from __future__ import unicode_literals import base64 import re from .common import InfoExtractor from ..compat import compat_urllib_parse_unquote from ..utils import ( ExtractorError, clean_html, determine_ext, int_or_none, js_to_json, parse_age_limit, parse_duration, ) class ViewLiftBaseIE(InfoExtractor): _DOMAINS_REGEX = r'(?:snagfilms|snagxtreme|funnyforfree|kiddovid|winnersview|(?:monumental|lax)sportsnetwork|vayafilm)\.com' class ViewLiftEmbedIE(ViewLiftBaseIE): _VALID_URL = r'https?://(?:(?:www|embed)\.)?(?:%s)/embed/player\?.*\bfilmId=(?P[\da-f]{8}-(?:[\da-f]{4}-){3}[\da-f]{12})' % ViewLiftBaseIE._DOMAINS_REGEX _TESTS = [{ 'url': 'http://embed.snagfilms.com/embed/player?filmId=74849a00-85a9-11e1-9660-123139220831&w=500', 'md5': '2924e9215c6eff7a55ed35b72276bd93', 'info_dict': { 'id': '74849a00-85a9-11e1-9660-123139220831', 'ext': 'mp4', 'title': '#whilewewatch', } }, { # invalid labels, 360p is better that 480p 'url': 'http://www.snagfilms.com/embed/player?filmId=17ca0950-a74a-11e0-a92a-0026bb61d036', 'md5': '882fca19b9eb27ef865efeeaed376a48', 'info_dict': { 'id': '17ca0950-a74a-11e0-a92a-0026bb61d036', 'ext': 'mp4', 'title': 'Life in Limbo', } }, { 'url': 'http://www.snagfilms.com/embed/player?filmId=0000014c-de2f-d5d6-abcf-ffef58af0017', 'only_matching': True, }] @staticmethod def _extract_url(webpage): mobj = re.search( r']+?src=(["\'])(?P(?:https?:)?//(?:embed\.)?(?:%s)/embed/player.+?)\1' % ViewLiftBaseIE._DOMAINS_REGEX, webpage) if mobj: return mobj.group('url') def _real_extract(self, url): video_id = self._match_id(url) webpage = self._download_webpage(url, video_id) if '>This film is not playable in your area.<' in webpage: raise ExtractorError( 'Film %s is not playable in your area.' % video_id, expected=True) formats = [] has_bitrate = False sources = self._parse_json(self._search_regex( r'(?s)sources:\s*(\[.+?\]),', webpage, 'sources', default='[]'), video_id, js_to_json) for source in sources: file_ = source.get('file') if not file_: continue type_ = source.get('type') ext = determine_ext(file_) format_id = source.get('label') or ext if all(v in ('m3u8', 'hls') for v in (type_, ext)): formats.extend(self._extract_m3u8_formats( file_, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) else: bitrate = int_or_none(self._search_regex( [r'(\d+)kbps', r'_\d{1,2}x\d{1,2}_(\d{3,})\.%s' % ext], file_, 'bitrate', default=None)) if not has_bitrate and bitrate: has_bitrate = True height = int_or_none(self._search_regex( r'^(\d+)[pP]$', format_id, 'height', default=None)) formats.append({ 'url': file_, 'format_id': 'http-%s%s' % (format_id, ('-%dk' % bitrate if bitrate else '')), 'tbr': bitrate, 'height': height, }) if not formats: hls_url = self._parse_json(self._search_regex( r'filmInfo\.src\s*=\s*({.+?});', webpage, 'src'), video_id, js_to_json)['src'] formats = self._extract_m3u8_formats( hls_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False) field_preference = None if has_bitrate else ('height', 'tbr', 'format_id') self._sort_formats(formats, field_preference) title = self._search_regex( [r"title\s*:\s*'([^']+)'", r'([^<]+)'], webpage, 'title') return { 'id': video_id, 'title': title, 'formats': formats, } class ViewLiftIE(ViewLiftBaseIE): _VALID_URL = r'https?://(?:www\.)?(?P%s)/(?:films/title|show|(?:news/)?videos?)/(?P[^?#]+)' % ViewLiftBaseIE._DOMAINS_REGEX _TESTS = [{ 'url': 'http://www.snagfilms.com/films/title/lost_for_life', 'md5': '19844f897b35af219773fd63bdec2942', 'info_dict': { 'id': '0000014c-de2f-d5d6-abcf-ffef58af0017', 'display_id': 'lost_for_life', 'ext': 'mp4', 'title': 'Lost for Life', 'description': 'md5:ea10b5a50405ae1f7b5269a6ec594102', 'thumbnail': r're:^https?://.*\.jpg', 'duration': 4489, 'categories': 'mincount:3', 'age_limit': 14, 'upload_date': '20150421', 'timestamp': 1429656819, } }, { 'url': 'http://www.snagfilms.com/show/the_world_cut_project/india', 'md5': 'e6292e5b837642bbda82d7f8bf3fbdfd', 'info_dict': { 'id': '00000145-d75c-d96e-a9c7-ff5c67b20000', 'display_id': 'the_world_cut_project/india', 'ext': 'mp4', 'title': 'India', 'description': 'md5:5c168c5a8f4719c146aad2e0dfac6f5f', 'thumbnail': r're:^https?://.*\.jpg', 'duration': 979, 'categories': 'mincount:2', 'timestamp': 1399478279, 'upload_date': '20140507', } }, { # Film is not playable in your area. 'url': 'http://www.snagfilms.com/films/title/inside_mecca', 'only_matching': True, }, { # Film is not available. 'url': 'http://www.snagfilms.com/show/augie_alone/flirting', 'only_matching': True, }, { 'url': 'http://www.winnersview.com/videos/the-good-son', 'only_matching': True, }, { # Was once Kaltura embed 'url': 'https://www.monumentalsportsnetwork.com/videos/john-carlson-postgame-2-25-15', 'only_matching': True, }] def _real_extract(self, url): domain, display_id = re.match(self._VALID_URL, url).groups() webpage = self._download_webpage(url, display_id) if ">Sorry, the Film you're looking for is not available.<" in webpage: raise ExtractorError( 'Film %s is not available.' % display_id, expected=True) initial_store_state = self._search_regex( r"window\.initialStoreState\s*=.*?JSON\.parse\(unescape\(atob\('([^']+)'\)\)\)", webpage, 'Initial Store State', default=None) if initial_store_state: modules = self._parse_json(compat_urllib_parse_unquote(base64.b64decode( initial_store_state).decode()), display_id)['page']['data']['modules'] content_data = next(m['contentData'][0] for m in modules if m.get('moduleType') == 'VideoDetailModule') gist = content_data['gist'] film_id = gist['id'] title = gist['title'] video_assets = content_data['streamingInfo']['videoAssets'] formats = [] mpeg_video_assets = video_assets.get('mpeg') or [] for video_asset in mpeg_video_assets: video_asset_url = video_asset.get('url') if not video_asset: continue bitrate = int_or_none(video_asset.get('bitrate')) height = int_or_none(self._search_regex( r'^_?(\d+)[pP]$', video_asset.get('renditionValue'), 'height', default=None)) formats.append({ 'url': video_asset_url, 'format_id': 'http%s' % ('-%d' % bitrate if bitrate else ''), 'tbr': bitrate, 'height': height, 'vcodec': video_asset.get('codec'), }) hls_url = video_assets.get('hls') if hls_url: formats.extend(self._extract_m3u8_formats( hls_url, film_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)) self._sort_formats(formats, ('height', 'tbr', 'format_id')) info = { 'id': film_id, 'display_id': display_id, 'title': title, 'description': gist.get('description'), 'thumbnail': gist.get('videoImageUrl'), 'duration': int_or_none(gist.get('runtime')), 'age_limit': parse_age_limit(content_data.get('parentalRating', '').replace('_', '-')), 'timestamp': int_or_none(gist.get('publishDate'), 1000), 'formats': formats, } for k in ('categories', 'tags'): info[k] = [v['title'] for v in content_data.get(k, []) if v.get('title')] return info else: film_id = self._search_regex(r'filmId=([\da-f-]{36})"', webpage, 'film id') snag = self._parse_json( self._search_regex( r'Snag\.page\.data\s*=\s*(\[.+?\]);', webpage, 'snag', default='[]'), display_id) for item in snag: if item.get('data', {}).get('film', {}).get('id') == film_id: data = item['data']['film'] title = data['title'] description = clean_html(data.get('synopsis')) thumbnail = data.get('image') duration = int_or_none(data.get('duration') or data.get('runtime')) categories = [ category['title'] for category in data.get('categories', []) if category.get('title')] break else: title = self._search_regex( r'itemprop="title">([^<]+)<', webpage, 'title') description = self._html_search_regex( r'(?s)
(.+?)
', webpage, 'description', default=None) or self._og_search_description(webpage) thumbnail = self._og_search_thumbnail(webpage) duration = parse_duration(self._search_regex( r'([^<]+)<', webpage, 'duration', fatal=False)) categories = re.findall(r'([^<]+)', webpage) return { '_type': 'url_transparent', 'url': 'http://%s/embed/player?filmId=%s' % (domain, film_id), 'id': film_id, 'display_id': display_id, 'title': title, 'description': description, 'thumbnail': thumbnail, 'duration': duration, 'categories': categories, 'ie_key': 'ViewLiftEmbed', }