X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/af478477605bdf3f5d57562035885cfee905f379..1931fa5f23ac75ea766517aa1b5ebf267ad3f97d:/youtube_dl/extractor/youtube.py?ds=sidebyside diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index a810368..3c9b0b5 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -1,19 +1,17 @@ # coding: utf-8 -import collections import errno import io import itertools import json import os.path import re -import string -import struct import traceback -import zlib from .common import InfoExtractor, SearchInfoExtractor from .subtitles import SubtitlesInfoExtractor +from ..jsinterp import JSInterpreter +from ..swfinterp import SWFInterpreter from ..utils import ( compat_chr, compat_parse_qs, @@ -29,7 +27,6 @@ from ..utils import ( ExtractorError, int_or_none, PagedList, - RegexNotFoundError, unescapeHTML, unified_strdate, orderedSet, @@ -138,19 +135,21 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): (?:(?:(?:(?:\w+\.)?[yY][oO][uU][tT][uU][bB][eE](?:-nocookie)?\.com/| (?:www\.)?deturl\.com/www\.youtube\.com/| (?:www\.)?pwnyoutube\.com/| + (?:www\.)?yourepeat\.com/| 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/ |(?: # or the v= param in all its forms - (?:(?:watch|movie)(?:_popup)?(?:\.php)?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx) + (?:(?:watch|movie)(?:_popup)?(?:\.php)?/?)? # preceding watch(_popup|.php) or nothing (like /?v=xxxx) (?:\?|\#!?) # the params delimiter ? or # or #! (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx) v= ) )) |youtu\.be/ # just youtu.be/xxxx + |https?://(?:www\.)?cleanvideosearch\.com/media/action/yt/watch\?videoId= ) )? # all until now is optional -> you can pass the naked ID ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID @@ -176,32 +175,32 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # 3d videos - '82': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20}, - '83': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20}, - '84': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20}, - '85': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': '3D', 'preference': -20}, - '100': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': '3D', 'preference': -20}, - '101': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': '3D', 'preference': -20}, - '102': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': '3D', 'preference': -20}, + '82': {'ext': 'mp4', 'height': 360, 'format_note': '3D', 'preference': -20}, + '83': {'ext': 'mp4', 'height': 480, 'format_note': '3D', 'preference': -20}, + '84': {'ext': 'mp4', 'height': 720, 'format_note': '3D', 'preference': -20}, + '85': {'ext': 'mp4', 'height': 1080, 'format_note': '3D', 'preference': -20}, + '100': {'ext': 'webm', 'height': 360, 'format_note': '3D', 'preference': -20}, + '101': {'ext': 'webm', 'height': 480, 'format_note': '3D', 'preference': -20}, + '102': {'ext': 'webm', 'height': 720, 'format_note': '3D', 'preference': -20}, # Apple HTTP Live Streaming - '92': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10}, - '93': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'HLS', 'preference': -10}, - '94': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'HLS', 'preference': -10}, - '95': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'HLS', 'preference': -10}, - '96': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'HLS', 'preference': -10}, - '132': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'HLS', 'preference': -10}, - '151': {'ext': 'mp4', 'height': 72, 'resolution': '72p', 'format_note': 'HLS', 'preference': -10}, + '92': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10}, + '93': {'ext': 'mp4', 'height': 360, 'format_note': 'HLS', 'preference': -10}, + '94': {'ext': 'mp4', 'height': 480, 'format_note': 'HLS', 'preference': -10}, + '95': {'ext': 'mp4', 'height': 720, 'format_note': 'HLS', 'preference': -10}, + '96': {'ext': 'mp4', 'height': 1080, 'format_note': 'HLS', 'preference': -10}, + '132': {'ext': 'mp4', 'height': 240, 'format_note': 'HLS', 'preference': -10}, + '151': {'ext': 'mp4', 'height': 72, 'format_note': 'HLS', 'preference': -10}, # DASH mp4 video - '133': {'ext': 'mp4', 'height': 240, 'resolution': '240p', 'format_note': 'DASH video', 'preference': -40}, - '134': {'ext': 'mp4', 'height': 360, 'resolution': '360p', 'format_note': 'DASH video', 'preference': -40}, - '135': {'ext': 'mp4', 'height': 480, 'resolution': '480p', 'format_note': 'DASH video', 'preference': -40}, - '136': {'ext': 'mp4', 'height': 720, 'resolution': '720p', 'format_note': 'DASH video', 'preference': -40}, - '137': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40}, - '138': {'ext': 'mp4', 'height': 1081, 'resolution': '>1080p', 'format_note': 'DASH video', 'preference': -40}, - '160': {'ext': 'mp4', 'height': 192, 'resolution': '192p', 'format_note': 'DASH video', 'preference': -40}, - '264': {'ext': 'mp4', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH video', 'preference': -40}, + '133': {'ext': 'mp4', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '134': {'ext': 'mp4', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '135': {'ext': 'mp4', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '136': {'ext': 'mp4', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '137': {'ext': 'mp4', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '138': {'ext': 'mp4', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '160': {'ext': 'mp4', 'height': 144, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '264': {'ext': 'mp4', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, # Dash mp4 audio '139': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 48, 'preference': -50}, @@ -209,23 +208,25 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): '141': {'ext': 'm4a', 'format_note': 'DASH audio', 'vcodec': 'none', 'abr': 256, 'preference': -50}, # Dash webm - '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, - '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, - '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, - '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, - '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, - '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'container': 'webm', 'vcodec': 'VP8', 'acodec': 'none', 'preference': -40}, - '242': {'ext': 'webm', 'height': 240, 'resolution': '240p', 'format_note': 'DASH webm', 'preference': -40}, - '243': {'ext': 'webm', 'height': 360, 'resolution': '360p', 'format_note': 'DASH webm', 'preference': -40}, - '244': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}, - '245': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}, - '246': {'ext': 'webm', 'height': 480, 'resolution': '480p', 'format_note': 'DASH webm', 'preference': -40}, - '247': {'ext': 'webm', 'height': 720, 'resolution': '720p', 'format_note': 'DASH webm', 'preference': -40}, - '248': {'ext': 'webm', 'height': 1080, 'resolution': '1080p', 'format_note': 'DASH webm', 'preference': -40}, + '167': {'ext': 'webm', 'height': 360, 'width': 640, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, + '168': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, + '169': {'ext': 'webm', 'height': 720, 'width': 1280, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, + '170': {'ext': 'webm', 'height': 1080, 'width': 1920, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, + '218': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, + '219': {'ext': 'webm', 'height': 480, 'width': 854, 'format_note': 'DASH video', 'acodec': 'none', 'container': 'webm', 'vcodec': 'VP8', 'preference': -40}, + '242': {'ext': 'webm', 'height': 240, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '243': {'ext': 'webm', 'height': 360, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '244': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '245': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '246': {'ext': 'webm', 'height': 480, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '247': {'ext': 'webm', 'height': 720, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '248': {'ext': 'webm', 'height': 1080, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '271': {'ext': 'webm', 'height': 1440, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, + '272': {'ext': 'webm', 'height': 2160, 'format_note': 'DASH video', 'acodec': 'none', 'preference': -40}, # Dash webm audio - '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 48, 'preference': -50}, - '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH webm audio', 'abr': 256, 'preference': -50}, + '171': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 48, 'preference': -50}, + '172': {'ext': 'webm', 'vcodec': 'none', 'format_note': 'DASH audio', 'abr': 256, 'preference': -50}, # RTMP (unnamed) '_rtmp': {'protocol': 'rtmp'}, @@ -241,7 +242,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): u"uploader": u"Philipp Hagemeister", u"uploader_id": u"phihag", u"upload_date": u"20121002", - u"description": u"test chars: \"'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de ." + u"description": u"test chars: \"'/\\ä↭𝕐\ntest URL: https://github.com/rg3/youtube-dl/issues/1892\n\nThis is a test video for youtube-dl.\n\nFor more information, contact phihag@phihag.de .", + u"categories": [u'Science & Technology'], } }, { @@ -251,7 +253,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): u"info_dict": { u"upload_date": u"20120506", u"title": u"Icona Pop - I Love It (feat. Charli XCX) [OFFICIAL VIDEO]", - u"description": u"md5:5b292926389560516e384ac437c0ec07", + u"description": u"md5:fea86fda2d5a5784273df5c7cc994d9f", u"uploader": u"Icona Pop", u"uploader_id": u"IconaPop" } @@ -296,6 +298,23 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): u"format": "141", }, }, + # DASH manifest with encrypted signature + { + u'url': u'https://www.youtube.com/watch?v=IB3lcPjvWLA', + u'info_dict': { + u'id': u'IB3lcPjvWLA', + u'ext': u'm4a', + u'title': u'Afrojack - The Spark ft. Spree Wilson', + u'description': u'md5:9717375db5a9a3992be4668bbf3bc0a8', + u'uploader': u'AfrojackVEVO', + u'uploader_id': u'AfrojackVEVO', + u'upload_date': u'20131011', + }, + u"params": { + u'youtube_include_dash_manifest': True, + u'format': '141', + }, + }, ] @@ -325,14 +344,22 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): """Indicate the download will use the RTMP protocol.""" self.to_screen(u'RTMP download detected') - def _extract_signature_function(self, video_id, player_url, slen): - id_m = re.match(r'.*-(?P[a-zA-Z0-9_-]+)\.(?P[a-z]+)$', - player_url) + def _signature_cache_id(self, example_sig): + """ Return a string representation of a signature """ + return u'.'.join(compat_str(len(part)) for part in example_sig.split('.')) + + def _extract_signature_function(self, video_id, player_url, example_sig): + id_m = re.match( + r'.*-(?P[a-zA-Z0-9_-]+)(?:/watch_as3|/html5player)?\.(?P[a-z]+)$', + player_url) + if not id_m: + raise ExtractorError('Cannot identify player %r' % player_url) player_type = id_m.group('ext') player_id = id_m.group('id') # Read from filesystem cache - func_id = '%s_%s_%d' % (player_type, player_id, slen) + func_id = '%s_%s_%s' % ( + player_type, player_id, self._signature_cache_id(example_sig)) assert os.path.basename(func_id) == func_id cache_dir = get_cachedir(self._downloader.params) @@ -366,7 +393,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): if cache_enabled: try: - test_string = u''.join(map(compat_chr, range(slen))) + test_string = u''.join(map(compat_chr, range(len(example_sig)))) cache_res = res(test_string) cache_spec = [ord(c) for c in cache_res] try: @@ -382,7 +409,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): return res - def _print_sig_code(self, func, slen): + def _print_sig_code(self, func, example_sig): def gen_sig_code(idxs): def _genslice(start, end, step): starts = u'' if start == 0 else str(start) @@ -411,607 +438,55 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): else: yield _genslice(start, i, step) - test_string = u''.join(map(compat_chr, range(slen))) + test_string = u''.join(map(compat_chr, range(len(example_sig)))) cache_res = func(test_string) cache_spec = [ord(c) for c in cache_res] expr_code = u' + '.join(gen_sig_code(cache_spec)) - code = u'if len(s) == %d:\n return %s\n' % (slen, expr_code) + signature_id_tuple = '(%s)' % ( + ', '.join(compat_str(len(p)) for p in example_sig.split('.'))) + code = (u'if tuple(len(p) for p in s.split(\'.\')) == %s:\n' + u' return %s\n') % (signature_id_tuple, expr_code) self.to_screen(u'Extracted signature function:\n' + code) 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 ExtractorError(u'Recursion limit reached') - - if stmt.startswith(u'var '): - stmt = stmt[len(u'var '):] - ass_m = re.match(r'^(?P[a-z]+)(?:\[(?P[^\]]+)\])?' + - r'=(?P.*)$', stmt) - if ass_m: - if ass_m.groupdict().get('index'): - def assign(val): - lvar = local_vars[ass_m.group('out')] - idx = interpret_expression(ass_m.group('index'), - local_vars, allow_recursion) - assert isinstance(idx, int) - lvar[idx] = val - return val - expr = ass_m.group('expr') - else: - def assign(val): - local_vars[ass_m.group('out')] = val - return val - expr = ass_m.group('expr') - elif stmt.startswith(u'return '): - assign = lambda v: v - expr = stmt[len(u'return '):] - else: - raise ExtractorError( - u'Cannot determine left side of statement in %r' % stmt) - - v = interpret_expression(expr, local_vars, allow_recursion) - return assign(v) - - def interpret_expression(expr, local_vars, allow_recursion): - if expr.isdigit(): - return int(expr) - - if expr.isalpha(): - return local_vars[expr] - - m = re.match(r'^(?P[a-z]+)\.(?P.*)$', 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\((?P.*)\)', member) - if slice_m: - idx = interpret_expression( - slice_m.group('idx'), local_vars, allow_recursion-1) - return val[idx:] - - m = re.match( - r'^(?P[a-z]+)\[(?P.+)\]$', expr) - if m: - val = local_vars[m.group('in')] - idx = interpret_expression(m.group('idx'), local_vars, - allow_recursion-1) - return val[idx] - - m = re.match(r'^(?P.+?)(?P[%])(?P.+?)$', 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[a-zA-Z$]+)\((?P[a-z0-9,]+)\)$', 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'\((?P[a-z,]+)\){(?P[^}]+)}', - 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) + r'signature=([$a-zA-Z]+)', jscode, + u'Initial JS player signature function name') + + jsi = JSInterpreter(jscode) + initial_function = jsi.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('> 6 - tag_len = header16 & 0x3f - if tag_len == 0x3f: - tag_len = struct.unpack('> 4 - methods = {} - if kind in [0x00, 0x06]: # Slot or Const - u30() # Slot id - u30() # type_name_idx - 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() # metadata index - - return methods - - # Classes + swfi = SWFInterpreter(file_contents) 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 - u30() # protected_ns_idx - 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() - u30() # max_stack - local_count = u30() - u30() # init_scope_depth - u30() # max_scope_depth - 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') + searched_class = swfi.extract_class(TARGET_CLASSNAME) + initial_function = swfi.extract_function(searched_class, 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: - if player_url.startswith(u'//'): - player_url = u'https:' + player_url - try: - player_id = (player_url, len(s)) - if player_id not in self._player_cache: - func = self._extract_signature_function( - video_id, player_url, len(s) - ) - self._player_cache[player_id] = func - func = self._player_cache[player_id] - if self._downloader.params.get('youtube_print_sig_code'): - self._print_sig_code(func, len(s)) - return func(s) - except Exception: - 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) == 93: - return s[86:29:-1] + s[88] + s[28:5:-1] - elif 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) == 91: - return s[84:27:-1] + s[86] + s[26:5:-1] - elif len(s) == 90: - return s[25] + s[3:25] + s[2] + s[26:40] + s[77] + s[41:77] + s[89] + s[78:81] - elif len(s) == 89: - return s[84:78:-1] + s[87] + s[77:60:-1] + s[0] + s[59:3:-1] - elif len(s) == 88: - return s[7:28] + s[87] + s[29:45] + s[55] + s[46:55] + s[2] + s[56:87] + s[28] - elif len(s) == 87: - return s[6:27] + s[4] + s[28:39] + s[27] + s[40:59] + s[2] + s[60:] - elif len(s) == 86: - return s[80:72:-1] + s[16] + s[71:39:-1] + s[72] + s[38:16:-1] + s[82] + s[15::-1] - elif len(s) == 85: - return s[3:11] + s[0] + s[12:55] + s[84] + s[56:84] - elif len(s) == 84: - return s[78:70:-1] + s[14] + s[69:37:-1] + s[70] + s[36:14:-1] + s[80] + s[:14][::-1] - elif len(s) == 83: - return s[80:63:-1] + s[0] + s[62:0:-1] + s[63] - elif len(s) == 82: - return s[80:37:-1] + s[7] + s[36:7:-1] + s[0] + s[6:0:-1] + s[37] - 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 s[1:19] + s[0] + s[20:68] + s[19] + s[69:80] - elif len(s) == 79: - return s[54] + s[77:54:-1] + s[39] + s[53:39:-1] + s[78] + s[38:34:-1] + s[0] + s[33:29:-1] + s[34] + s[28:9:-1] + s[29] + s[8:0:-1] + s[9] + if player_url is None: + raise ExtractorError(u'Cannot decrypt signature without player_url') - else: - raise ExtractorError(u'Unable to decrypt signature, key length %d not supported; retrying might work' % (len(s))) + if player_url.startswith(u'//'): + player_url = u'https:' + player_url + try: + player_id = (player_url, self._signature_cache_id(s)) + if player_id not in self._player_cache: + func = self._extract_signature_function( + video_id, player_url, s + ) + self._player_cache[player_id] = func + func = self._player_cache[player_id] + if self._downloader.params.get('youtube_print_sig_code'): + self._print_sig_code(func, s) + return func(s) + except Exception as e: + tb = traceback.format_exc() + raise ExtractorError( + u'Signature extraction failed: ' + tb, cause=e) def _get_available_subtitles(self, video_id, webpage): try: @@ -1113,14 +588,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): return self._download_webpage(url, video_id, note=u'Searching for annotations.', errnote=u'Unable to download video annotations.') def _real_extract(self, url): + proto = ( + u'http' if self._downloader.params.get('prefer_insecure', False) + else u'https') + # Extract original video URL from URL with redirection, like age verification, using next_url parameter mobj = re.search(self._NEXT_URL_RE, url) if mobj: - url = 'https://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/') + url = proto + '://www.youtube.com/' + compat_urllib_parse.unquote(mobj.group(1)).lstrip('/') video_id = self.extract_id(url) # Get video webpage - url = 'https://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id + url = proto + '://www.youtube.com/watch?v=%s&gl=US&hl=en&has_verified=1' % video_id video_webpage = self._download_webpage(url, video_id) # Attempt to extract SWF player URL @@ -1137,15 +616,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): age_gate = True # We simulate the access to the video from www.youtube.com/v/{video_id} # this can be viewed without login into Youtube - data = compat_urllib_parse.urlencode({'video_id': video_id, - 'el': 'player_embedded', - 'gl': 'US', - 'hl': 'en', - 'eurl': 'https://youtube.googleapis.com/v/' + video_id, - 'asv': 3, - 'sts':'1588', - }) - video_info_url = 'https://www.youtube.com/get_video_info?' + data + data = compat_urllib_parse.urlencode({ + 'video_id': video_id, + 'eurl': 'https://youtube.googleapis.com/v/' + video_id, + 'sts': self._search_regex( + r'"sts"\s*:\s*(\d+)', video_webpage, 'sts'), + }) + video_info_url = proto + '://www.youtube.com/get_video_info?' + data video_info_webpage = self._download_webpage(video_info_url, video_id, note=False, errnote='unable to download video info webpage') @@ -1153,7 +630,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): else: age_gate = False for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']: - video_info_url = ('https://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en' + video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en' % (video_id, el_type)) video_info_webpage = self._download_webpage(video_info_url, video_id, note=False, @@ -1163,9 +640,13 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): break if 'token' not in video_info: if 'reason' in video_info: - raise ExtractorError(u'YouTube said: %s' % video_info['reason'][0], expected=True) + raise ExtractorError( + u'YouTube said: %s' % video_info['reason'][0], + expected=True, video_id=video_id) else: - raise ExtractorError(u'"token" parameter not in video info for unknown reason') + raise ExtractorError( + u'"token" parameter not in video info for unknown reason', + video_id=video_id) if 'view_count' in video_info: view_count = int(video_info['view_count'][0]) @@ -1194,7 +675,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # title if 'title' in video_info: - video_title = compat_urllib_parse.unquote_plus(video_info['title'][0]) + video_title = video_info['title'][0] else: self._downloader.report_warning(u'Unable to extract video title') video_title = u'_' @@ -1213,11 +694,24 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # upload date upload_date = None - mobj = re.search(r'id="eow-date.*?>(.*?)', video_webpage, re.DOTALL) + mobj = re.search(r'(?s)id="eow-date.*?>(.*?)', video_webpage) + if mobj is None: + mobj = re.search( + r'(?s)id="watch-uploader-info".*?>.*?(?:Published|Uploaded|Streamed live) on (.*?)', + video_webpage) if mobj is not None: upload_date = ' '.join(re.sub(r'[/,-]', r' ', mobj.group(1)).split()) upload_date = unified_strdate(upload_date) + m_cat_container = get_element_by_id("eow-category", video_webpage) + if m_cat_container: + category = self._html_search_regex( + r'(?s)(.*?)', m_cat_container, 'category', + default=None) + video_categories = None if category is None else [category] + else: + video_categories = None + # description video_description = get_element_by_id("eow-description", video_webpage) if video_description: @@ -1268,11 +762,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): # Decide which formats to download try: - mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage) + mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage) if not mobj: raise ValueError('Could not find vevo ID') - info = json.loads(mobj.group(1)) - args = info['args'] + json_code = uppercase_escape(mobj.group(1)) + ytplayer_config = json.loads(json_code) + args = ytplayer_config['args'] # Easy way to know if the 's' value is in url_encoded_fmt_stream_map # this signatures are encrypted if 'url_encoded_fmt_stream_map' not in args: @@ -1319,43 +814,54 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): url_map = {} for url_data_str in encoded_url_map.split(','): url_data = compat_parse_qs(url_data_str) - if 'itag' in url_data and 'url' in url_data: - url = url_data['url'][0] - 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'): - if age_gate: - if player_url is None: - player_version = 'unknown' - else: - player_version = self._search_regex( - r'-(.+)\.swf$', player_url, - u'flash player', fatal=False) + if 'itag' not in url_data or 'url' not in url_data: + continue + format_id = url_data['itag'][0] + url = url_data['url'][0] + + if 'sig' in url_data: + url += '&signature=' + url_data['sig'][0] + elif 's' in url_data: + encrypted_sig = url_data['s'][0] + + 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) + if player_url is None: + player_url_json = self._search_regex( + r'ytplayer\.config.*?"url"\s*:\s*("[^"]+")', + video_webpage, u'age gate player URL') + player_url = json.loads(player_url_json) + + if self._downloader.params.get('verbose'): + if player_url is None: + player_version = 'unknown' + player_desc = 'unknown' + else: + if player_url.endswith('swf'): + player_version = self._search_regex( + r'-(.+?)(?:/watch_as3)?\.swf$', player_url, + u'flash player', fatal=False) player_desc = 'flash player %s' % player_version else: player_version = self._search_regex( - r'html5player-(.+?)\.js', video_webpage, + r'html5player-([^/]+?)(?:/html5player)?\.js', + player_url, 'html5 player', fatal=False) 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(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' - url_map[url_data['itag'][0]] = url + parts_sizes = self._signature_cache_id(encrypted_sig) + self.to_screen(u'{%s} signature length %s, %s' % + (format_id, parts_sizes, player_desc)) + + signature = self._decrypt_signature( + encrypted_sig, video_id, player_url, age_gate) + url += '&signature=' + signature + if 'ratebypass' not in url: + url += '&ratebypass=yes' + url_map[format_id] = url formats = _map_to_format_list(url_map) elif video_info.get('hlsvp'): manifest_url = video_info['hlsvp'][0] @@ -1365,12 +871,24 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info') # Look for the DASH manifest - dash_manifest_url_lst = video_info.get('dashmpd') - if (dash_manifest_url_lst and dash_manifest_url_lst[0] and - self._downloader.params.get('youtube_include_dash_manifest', False)): + if (self._downloader.params.get('youtube_include_dash_manifest', False)): try: + # The DASH manifest used needs to be the one from the original video_webpage. + # The one found in get_video_info seems to be using different signatures. + # However, in the case of an age restriction there won't be any embedded dashmpd in the video_webpage. + # Luckily, it seems, this case uses some kind of default signature (len == 86), so the + # combination of get_video_info and the _static_decrypt_signature() decryption fallback will work here. + if age_gate: + dash_manifest_url = video_info.get('dashmpd')[0] + else: + dash_manifest_url = ytplayer_config['args']['dashmpd'] + def decrypt_sig(mobj): + s = mobj.group(1) + dec_s = self._decrypt_signature(s, video_id, player_url, age_gate) + return '/signature/%s' % dec_s + dash_manifest_url = re.sub(r'/s/([\w\.]+)', decrypt_sig, dash_manifest_url) dash_doc = self._download_xml( - dash_manifest_url_lst[0], video_id, + dash_manifest_url, video_id, note=u'Downloading DASH manifest', errnote=u'Could not download DASH manifest') for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'): @@ -1411,11 +929,12 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): 'title': video_title, 'thumbnail': video_thumbnail, 'description': video_description, + 'categories': video_categories, 'subtitles': video_subtitles, 'duration': video_duration, 'age_limit': 18 if age_gate else 0, 'annotations': video_annotations, - 'webpage_url': 'https://www.youtube.com/watch?v=%s' % video_id, + 'webpage_url': proto + '://www.youtube.com/watch?v=%s' % video_id, 'view_count': view_count, 'like_count': like_count, 'dislike_count': dislike_count, @@ -1434,17 +953,17 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): | p/ ) ( - (?:PL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,} + (?:PL|LL|EC|UU|FL|RD)?[0-9A-Za-z-_]{10,} # Top tracks, they can also include dots |(?:MC)[\w\.]* ) .* | - ((?:PL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,}) + ((?:PL|LL|EC|UU|FL|RD)[0-9A-Za-z-_]{10,}) )""" - _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s&page=%s' + _TEMPLATE_URL = 'https://www.youtube.com/playlist?list=%s' _MORE_PAGES_INDICATOR = r'data-link-type="next"' - _VIDEO_RE = r'href="/watch\?v=(?P[0-9A-Za-z_-]{11})&[^"]*?index=(?P\d+)' + _VIDEO_RE = r'href="\s*/watch\?v=(?P[0-9A-Za-z_-]{11})&[^"]*?index=(?P\d+)' IE_NAME = u'youtube:playlist' def _real_initialize(self): @@ -1459,11 +978,13 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): # the id of the playlist is just 'RD' + video_id url = 'https://youtube.com/watch?v=%s&list=%s' % (playlist_id[-11:], playlist_id) webpage = self._download_webpage(url, playlist_id, u'Downloading Youtube mix') - title_span = (get_element_by_attribute('class', 'title long-title', webpage) or - get_element_by_attribute('class', 'title ', webpage)) + search_title = lambda class_name: get_element_by_attribute('class', class_name, webpage) + title_span = (search_title('playlist-title') or + search_title('title long-title') or search_title('title')) title = clean_html(title_span) - video_re = r'data-index="\d+".*?href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s' % re.escape(playlist_id) - ids = orderedSet(re.findall(video_re, webpage)) + video_re = r'''(?x)data-video-username=".*?".*? + href="/watch\?v=([0-9A-Za-z_-]{11})&[^"]*?list=%s''' % re.escape(playlist_id) + ids = orderedSet(re.findall(video_re, webpage, flags=re.DOTALL)) url_results = self._ids_to_results(ids) return self.playlist_result(url_results, playlist_id, title) @@ -1483,7 +1004,7 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): self.to_screen(u'Downloading just video %s because of --no-playlist' % video_id) return self.url_result(video_id, 'Youtube', video_id=video_id) else: - self.to_screen(u'Downloading playlist PL%s - add --no-playlist to just download video %s' % (playlist_id, video_id)) + self.to_screen(u'Downloading playlist %s - add --no-playlist to just download video %s' % (playlist_id, video_id)) if playlist_id.startswith('RD'): # Mixes require a custom extraction process @@ -1492,29 +1013,41 @@ class YoutubePlaylistIE(YoutubeBaseInfoExtractor): raise ExtractorError(u'For downloading YouTube.com top lists, use ' u'the "yttoplist" keyword, for example "youtube-dl \'yttoplist:music:Top Tracks\'"', expected=True) + url = self._TEMPLATE_URL % playlist_id + page = self._download_webpage(url, playlist_id) + more_widget_html = content_html = page + + # Check if the playlist exists or is private + if re.search(r'
[^<]*?(The|This) playlist (does not exist|is private)[^<]*?
', page) is not None: + raise ExtractorError( + u'The playlist doesn\'t exist or is private, use --username or ' + '--netrc to access it.', + expected=True) + # Extract the video ids from the playlist pages ids = [] for page_num in itertools.count(1): - url = self._TEMPLATE_URL % (playlist_id, page_num) - page = self._download_webpage(url, playlist_id, u'Downloading page #%s' % page_num) - matches = re.finditer(self._VIDEO_RE, page) + matches = re.finditer(self._VIDEO_RE, content_html) # We remove the duplicates and the link with index 0 # (it's not the first video of the playlist) new_ids = orderedSet(m.group('id') for m in matches if m.group('index') != '0') ids.extend(new_ids) - if re.search(self._MORE_PAGES_INDICATOR, page) is None: + mobj = re.search(r'data-uix-load-more-href="/?(?P[^"]+)"', more_widget_html) + if not mobj: break - try: - playlist_title = self._og_search_title(page) - except RegexNotFoundError: - self.report_warning( - u'Playlist page is missing OpenGraph title, falling back ...', - playlist_id) - playlist_title = self._html_search_regex( - r'

(.*?)

', page, u'title') + more = self._download_json( + 'https://youtube.com/%s' % mobj.group('more'), playlist_id, + 'Downloading page #%s' % page_num, + transform_source=uppercase_escape) + content_html = more['content_html'] + more_widget_html = more['load_more_widget_html'] + + playlist_title = self._html_search_regex( + r'(?s)

\s*(.*?)\s*

', + page, u'title') url_results = self._ids_to_results(ids) return self.playlist_result(url_results, playlist_id, playlist_title) @@ -1610,7 +1143,7 @@ class YoutubeChannelIE(InfoExtractor): class YoutubeUserIE(InfoExtractor): IE_DESC = u'YouTube.com user videos (URL or "ytuser" keyword)' - _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)' + _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?(?!(?:attribution_link|watch|results)(?:$|[^a-z_A-Z0-9-])))|ytuser:)(?!feed/)([A-Za-z0-9_-]+)' _TEMPLATE_URL = 'https://gdata.youtube.com/feeds/api/users/%s' _GDATA_PAGE_SIZE = 50 _GDATA_URL = 'https://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json' @@ -1672,7 +1205,7 @@ class YoutubeUserIE(InfoExtractor): class YoutubeSearchIE(SearchInfoExtractor): IE_DESC = u'YouTube.com searches' - _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc' + _API_URL = u'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc' _MAX_RESULTS = 1000 IE_NAME = u'youtube:search' _SEARCH_KEY = 'ytsearch' @@ -1683,9 +1216,12 @@ class YoutubeSearchIE(SearchInfoExtractor): video_ids = [] pagenum = 0 limit = n + PAGE_SIZE = 50 - while (50 * pagenum) < limit: - result_url = self._API_URL % (compat_urllib_parse.quote_plus(query), (50*pagenum)+1) + while (PAGE_SIZE * pagenum) < limit: + result_url = self._API_URL % ( + compat_urllib_parse.quote_plus(query.encode('utf-8')), + (PAGE_SIZE * pagenum) + 1) data_json = self._download_webpage( result_url, video_id=u'query "%s"' % query, note=u'Downloading page %s' % (pagenum + 1), @@ -1709,12 +1245,50 @@ class YoutubeSearchIE(SearchInfoExtractor): for video_id in video_ids] return self.playlist_result(videos, query) + class YoutubeSearchDateIE(YoutubeSearchIE): IE_NAME = YoutubeSearchIE.IE_NAME + ':date' _API_URL = 'https://gdata.youtube.com/feeds/api/videos?q=%s&start-index=%i&max-results=50&v=2&alt=jsonc&orderby=published' _SEARCH_KEY = 'ytsearchdate' IE_DESC = u'YouTube.com searches, newest videos first' + +class YoutubeSearchURLIE(InfoExtractor): + IE_DESC = u'YouTube.com search URLs' + IE_NAME = u'youtube:search_url' + _VALID_URL = r'https?://(?:www\.)?youtube\.com/results\?(.*?&)?search_query=(?P[^&]+)(?:[&]|$)' + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + query = compat_urllib_parse.unquote_plus(mobj.group('query')) + + webpage = self._download_webpage(url, query) + result_code = self._search_regex( + r'(?s)
    ', webpage, u'result HTML') + + part_codes = re.findall( + r'(?s)

    (.*?)

    ', result_code) + entries = [] + for part_code in part_codes: + part_title = self._html_search_regex( + [r'(?s)title="([^"]+)"', r'>([^<]+)'], part_code, 'item title', fatal=False) + part_url_snippet = self._html_search_regex( + r'(?s)href="([^"]+)"', part_code, 'item URL') + part_url = compat_urlparse.urljoin( + 'https://www.youtube.com/', part_url_snippet) + entries.append({ + '_type': 'url', + 'url': part_url, + 'title': part_title, + }) + + return { + '_type': 'playlist', + 'entries': entries, + 'title': query, + } + + class YoutubeShowIE(InfoExtractor): IE_DESC = u'YouTube.com (multi-season) shows' _VALID_URL = r'https?://www\.youtube\.com/show/(.*)' @@ -1758,23 +1332,25 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): feed_entries = [] paging = 0 for i in itertools.count(1): - info = self._download_webpage(self._FEED_TEMPLATE % paging, + info = self._download_json(self._FEED_TEMPLATE % paging, u'%s feed' % self._FEED_NAME, u'Downloading page %s' % i) - info = json.loads(info) - feed_html = info['feed_html'] + feed_html = info.get('feed_html') or info.get('content_html') m_ids = re.finditer(r'"/watch\?v=(.*?)["&]', feed_html) ids = orderedSet(m.group(1) for m in m_ids) feed_entries.extend( self.url_result(video_id, 'Youtube', video_id=video_id) for video_id in ids) - if info['paging'] is None: + mobj = re.search( + r'data-uix-load-more-href="/?[^"]+paging=(?P\d+)', + feed_html) + if mobj is None: break - paging = info['paging'] + paging = mobj.group('paging') return self.playlist_result(feed_entries, playlist_title=self._PLAYLIST_TITLE) class YoutubeSubscriptionsIE(YoutubeFeedsInfoExtractor): - IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword(requires authentication)' + IE_DESC = u'YouTube.com subscriptions feed, "ytsubs" keyword (requires authentication)' _VALID_URL = r'https?://www\.youtube\.com/feed/subscriptions|:ytsubs(?:criptions)?' _FEED_NAME = 'subscriptions' _PLAYLIST_TITLE = u'Youtube Subscriptions' @@ -1815,10 +1391,21 @@ class YoutubeTruncatedURLIE(InfoExtractor): IE_NAME = 'youtube:truncated_url' IE_DESC = False # Do not list _VALID_URL = r'''(?x) - (?:https?://)?[^/]+/watch\?feature=[a-z_]+$| + (?:https?://)?[^/]+/watch\?(?: + feature=[a-z_]+| + annotation_id=annotation_[^&]+ + )?$| (?:https?://)?(?:www\.)?youtube\.com/attribution_link\?a=[^&]+$ ''' + _TESTS = [{ + 'url': 'http://www.youtube.com/watch?annotation_id=annotation_3951667041', + 'only_matching': True, + }, { + 'url': 'http://www.youtube.com/watch?', + 'only_matching': True, + }] + def _real_extract(self, url): raise ExtractorError( u'Did you forget to quote the URL? Remember that & is a meta '