diff options
Diffstat (limited to 'youtube_dl/extractor/youtube.py')
-rw-r--r-- | youtube_dl/extractor/youtube.py | 235 |
1 files changed, 112 insertions, 123 deletions
diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 9e2373bd5..f49665925 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -5,8 +5,10 @@ import netrc import re import socket import itertools +import xml.etree.ElementTree from .common import InfoExtractor, SearchInfoExtractor +from .subtitles import SubtitlesInfoExtractor from ..utils import ( compat_http_client, compat_parse_qs, @@ -130,12 +132,13 @@ class YoutubeBaseInfoExtractor(InfoExtractor): return self._confirm_age() -class YoutubeIE(YoutubeBaseInfoExtractor): + +class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor): IE_DESC = u'YouTube.com' _VALID_URL = r"""^ ( (?:https?://)? # http(s):// (optional) - (?:youtu\.be/|(?:\w+\.)?youtube(?:-nocookie)?\.com/| + (?:(?:(?:(?:\w+\.)?youtube(?:-nocookie)?\.com/| tube\.majestyc\.net/) # the various hostnames, with wildcard subdomains (?:.*?\#/)? # handle anchor (#/) redirect urls (?: # the various things that can precede the ID: @@ -146,15 +149,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor): (?:.*?&)? # any other preceding param (like /?s=tuff&v=xxxx) v= ) - )? # optional -> youtube.com/xxxx is OK + )) + |youtu\.be/ # just youtu.be/xxxx + ) )? # all until now is optional -> you can pass the naked ID - ([0-9A-Za-z_-]+) # here is it! the YouTube video ID + ([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID (?(1).+)? # if we found the ID, everything can follow $""" _NEXT_URL_RE = r'[\?&]next_url=([^&]+)' # Listed in order of quality - _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '17', '13', - '95', '94', '93', '92', '132', '151', + _available_formats = ['38', '37', '46', '22', '45', '35', '44', '34', '18', '43', '6', '5', '36', '17', '13', + # Apple HTTP Live Streaming + '96', '95', '94', '93', '92', '132', '151', # 3D '85', '84', '102', '83', '101', '82', '100', # Dash video @@ -163,8 +169,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # Dash audio '141', '172', '140', '171', '139', ] - _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '17', '13', - '95', '94', '93', '92', '132', '151', + _available_formats_prefer_free = ['38', '46', '37', '45', '22', '44', '35', '43', '34', '18', '6', '5', '36', '17', '13', + # Apple HTTP Live Streaming + '96', '95', '94', '93', '92', '132', '151', + # 3D '85', '102', '84', '101', '83', '100', '82', # Dash video '138', '248', '137', '247', '136', '246', '245', @@ -172,11 +180,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # Dash audio '172', '141', '171', '140', '139', ] + _video_formats_map = { + 'flv': ['35', '34', '6', '5'], + '3gp': ['36', '17', '13'], + 'mp4': ['38', '37', '22', '18'], + 'webm': ['46', '45', '44', '43'], + } _video_extensions = { '13': '3gp', - '17': 'mp4', + '17': '3gp', '18': 'mp4', '22': 'mp4', + '36': '3gp', '37': 'mp4', '38': 'mp4', '43': 'webm', @@ -193,7 +208,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): '101': 'webm', '102': 'webm', - # videos that use m3u8 + # Apple HTTP Live Streaming '92': 'mp4', '93': 'mp4', '94': 'mp4', @@ -234,6 +249,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): '22': '720x1280', '34': '360x640', '35': '480x854', + '36': '240x320', '37': '1080x1920', '38': '3072x4096', '43': '360x640', @@ -373,7 +389,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): @classmethod def suitable(cls, url): """Receives a URL and returns True if suitable for this IE.""" - if YoutubePlaylistIE.suitable(url) or YoutubeSubscriptionsIE.suitable(url): return False + if YoutubePlaylistIE.suitable(url): return False return re.match(cls._VALID_URL, url, re.VERBOSE) is not None def report_video_webpage_download(self, video_id): @@ -384,19 +400,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor): """Report attempt to download video info webpage.""" self.to_screen(u'%s: Downloading video info webpage' % video_id) - def report_video_subtitles_download(self, video_id): - """Report attempt to download video info webpage.""" - self.to_screen(u'%s: Checking available subtitles' % video_id) - - def report_video_subtitles_request(self, video_id, sub_lang, format): - """Report attempt to download video info webpage.""" - self.to_screen(u'%s: Downloading video subtitles for %s.%s' % (video_id, sub_lang, format)) - - def report_video_subtitles_available(self, video_id, sub_lang_list): - """Report available subtitles.""" - sub_lang = ",".join(list(sub_lang_list.keys())) - self.to_screen(u'%s: Available subtitles for video: %s' % (video_id, sub_lang)) - def report_information_extraction(self, video_id): """Report attempt to extract video information.""" self.to_screen(u'%s: Extracting video information' % video_id) @@ -423,15 +426,15 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 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[83:36:-1] + s[0] + s[35:2:-1] + 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[83:34:-1] + s[0] + s[33:27:-1] + s[3] + s[26:19:-1] + s[34] + s[18:3:-1] + s[27] + 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] 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: @@ -451,56 +454,38 @@ class YoutubeIE(YoutubeBaseInfoExtractor): # Fallback to the other algortihms return self._decrypt_signature(s) - def _get_available_subtitles(self, video_id): - self.report_video_subtitles_download(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 {} - sub_lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list) - sub_lang_list = dict((l[1], l[0]) for l in sub_lang_list) + lang_list = re.findall(r'name="([^"]*)"[^>]+lang_code="([\w\-]+)"', sub_list) + + sub_lang_list = {} + for l in lang_list: + lang = l[1] + params = compat_urllib_parse.urlencode({ + 'lang': lang, + 'v': video_id, + 'fmt': self._downloader.params.get('subtitlesformat'), + }) + url = u'http://www.youtube.com/api/timedtext?' + params + sub_lang_list[lang] = url if not sub_lang_list: self._downloader.report_warning(u'video doesn\'t have subtitles') return {} return sub_lang_list - def _list_available_subtitles(self, video_id): - sub_lang_list = self._get_available_subtitles(video_id) - self.report_video_subtitles_available(video_id, sub_lang_list) - - def _request_subtitle(self, sub_lang, sub_name, video_id, format): - """ - Return the subtitle as a string or None if they are not found - """ - self.report_video_subtitles_request(video_id, sub_lang, format) - params = compat_urllib_parse.urlencode({ - 'lang': sub_lang, - 'name': sub_name, - 'v': video_id, - 'fmt': format, - }) - url = 'http://www.youtube.com/api/timedtext?' + params - try: - sub = compat_urllib_request.urlopen(url).read().decode('utf-8') - except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: - self._downloader.report_warning(u'unable to download video subtitles for %s: %s' % (sub_lang, compat_str(err))) - return - if not sub: - self._downloader.report_warning(u'Did not fetch video subtitles') - return - return sub - - 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 {} @@ -509,53 +494,38 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 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): self._downloader.report_warning(err_msg) return {} - - def _extract_subtitles(self, video_id): - """ - Return a dictionary: {language: subtitles} or {} if the subtitles - couldn't be found - """ - available_subs_list = self._get_available_subtitles(video_id) - sub_format = self._downloader.params.get('subtitlesformat') - if not available_subs_list: #There was some error, it didn't get the available subtitles - return {} - if self._downloader.params.get('allsubtitles', False): - sub_lang_list = available_subs_list - else: - if self._downloader.params.get('subtitleslangs', False): - reqested_langs = self._downloader.params.get('subtitleslangs') - elif 'en' in available_subs_list: - reqested_langs = ['en'] - else: - reqested_langs = [list(available_subs_list.keys())[0]] - - sub_lang_list = {} - for sub_lang in reqested_langs: - if not sub_lang in available_subs_list: - self._downloader.report_warning(u'no closed captions found in the specified language "%s"' % sub_lang) - continue - sub_lang_list[sub_lang] = available_subs_list[sub_lang] - subtitles = {} - for sub_lang in sub_lang_list: - subtitle = self._request_subtitle(sub_lang, sub_lang_list[sub_lang].encode('utf-8'), video_id, sub_format) - if subtitle: - subtitles[sub_lang] = subtitle - return subtitles def _print_formats(self, formats): print('Available formats:') @@ -597,13 +567,25 @@ class YoutubeIE(YoutubeBaseInfoExtractor): video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats else: # Specific formats. We pick the first in a slash-delimeted sequence. - # For example, if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'. + # Format can be specified as itag or 'mp4' or 'flv' etc. We pick the highest quality + # available in the specified format. For example, + # if '1/2/3/4' is requested and '2' and '4' are available, we pick '2'. + # if '1/mp4/3/4' is requested and '1' and '5' (is a mp4) are available, we pick '1'. + # if '1/mp4/3/4' is requested and '4' and '5' (is a mp4) are available, we pick '5'. req_formats = req_format.split('/') video_url_list = None for rf in req_formats: if rf in url_map: video_url_list = [(rf, url_map[rf])] break + if rf in self._video_formats_map: + for srf in self._video_formats_map[rf]: + if srf in url_map: + video_url_list = [(srf, url_map[srf])] + break + else: + continue + break if video_url_list is None: raise ExtractorError(u'requested format not available') return video_url_list @@ -743,15 +725,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 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: @@ -920,8 +897,11 @@ class YoutubePlaylistIE(InfoExtractor): for entry in response['feed']['entry']: index = entry['yt$position']['$t'] - if 'media$group' in entry and 'media$player' in entry['media$group']: - videos.append((index, entry['media$group']['media$player']['url'])) + if 'media$group' in entry and 'yt$videoid' in entry['media$group']: + videos.append(( + index, + 'https://www.youtube.com/watch?v=' + entry['media$group']['yt$videoid']['$t'] + )) videos = [v[1] for v in sorted(videos)] @@ -987,13 +967,20 @@ 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/)|ytuser:)([A-Za-z0-9_-]+)' + _VALID_URL = r'(?:(?:(?:https?://)?(?:\w+\.)?youtube\.com/(?:user/)?)|ytuser:)(?!feed/)([A-Za-z0-9_-]+)' _TEMPLATE_URL = 'http://gdata.youtube.com/feeds/api/users/%s' _GDATA_PAGE_SIZE = 50 - _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d' - _VIDEO_INDICATOR = r'/watch\?v=(.+?)[\<&]' + _GDATA_URL = 'http://gdata.youtube.com/feeds/api/users/%s/uploads?max-results=%d&start-index=%d&alt=json' IE_NAME = u'youtube:user' + @classmethod + def suitable(cls, url): + # Don't return True if the url can be extracted with other youtube + # extractor, the regex would is too permissive and it would match. + other_ies = iter(klass for (name, klass) in globals().items() if name.endswith('IE') and klass is not cls) + if any(ie.suitable(url) for ie in other_ies): return False + else: return super(YoutubeUserIE, cls).suitable(url) + def _real_extract(self, url): # Extract username mobj = re.match(self._VALID_URL, url) @@ -1016,13 +1003,15 @@ class YoutubeUserIE(InfoExtractor): page = self._download_webpage(gdata_url, username, u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE)) + try: + response = json.loads(page) + except ValueError as err: + raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err)) + # Extract video identifiers ids_in_page = [] - - for mobj in re.finditer(self._VIDEO_INDICATOR, page): - if mobj.group(1) not in ids_in_page: - ids_in_page.append(mobj.group(1)) - + for entry in response['feed']['entry']: + ids_in_page.append(entry['id']['$t'].split('/')[-1]) video_ids.extend(ids_in_page) # A little optimization - if current page is not @@ -1161,7 +1150,7 @@ class YoutubeWatchLaterIE(YoutubeFeedsInfoExtractor): class YoutubeFavouritesIE(YoutubeBaseInfoExtractor): IE_NAME = u'youtube:favorites' IE_DESC = u'YouTube.com favourite videos, "ytfav" keyword (requires authentication)' - _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:o?rites)?' + _VALID_URL = r'https?://www\.youtube\.com/my_favorites|:ytfav(?:ou?rites)?' _LOGIN_REQUIRED = True def _real_extract(self, url): |