X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/e8cd8c4bd832446f1971215b9fedc4531555dc1a..15b1d10671b48df598afd70e17ba21e9e64ac766:/youtube_dl/extractor/yahoo.py diff --git a/youtube_dl/extractor/yahoo.py b/youtube_dl/extractor/yahoo.py index e17a397..0fdb122 100644 --- a/youtube_dl/extractor/yahoo.py +++ b/youtube_dl/extractor/yahoo.py @@ -1,9 +1,13 @@ +# coding: utf-8 +from __future__ import unicode_literals + import itertools import json import re from .common import InfoExtractor, SearchInfoExtractor from ..utils import ( + ExtractorError, compat_urllib_parse, compat_urlparse, clean_html, @@ -12,58 +16,172 @@ from ..utils import ( class YahooIE(InfoExtractor): - IE_DESC = u'Yahoo screen' - _VALID_URL = r'http://screen\.yahoo\.com/.*?-(?P\d*?)\.html' + 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)' _TESTS = [ { - u'url': u'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html', - u'file': u'214727115.mp4', - u'md5': u'4962b075c08be8690a922ee026d05e69', - u'info_dict': { - u'title': u'Julian Smith & Travis Legg Watch Julian Smith', - u'description': u'Julian and Travis watch Julian Smith', + 'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html', + 'md5': '4962b075c08be8690a922ee026d05e69', + 'info_dict': { + 'id': '2d25e626-2378-391f-ada0-ddaf1417e588', + 'ext': 'mp4', + 'title': 'Julian Smith & Travis Legg Watch Julian Smith', + 'description': 'Julian and Travis watch Julian Smith', + 'duration': 6863, }, }, { - u'url': u'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html', - u'file': u'103000935.mp4', - u'md5': u'd6e6fc6e1313c608f316ddad7b82b306', - u'info_dict': { - u'title': u'Codefellas - The Cougar Lies with Spanish Moss', - u'description': u'Agent Topple\'s mustache does its dirty work, and Nicole brokers a deal for peace. But why is the NSA collecting millions of Instagram brunch photos? And if your waffles have nothing to hide, what are they so worried about?', + 'url': 'http://screen.yahoo.com/wired/codefellas-s1-ep12-cougar-lies-103000935.html', + 'md5': 'd6e6fc6e1313c608f316ddad7b82b306', + 'info_dict': { + 'id': 'd1dedf8c-d58c-38c3-8963-e899929ae0a9', + 'ext': 'mp4', + 'title': 'Codefellas - The Cougar Lies with Spanish Moss', + 'description': 'md5:66b627ab0a282b26352136ca96ce73c1', + 'duration': 151, }, }, + { + 'url': 'https://screen.yahoo.com/community/community-sizzle-reel-203225340.html?format=embed', + 'md5': '60e8ac193d8fb71997caa8fce54c6460', + 'info_dict': { + 'id': '4fe78544-8d48-39d8-97cd-13f205d9fcdb', + 'ext': 'mp4', + 'title': "Yahoo Saves 'Community'", + 'description': 'md5:4d4145af2fd3de00cbb6c1d664105053', + 'duration': 170, + } + }, + { + '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', + 'info_dict': { + 'id': '7a23b569-7bea-36cb-85b9-bd5301a0a1fb', + '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, + } + }, + { + 'url': 'https://uk.screen.yahoo.com/editor-picks/cute-raccoon-freed-drain-using-091756545.html', + 'md5': '0b51660361f0e27c9789e7037ef76f4b', + 'info_dict': { + 'id': 'b3affa53-2e14-3590-852b-0e0db6cd1a58', + 'ext': 'mp4', + 'title': 'Cute Raccoon Freed From Drain\u00a0Using Angle Grinder', + 'description': 'md5:f66c890e1490f4910a9953c941dee944', + 'duration': 97, + } + }, + { + 'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html', + 'md5': '57e06440778b1828a6079d2f744212c4', + 'info_dict': { + 'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73', + 'ext': 'mp4', + 'title': 'Program that makes hockey more affordable not offered in Manitoba', + 'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4', + 'duration': 121, + } + }, { + 'url': 'https://ca.finance.yahoo.com/news/20-most-valuable-brands-world-112600775.html', + 'md5': '3e401e4eed6325aa29d9b96125fd5b4f', + 'info_dict': { + 'id': 'c1b4c09c-8ed8-3b65-8b05-169c55358a83', + 'ext': 'mp4', + 'title': "Apple Is The World's Most Valuable Brand", + 'description': 'md5:73eabc1a11c6f59752593b2ceefa1262', + 'duration': 21, + } + }, { + 'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html', + 'md5': '67010fdf3a08d290e060a4dd96baa07b', + 'info_dict': { + 'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521', + 'ext': 'mp4', + 'title': 'China Moses Is Crazy About the Blues', + 'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0', + 'duration': 128, + } + }, { + 'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html', + 'md5': 'd9a083ccf1379127bf25699d67e4791b', + 'info_dict': { + 'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c', + 'ext': 'mp4', + 'title': 'Connect the Dots: Dark Side of Virgo', + 'description': 'md5:1428185051cfd1949807ad4ff6d3686a', + 'duration': 201, + } + }, { + 'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html', + 'only_matching': True, + } ] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - video_id = mobj.group('id') - webpage = self._download_webpage(url, video_id) - - items_json = self._search_regex(r'mediaItems: ({.*?})$', - webpage, u'items', flags=re.MULTILINE) - items = json.loads(items_json) - info = items['mediaItems']['query']['results']['mediaObj'][0] - # The 'meta' field is not always in the video webpage, we request it - # from another page - long_id = info['id'] - return self._get_info(long_id, video_id) - - def _get_info(self, long_id, video_id): + display_id = mobj.group('display_id') + url = mobj.group('url') + host = mobj.group('host') + webpage = self._download_webpage(url, display_id) + + # Look for iframed media first + iframe_m = re.search(r']+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage) + if iframe_m: + iframepage = self._download_webpage( + host + iframe_m.group(1), display_id, 'Downloading iframe webpage') + items_json = self._search_regex( + r'mediaItems: (\[.+?\])$', iframepage, 'items', flags=re.MULTILINE, default=None) + if items_json: + items = json.loads(items_json) + video_id = items[0]['id'] + return self._get_info(video_id, display_id, webpage) + + items_json = self._search_regex( + r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE, + default=None) + if items_json is None: + CONTENT_ID_REGEXES = [ + r'YUI\.namespace\("Media"\)\.CONTENT_ID\s*=\s*"([^"]+)"', + r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"', + r'"first_videoid"\s*:\s*"([^"]+)"', + ] + video_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID') + else: + items = json.loads(items_json) + info = items['mediaItems']['query']['results']['mediaObj'][0] + # The 'meta' field is not always in the video webpage, we request it + # from another page + 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="US"' - ' AND protocol="http"' % long_id) + ' 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_json = self._download_webpage( + query_result = self._download_json( 'http://video.query.yahoo.com/v1/public/yql?' + data, - video_id, u'Downloading video info') - query_result = json.loads(query_result_json) + display_id, 'Downloading video info') + info = query_result['query']['results']['mediaObj'][0] - meta = info['meta'] + meta = info.get('meta') + + if not meta: + msg = info['status'].get('msg') + if msg: + raise ExtractorError( + '%s returned error: %s' % (self.IE_NAME, msg), expected=True) + raise ExtractorError('Unable to extract media object meta') formats = [] for s in info['streams']: @@ -84,75 +202,48 @@ class YahooIE(InfoExtractor): else: format_url = compat_urlparse.urljoin(host, path) format_info['url'] = format_url - formats.append(format_info) self._sort_formats(formats) return { 'id': video_id, + 'display_id': display_id, 'title': meta['title'], 'formats': formats, 'description': clean_html(meta['description']), - 'thumbnail': meta['thumbnail'], + 'thumbnail': meta['thumbnail'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage), + 'duration': int_or_none(meta.get('duration')), } -class YahooNewsIE(YahooIE): - IE_NAME = 'yahoo:news' - _VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P\d*?)\.html' - - _TEST = { - u'url': u'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html', - u'md5': u'67010fdf3a08d290e060a4dd96baa07b', - u'info_dict': { - u'id': u'104538833', - u'ext': u'mp4', - u'title': u'China Moses Is Crazy About the Blues', - u'description': u'md5:9900ab8cd5808175c7b3fe55b979bed0', - }, - } - - # Overwrite YahooIE properties we don't want - _TESTS = [] - - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - video_id = mobj.group('id') - webpage = self._download_webpage(url, video_id) - long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, u'long id') - return self._get_info(long_id, video_id) - - class YahooSearchIE(SearchInfoExtractor): - IE_DESC = u'Yahoo screen search' + IE_DESC = 'Yahoo screen search' _MAX_RESULTS = 1000 - IE_NAME = u'screen.yahoo:search' + IE_NAME = 'screen.yahoo:search' _SEARCH_KEY = 'yvsearch' def _get_n_results(self, query, n): """Get a specified number of results for a query""" - - res = { - '_type': 'playlist', - 'id': query, - 'entries': [] - } - for pagenum in itertools.count(0): - result_url = u'http://video.search.yahoo.com/search/?p=%s&fr=screen&o=js&gs=0&b=%d' % (compat_urllib_parse.quote_plus(query), pagenum * 30) - webpage = self._download_webpage(result_url, query, - note='Downloading results page '+str(pagenum+1)) - info = json.loads(webpage) - m = info[u'm'] - results = info[u'results'] + entries = [] + 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)) + m = info['m'] + results = info['results'] for (i, r) in enumerate(results): - if (pagenum * 30) +i >= n: + if (pagenum * 30) + i >= n: break mobj = re.search(r'(?Pscreen\.yahoo\.com/.*?-\d*?\.html)"', r) e = self.url_result('http://' + mobj.group('url'), 'Yahoo') - res['entries'].append(e) - if (pagenum * 30 +i >= n) or (m[u'last'] >= (m[u'total'] -1)): + entries.append(e) + if (pagenum * 30 + i >= n) or (m['last'] >= (m['total'] - 1)): break - return res + return { + '_type': 'playlist', + 'id': query, + 'entries': entries, + }