X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/46113edab215c2211a604c06245c16d5d4e57dcf..d018d3313032e12968a6add6800e51d412e2f602:/youtube_dl/extractor/yahoo.py diff --git a/youtube_dl/extractor/yahoo.py b/youtube_dl/extractor/yahoo.py index 117f085..fca5ddc 100644 --- a/youtube_dl/extractor/yahoo.py +++ b/youtube_dl/extractor/yahoo.py @@ -6,22 +6,27 @@ import json import re from .common import InfoExtractor, SearchInfoExtractor -from ..utils import ( - ExtractorError, +from ..compat import ( compat_urllib_parse, compat_urlparse, +) +from ..utils import ( clean_html, + unescapeHTML, + ExtractorError, int_or_none, + mimetype2ext, ) +from .nbc import NBCSportsVPlayerIE + class YahooIE(InfoExtractor): IE_DESC = 'Yahoo screen and movies' - _VALID_URL = r'(?P(?Phttps?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P.+?)-(?P[0-9]+)(?:-[a-z]+)?\.html)' + _VALID_URL = r'(?P(?Phttps?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P.+)?-(?P[0-9]+)(?:-[a-z]+)?\.html)' _TESTS = [ { 'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html', - 'md5': '4962b075c08be8690a922ee026d05e69', 'info_dict': { 'id': '2d25e626-2378-391f-ada0-ddaf1417e588', 'ext': 'mp4', @@ -53,14 +58,14 @@ class YahooIE(InfoExtractor): } }, { - 'url': 'https://tw.screen.yahoo.com/taipei-opinion-poll/選情站報-街頭民調-台北市篇-102823042.html', - 'md5': '92a7fdd8a08783c68a174d7aa067dde8', + 'url': 'https://tw.screen.yahoo.com/election-2014-askmayor/敢問市長-黃秀霜批賴清德-非常高傲-033009720.html', + 'md5': '3a09cf59349cfaddae1797acc3c087fc', 'info_dict': { - 'id': '7a23b569-7bea-36cb-85b9-bd5301a0a1fb', + 'id': 'cac903b3-fcf4-3c14-b632-643ab541712f', 'ext': 'mp4', - 'title': '選情站報 街頭民調 台北市篇', - 'description': '選情站報 街頭民調 台北市篇', - 'duration': 429, + 'title': '敢問市長／黃秀霜批賴清德「非常高傲」', + 'description': '直言台南沒捷運 交通居五都之末', + 'duration': 396, } }, { @@ -85,18 +90,18 @@ class YahooIE(InfoExtractor): 'duration': 121, } }, { - 'url': 'https://ca.finance.yahoo.com/news/20-most-valuable-brands-world-112600775.html', - 'md5': '3e401e4eed6325aa29d9b96125fd5b4f', + 'url': 'https://ca.finance.yahoo.com/news/hackers-sony-more-trouble-well-154609075.html', + 'md5': '226a895aae7e21b0129e2a2006fe9690', 'info_dict': { - 'id': 'c1b4c09c-8ed8-3b65-8b05-169c55358a83', + 'id': 'e624c4bc-3389-34de-9dfc-025f74943409', 'ext': 'mp4', - 'title': "Apple Is The World's Most Valuable Brand", - 'description': 'md5:73eabc1a11c6f59752593b2ceefa1262', - 'duration': 21, + 'title': '\'The Interview\' TV Spot: War', + 'description': 'The Interview', + 'duration': 30, } }, { 'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html', - 'md5': '67010fdf3a08d290e060a4dd96baa07b', + 'md5': '88e209b417f173d86186bef6e4d1f160', 'info_dict': { 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521', 'ext': 'mp4', @@ -114,15 +119,49 @@ class YahooIE(InfoExtractor): 'description': 'md5:1428185051cfd1949807ad4ff6d3686a', 'duration': 201, } + }, { + 'url': 'https://www.yahoo.com/movies/v/true-story-trailer-173000497.html', + 'md5': '989396ae73d20c6f057746fb226aa215', + 'info_dict': { + 'id': '071c4013-ce30-3a93-a5b2-e0413cd4a9d1', + 'ext': 'mp4', + 'title': '\'True Story\' Trailer', + 'description': 'True Story', + 'duration': 150, + }, }, { 'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html', 'only_matching': True, + }, { + 'note': 'NBC Sports embeds', + 'url': 'http://sports.yahoo.com/blogs/ncaab-the-dagger/tyler-kalinoski-s-buzzer-beater-caps-davidson-s-comeback-win-185609842.html?guid=nbc_cbk_davidsonbuzzerbeater_150313', + 'info_dict': { + 'id': '9CsDKds0kvHI', + 'ext': 'flv', + 'description': 'md5:df390f70a9ba7c95ff1daace988f0d8d', + 'title': 'Tyler Kalinoski hits buzzer-beater to lift Davidson', + } + }, { + 'url': 'https://tw.news.yahoo.com/-100120367.html', + 'only_matching': True, + }, { + # Query result is embedded in webpage, but explicit request to video API fails with geo restriction + 'url': 'https://screen.yahoo.com/community/communitary-community-episode-1-ladders-154501237.html', + 'md5': '4fbafb9c9b6f07aa8f870629f6671b35', + 'info_dict': { + 'id': '1f32853c-a271-3eef-8cb6-f6d6872cb504', + 'ext': 'mp4', + 'title': 'Communitary - Community Episode 1: Ladders', + 'description': 'md5:8fc39608213295748e1e289807838c97', + 'duration': 1646, + }, } ] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - display_id = mobj.group('display_id') + display_id = mobj.group('display_id') or self._match_id(url) + page_id = mobj.group('id') url = mobj.group('url') host = mobj.group('host') webpage = self._download_webpage(url, display_id) @@ -138,6 +177,23 @@ class YahooIE(InfoExtractor): items = json.loads(items_json) video_id = items[0]['id'] return self._get_info(video_id, display_id, webpage) + # Look for NBCSports iframes + nbc_sports_url = NBCSportsVPlayerIE._extract_url(webpage) + if nbc_sports_url: + return self.url_result(nbc_sports_url, 'NBCSportsVPlayer') + + # Query result is often embedded in webpage as JSON. Sometimes explicit requests + # to video API results in a failure with geo restriction reason therefore using + # embedded query result when present sounds reasonable. + config_json = self._search_regex( + r'window\.Af\.bootstrap\[[^\]]+\]\s*=\s*({.*?"applet_type"\s*:\s*"td-applet-videoplayer".*?});(?:|$)', + webpage, 'videoplayer applet', default=None) + if config_json: + config = self._parse_json(config_json, display_id, fatal=False) + if config: + sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi') + if sapi: + return self._extract_info(display_id, sapi, webpage) items_json = self._search_regex( r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE, @@ -147,6 +203,7 @@ class YahooIE(InfoExtractor): r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"', r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"', r'"first_videoid"\s*:\s*"([^"]+)"', + r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id), ] video_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID') else: @@ -157,24 +214,10 @@ class YahooIE(InfoExtractor): video_id = info['id'] return self._get_info(video_id, display_id, webpage) - def _get_info(self, video_id, display_id, webpage): - region = self._search_regex( - r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"', - webpage, 'region', fatal=False, default='US') - query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"' - ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="%s"' - ' AND protocol="http"' % (video_id, region)) - data = compat_urllib_parse.urlencode({ - 'q': query, - 'env': 'prod', - 'format': 'json', - }) - query_result = self._download_json( - 'http://video.query.yahoo.com/v1/public/yql?' + data, - display_id, 'Downloading video info') - - info = query_result['query']['results']['mediaObj'][0] + def _extract_info(self, display_id, query, webpage): + info = query['query']['results']['mediaObj'][0] meta = info.get('meta') + video_id = info.get('id') if not meta: msg = info['status'].get('msg') @@ -200,22 +243,57 @@ class YahooIE(InfoExtractor): 'ext': 'flv', }) else: + if s.get('format') == 'm3u8_playlist': + format_info['protocol'] = 'm3u8_native' + format_info['ext'] = 'mp4' format_url = compat_urlparse.urljoin(host, path) format_info['url'] = format_url formats.append(format_info) self._sort_formats(formats) + closed_captions = self._html_search_regex( + r'"closedcaptions":(\[[^\]]+\])', webpage, 'closed captions', + default='[]') + + cc_json = self._parse_json(closed_captions, video_id, fatal=False) + subtitles = {} + if cc_json: + for closed_caption in cc_json: + lang = closed_caption['lang'] + if lang not in subtitles: + subtitles[lang] = [] + subtitles[lang].append({ + 'url': closed_caption['url'], + 'ext': mimetype2ext(closed_caption['content_type']), + }) + return { 'id': video_id, 'display_id': display_id, - 'title': meta['title'], + 'title': unescapeHTML(meta['title']), 'formats': formats, 'description': clean_html(meta['description']), 'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage), 'duration': int_or_none(meta.get('duration')), + 'subtitles': subtitles, } + def _get_info(self, video_id, display_id, webpage): + region = self._search_regex( + r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"', + webpage, 'region', fatal=False, default='US') + data = compat_urllib_parse.urlencode({ + 'protocol': 'http', + 'region': region, + }) + query_url = ( + 'https://video.media.yql.yahoo.com/v1/video/sapi/streams/' + '{id}?{data}'.format(id=video_id, data=data)) + query_result = self._download_json( + query_url, display_id, 'Downloading video info') + return self._extract_info(display_id, query_result, webpage) + class YahooSearchIE(SearchInfoExtractor): IE_DESC = 'Yahoo screen search' @@ -229,7 +307,7 @@ class YahooSearchIE(SearchInfoExtractor): for pagenum in itertools.count(0): result_url = 'http://video.search.yahoo.com/search/?p=%s&fr=screen&o=js&gs=0&b=%d' % (compat_urllib_parse.quote_plus(query), pagenum * 30) info = self._download_json(result_url, query, - note='Downloading results page '+str(pagenum+1)) + note='Downloading results page ' + str(pagenum + 1)) m = info['m'] results = info['results']