x08 != 0: # Protected namespace is present
+ protected_ns_idx = u30()
+ intrf_count = u30()
+ for _c2 in range(intrf_count):
+ _ = u30()
+ _ = u30() # iinit
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ _ = parse_traits_info()
+
+ if searched_class_id is None:
+ raise ExtractorError(u'Target class %r not found' %
+ TARGET_CLASSNAME)
+
+ method_names = {}
+ method_idxs = {}
+ for class_id in range(class_count):
+ _ = u30() # cinit
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ trait_methods = parse_traits_info()
+ if class_id == searched_class_id:
+ method_names.update(trait_methods.items())
+ method_idxs.update(dict(
+ (idx, name)
+ for name, idx in trait_methods.items()))
+
+ # Scripts
+ script_count = u30()
+ for _c in range(script_count):
+ _ = u30() # init
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ _ = parse_traits_info()
+
+ # Method bodies
+ method_body_count = u30()
+ Method = collections.namedtuple('Method', ['code', 'local_count'])
+ methods = {}
+ for _c in range(method_body_count):
+ method_idx = u30()
+ max_stack = u30()
+ local_count = u30()
+ init_scope_depth = u30()
+ max_scope_depth = u30()
+ code_length = u30()
+ code = read_bytes(code_length)
+ if method_idx in method_idxs:
+ m = Method(code, local_count)
+ methods[method_idxs[method_idx]] = m
+ exception_count = u30()
+ for _c2 in range(exception_count):
+ _ = u30() # from
+ _ = u30() # to
+ _ = u30() # target
+ _ = u30() # exc_type
+ _ = u30() # var_name
+ trait_count = u30()
+ for _c2 in range(trait_count):
+ _ = parse_traits_info()
+
+ assert p + code_reader.tell() == len(code_tag)
+ assert len(methods) == len(method_idxs)
+
+ method_pyfunctions = {}
+
+ def extract_function(func_name):
+ if func_name in method_pyfunctions:
+ return method_pyfunctions[func_name]
+ if func_name not in methods:
+ raise ExtractorError(u'Cannot find function %r' % func_name)
+ m = methods[func_name]
+
+ def resfunc(args):
+ registers = ['(this)'] + list(args) + [None] * m.local_count
+ stack = []
+ coder = io.BytesIO(m.code)
+ while True:
+ opcode = struct.unpack('!B', coder.read(1))[0]
+ if opcode == 36: # pushbyte
+ v = struct.unpack('!B', coder.read(1))[0]
+ stack.append(v)
+ elif opcode == 44: # pushstring
+ idx = u30(coder)
+ stack.append(constant_strings[idx])
+ elif opcode == 48: # pushscope
+ # We don't implement the scope register, so we'll just
+ # ignore the popped value
+ stack.pop()
+ elif opcode == 70: # callproperty
+ index = u30(coder)
+ mname = multinames[index]
+ arg_count = u30(coder)
+ args = list(reversed(
+ [stack.pop() for _ in range(arg_count)]))
+ obj = stack.pop()
+ if mname == u'split':
+ assert len(args) == 1
+ assert isinstance(args[0], compat_str)
+ assert isinstance(obj, compat_str)
+ if args[0] == u'':
+ res = list(obj)
+ else:
+ res = obj.split(args[0])
+ stack.append(res)
+ elif mname == u'slice':
+ assert len(args) == 1
+ assert isinstance(args[0], int)
+ assert isinstance(obj, list)
+ res = obj[args[0]:]
+ stack.append(res)
+ elif mname == u'join':
+ assert len(args) == 1
+ assert isinstance(args[0], compat_str)
+ assert isinstance(obj, list)
+ res = args[0].join(obj)
+ stack.append(res)
+ elif mname in method_pyfunctions:
+ stack.append(method_pyfunctions[mname](args))
+ else:
+ raise NotImplementedError(
+ u'Unsupported property %r on %r'
+ % (mname, obj))
+ elif opcode == 72: # returnvalue
+ res = stack.pop()
+ return res
+ elif opcode == 79: # callpropvoid
+ index = u30(coder)
+ mname = multinames[index]
+ arg_count = u30(coder)
+ args = list(reversed(
+ [stack.pop() for _ in range(arg_count)]))
+ obj = stack.pop()
+ if mname == u'reverse':
+ assert isinstance(obj, list)
+ obj.reverse()
+ else:
+ raise NotImplementedError(
+ u'Unsupported (void) property %r on %r'
+ % (mname, obj))
+ elif opcode == 93: # findpropstrict
+ index = u30(coder)
+ mname = multinames[index]
+ res = extract_function(mname)
+ stack.append(res)
+ elif opcode == 97: # setproperty
+ index = u30(coder)
+ value = stack.pop()
+ idx = stack.pop()
+ obj = stack.pop()
+ assert isinstance(obj, list)
+ assert isinstance(idx, int)
+ obj[idx] = value
+ elif opcode == 98: # getlocal
+ index = u30(coder)
+ stack.append(registers[index])
+ elif opcode == 99: # setlocal
+ index = u30(coder)
+ value = stack.pop()
+ registers[index] = value
+ elif opcode == 102: # getproperty
+ index = u30(coder)
+ pname = multinames[index]
+ if pname == u'length':
+ obj = stack.pop()
+ assert isinstance(obj, list)
+ stack.append(len(obj))
+ else: # Assume attribute access
+ idx = stack.pop()
+ assert isinstance(idx, int)
+ obj = stack.pop()
+ assert isinstance(obj, list)
+ stack.append(obj[idx])
+ elif opcode == 128: # coerce
+ _ = u30(coder)
+ elif opcode == 133: # coerce_s
+ assert isinstance(stack[-1], (type(None), compat_str))
+ elif opcode == 164: # modulo
+ value2 = stack.pop()
+ value1 = stack.pop()
+ res = value1 % value2
+ stack.append(res)
+ elif opcode == 208: # getlocal_0
+ stack.append(registers[0])
+ elif opcode == 209: # getlocal_1
+ stack.append(registers[1])
+ elif opcode == 210: # getlocal_2
+ stack.append(registers[2])
+ elif opcode == 211: # getlocal_3
+ stack.append(registers[3])
+ elif opcode == 214: # setlocal_2
+ registers[2] = stack.pop()
+ elif opcode == 215: # setlocal_3
+ registers[3] = stack.pop()
+ else:
+ raise NotImplementedError(
+ u'Unsupported opcode %d' % opcode)
+
+ method_pyfunctions[func_name] = resfunc
+ return resfunc
+
+ initial_function = extract_function(u'decipher')
+ return lambda s: initial_function([s])
+
+ def _decrypt_signature(self, s, video_id, player_url, age_gate=False):
"""Turn the encrypted s field into a working signature"""
+ if player_url is not None:
+ try:
+ if player_url not in self._player_cache:
+ func = self._extract_signature_function(
+ video_id, player_url
+ )
+ self._player_cache[player_url] = func
+ return self._player_cache[player_url](s)
+ except Exception as e:
+ tb = traceback.format_exc()
+ self._downloader.report_warning(
+ u'Automatic signature extraction failed: ' + tb)
+
+ self._downloader.report_warning(
+ u'Warning: Falling back to static signature algorithm')
+ return self._static_decrypt_signature(
+ s, video_id, player_url, age_gate)
+
+ def _static_decrypt_signature(self, s, video_id, player_url, age_gate):
+ if age_gate:
+ # The videos with age protection use another player, so the
+ # algorithms can be different.
+ if len(s) == 86:
+ return s[2:63] + s[82] + s[64:82] + s[63]
+
if len(s) == 92:
return s[25] + s[3:25] + s[0] + s[26:42] + s[79] + s[43:79] + s[91] + s[80:83]
elif len(s) == 90:
elif len(s) == 86:
return s[5:34] + s[0] + s[35:38] + s[3] + s[39:45] + s[38] + s[46:53] + s[73] + s[54:73] + s[85] + s[74:85] + s[53]
elif len(s) == 85:
- return s[40] + s[82:43:-1] + s[22] + s[42:40:-1] + s[83] + s[39:22:-1] + s[0] + s[21:2:-1]
+ return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84]
elif len(s) == 84:
return s[81:36:-1] + s[0] + s[35:2:-1]
elif len(s) == 83:
return s[81:64:-1] + s[82] + s[63:52:-1] + s[45] + s[51:45:-1] + s[1] + s[44:1:-1] + s[0]
elif len(s) == 82:
- return s[1:19] + s[0] + s[20:68] + s[19] + s[69:82]
+ return s[80:73:-1] + s[81] + s[72:54:-1] + s[2] + s[53:43:-1] + s[0] + s[42:2:-1] + s[43] + s[1] + s[54]
elif len(s) == 81:
return s[56] + s[79:56:-1] + s[41] + s[55:41:-1] + s[80] + s[40:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9]
elif len(s) == 80:
return self._decrypt_signature(s)
def _get_available_subtitles(self, video_id):
- request = compat_urllib_request.Request('http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id)
try:
- sub_list = compat_urllib_request.urlopen(request).read().decode('utf-8')
- except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
+ sub_list = self._download_webpage(
+ 'http://video.google.com/timedtext?hl=en&type=list&v=%s' % video_id,
+ video_id, note=False)
+ except ExtractorError as err:
self._downloader.report_warning(u'unable to download video subtitles: %s' % compat_str(err))
return {}
lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list)
return {}
return sub_lang_list
- def _request_automatic_caption(self, video_id, webpage):
+ def _get_available_automatic_caption(self, video_id, webpage):
"""We need the webpage for getting the captions url, pass it as an
argument to speed up the process."""
- sub_lang = (self._downloader.params.get('subtitleslangs') or ['en'])[0]
sub_format = self._downloader.params.get('subtitlesformat')
self.to_screen(u'%s: Looking for automatic captions' % video_id)
mobj = re.search(r';ytplayer.config = ({.*?});', webpage)
- err_msg = u'Couldn\'t find automatic captions for "%s"' % sub_lang
+ err_msg = u'Couldn\'t find automatic captions for %s' % video_id
if mobj is None:
self._downloader.report_warning(err_msg)
return {}
args = player_config[u'args']
caption_url = args[u'ttsurl']
timestamp = args[u'timestamp']
- params = compat_urllib_parse.urlencode({
- 'lang': 'en',
- 'tlang': sub_lang,
- 'fmt': sub_format,
- 'ts': timestamp,
- 'kind': 'asr',
+ # We get the available subtitles
+ list_params = compat_urllib_parse.urlencode({
+ 'type': 'list',
+ 'tlangs': 1,
+ 'asrs': 1,
})
- subtitles_url = caption_url + '&' + params
- sub = self._download_webpage(subtitles_url, video_id, u'Downloading automatic captions')
- return {sub_lang: sub}
+ list_url = caption_url + '&' + list_params
+ list_page = self._download_webpage(list_url, video_id)
+ caption_list = xml.etree.ElementTree.fromstring(list_page.encode('utf-8'))
+ original_lang_node = caption_list.find('track')
+ if original_lang_node.attrib.get('kind') != 'asr' :
+ self._downloader.report_warning(u'Video doesn\'t have automatic captions')
+ return {}
+ original_lang = original_lang_node.attrib['lang_code']
+
+ sub_lang_list = {}
+ for lang_node in caption_list.findall('target'):
+ sub_lang = lang_node.attrib['lang_code']
+ params = compat_urllib_parse.urlencode({
+ 'lang': original_lang,
+ 'tlang': sub_lang,
+ 'fmt': sub_format,
+ 'ts': timestamp,
+ 'kind': 'asr',
+ })
+ sub_lang_list[sub_lang] = caption_url + '&' + params
+ return sub_lang_list
# An extractor error can be raise by the download process if there are
# no automatic captions but there are subtitles
except (KeyError, ExtractorError):
video_webpage = video_webpage_bytes.decode('utf-8', 'ignore')
# Attempt to extract SWF player URL
- mobj = re.search(r'swfConfig.*?"(http:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
+ mobj = re.search(r'swfConfig.*?"(https?:\\/\\/.*?watch.*?-.*?\.swf)"', video_webpage)
if mobj is not None:
player_url = re.sub(r'\\(.)', r'\1', mobj.group(1))
else:
video_description = u''
# subtitles
- video_subtitles = None
-
- if self._downloader.params.get('writesubtitles', False) or self._downloader.params.get('allsubtitles', False):
- video_subtitles = self._extract_subtitles(video_id)
- elif self._downloader.params.get('writeautomaticsub', False):
- video_subtitles = self._request_automatic_caption(video_id, video_webpage)
+ video_subtitles = self.extract_subtitles(video_id, video_webpage)
if self._downloader.params.get('listsubtitles', False):
- self._list_available_subtitles(video_id)
+ self._list_available_subtitles(video_id, video_webpage)
return
if 'length_seconds' not in video_info:
if 'sig' in url_data:
url += '&signature=' + url_data['sig'][0]
elif 's' in url_data:
+ encrypted_sig = url_data['s'][0]
if self._downloader.params.get('verbose'):
- s = url_data['s'][0]
if age_gate:
- player_version = self._search_regex(r'ad3-(.+?)\.swf',
- video_info['ad3_module'][0] if 'ad3_module' in video_info else 'NOT FOUND',
+ player_version = self._search_regex(
+ r'-(.+)\.swf$',
+ player_url if player_url else None,
'flash player', fatal=False)
- player = 'flash player %s' % player_version
+ player_desc = 'flash player %s' % player_version
else:
- player = u'html5 player %s' % self._search_regex(r'html5player-(.+?)\.js', video_webpage,
+ player_version = self._search_regex(
+ r'html5player-(.+?)\.js', video_webpage,
'html5 player', fatal=False)
- parts_sizes = u'.'.join(compat_str(len(part)) for part in s.split('.'))
+ player_desc = u'html5 player %s' % player_version
+
+ parts_sizes = u'.'.join(compat_str(len(part)) for part in encrypted_sig.split('.'))
self.to_screen(u'encrypted signature length %d (%s), itag %s, %s' %
- (len(s), parts_sizes, url_data['itag'][0], player))
- encrypted_sig = url_data['s'][0]
- if age_gate:
- signature = self._decrypt_signature_age_gate(encrypted_sig)
- else:
- signature = self._decrypt_signature(encrypted_sig)
+ (len(encrypted_sig), parts_sizes, url_data['itag'][0], player_desc))
+
+ if not age_gate:
+ jsplayer_url_json = self._search_regex(
+ r'"assets":.+?"js":\s*("[^"]+")',
+ video_webpage, u'JS player URL')
+ player_url = json.loads(jsplayer_url_json)
+
+ signature = self._decrypt_signature(
+ encrypted_sig, video_id, player_url, age_gate)
url += '&signature=' + signature
if 'ratebypass' not in url:
url += '&ratebypass=yes'
response = json.loads(page)
except ValueError as err:
raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
+ if 'entry' not in response['feed']:
+ # Number of videos is a multiple of self._MAX_RESULTS
+ break
# Extract video identifiers
ids_in_page = []