From c8a80fd8184cddbd30ac8188ce04486886156365 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sergey=20M=E2=80=A4?= Date: Mon, 15 Feb 2016 22:29:28 +0600 Subject: [PATCH] [screenjunkies] Improve, extract more metadata and workaround subscription (Closes #8505) --- youtube_dl/extractor/screenjunkies.py | 122 ++++++++++++++++++++------ 1 file changed, 93 insertions(+), 29 deletions(-) diff --git a/youtube_dl/extractor/screenjunkies.py b/youtube_dl/extractor/screenjunkies.py index 5da3f9bcd..e0db180b8 100644 --- a/youtube_dl/extractor/screenjunkies.py +++ b/youtube_dl/extractor/screenjunkies.py @@ -1,74 +1,138 @@ from __future__ import unicode_literals import re -import json - -from .common import ( - InfoExtractor, - ExtractorError, -) +from .common import InfoExtractor +from ..compat import compat_str from ..utils import ( int_or_none, + parse_age_limit, ) -class ScreenJunkiesIE(InfoExtractor): - _VALID_URL = r'http://www.screenjunkies.com/video/(.+-(?P\d+)|.+)' +class ScreenJunkiesIE(InfoExtractor): + _VALID_URL = r'http://www.screenjunkies.com/video/(?P[^/]+?)(?:-(?P\d+))?(?:[/?#&]|$)' _TESTS = [{ 'url': 'http://www.screenjunkies.com/video/best-quentin-tarantino-movie-2841915', + 'md5': '5c2b686bec3d43de42bde9ec047536b0', 'info_dict': { 'id': '2841915', + 'display_id': 'best-quentin-tarantino-movie', 'ext': 'mp4', 'title': 'Best Quentin Tarantino Movie', + 'thumbnail': 're:^https?://.*\.jpg', + 'duration': 3671, + 'age_limit': 13, + 'tags': list, }, }, { 'url': 'http://www.screenjunkies.com/video/honest-trailers-the-dark-knight', 'info_dict': { 'id': '2348808', + 'display_id': 'honest-trailers-the-dark-knight', 'ext': 'mp4', - 'title': "Honest Trailers - 'The Dark Knight'", + 'title': "Honest Trailers: 'The Dark Knight'", + 'thumbnail': 're:^https?://.*\.jpg', + 'age_limit': 10, + 'tags': list, }, }, { + # requires subscription but worked around 'url': 'http://www.screenjunkies.com/video/knocking-dead-ep-1-the-show-so-far-3003285', - 'only_matching': True, + 'info_dict': { + 'id': '3003285', + 'display_id': 'knocking-dead-ep-1-the-show-so-far', + 'ext': 'mp4', + 'title': 'Knocking Dead Ep 1: State of The Dead Recap', + 'thumbnail': 're:^https?://.*\.jpg', + 'duration': 3307, + 'age_limit': 13, + 'tags': list, + }, }] + _DEFAULT_BITRATES = (48, 150, 496, 864, 2240) + def _real_extract(self, url): - video_id = self._match_id(url) + mobj = re.match(self._VALID_URL, url) + video_id = mobj.group('id') + display_id = mobj.group('display_id') - if not video_id: # Older urls didn't have the id in them, but we can grab it manually - webpage = self._download_webpage(url, url) - video_id = self._html_search_regex(r'src="/embed/(\d+)"', webpage, 'video id') + if not video_id: + webpage = self._download_webpage(url, display_id) + video_id = self._search_regex( + (r'src=["\']/embed/(\d+)', r'data-video-content-id=["\'](\d+)'), + webpage, 'video id') - webpage = self._download_webpage('http://www.screenjunkies.com/embed/%s' %video_id, video_id) - video_vars_str = self._html_search_regex(r'embedVars = (\{.+\})\s*', webpage, 'video variables', flags=re.S) - video_vars = self._parse_json(video_vars_str, video_id) + webpage = self._download_webpage( + 'http://www.screenjunkies.com/embed/%s' % video_id, + display_id, 'Downloading video embed page') + embed_vars = self._parse_json( + self._search_regex( + r'(?s)embedVars\s*=\s*({.+?})\s*', webpage, 'embed vars'), + display_id) - # TODO: Figure out required cookies - if video_vars['subscriptionLevel'] > 0: - raise ExtractorError('This video requires ScreenJunkiesPlus', expected=True) + title = embed_vars['contentName'] formats = [] - for f in video_vars['media']: - if not f['mediaPurpose'] == 'play': + bitrates = [] + for f in embed_vars.get('media', []): + if not f.get('uri') or f.get('mediaPurpose') != 'play': continue - + bitrate = int_or_none(f.get('bitRate')) + if bitrate: + bitrates.append(bitrate) formats.append({ 'url': f['uri'], + 'format_id': 'http-%d' % bitrate if bitrate else 'http', 'width': int_or_none(f.get('width')), 'height': int_or_none(f.get('height')), - 'tbr': int_or_none(f.get('bitRate')), + 'tbr': bitrate, 'format': 'mp4', }) + + if not bitrates: + # When subscriptionLevel > 0, i.e. plus subscription is required + # media list will be empty. However, hds and hls uri are still + # available. We can grabbing them assuming bitrates to be default. + bitrates = self._DEFAULT_BITRATES + + auth_token = embed_vars.get('AuthToken') + + def construct_manifest_url(base_url, ext): + pieces = [base_url] + pieces.extend([compat_str(b) for b in bitrates]) + pieces.append('_kbps.mp4.%s?%s' % (ext, auth_token)) + return ','.join(pieces) + + if bitrates and auth_token: + hds_url = embed_vars.get('hdsUri') + if hds_url: + f4m_formats = self._extract_f4m_formats( + construct_manifest_url(hds_url, 'f4m'), + display_id, f4m_id='hds', fatal=False) + if len(f4m_formats) == len(bitrates): + for f, bitrate in zip(f4m_formats, bitrates): + if not f.get('tbr'): + f['format_id'] = 'hds-%d' % bitrate + f['tbr'] = bitrate + # TODO: fix f4m downloader to handle manifests without bitrates if possible + # formats.extend(f4m_formats) + + hls_url = embed_vars.get('hlsUri') + if hls_url: + formats.extend(self._extract_m3u8_formats( + construct_manifest_url(hls_url, 'm3u8'), + display_id, 'mp4', entry_protocol='m3u8_native', m3u8_id='hls', fatal=False)) self._sort_formats(formats) return { 'id': video_id, - 'title': video_vars['contentName'], + 'display_id': display_id, + 'title': title, + 'thumbnail': embed_vars.get('thumbUri'), + 'duration': int_or_none(embed_vars.get('videoLengthInSeconds')) or None, + 'age_limit': parse_age_limit(embed_vars.get('audienceRating')), + 'tags': embed_vars.get('tags', '').split(','), 'formats': formats, - 'duration': int_or_none(video_vars.get('videoLengthInSeconds')), - 'thumbnail': video_vars.get('thumbUri'), - 'tags': video_vars['tags'].split(',') if 'tags' in video_vars else [], } -