X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/63a6927374492ef47c8fd6de67d0760ace4dd0ed..22bdf6fb907cce994fbfe062879d2a991e597a3f:/youtube_dl/extractor/yahoo.py diff --git a/youtube_dl/extractor/yahoo.py b/youtube_dl/extractor/yahoo.py index 5c9c361..d84be25 100644 --- a/youtube_dl/extractor/yahoo.py +++ b/youtube_dl/extractor/yahoo.py @@ -1,3 +1,5 @@ +from __future__ import unicode_literals + import itertools import json import re @@ -6,33 +8,45 @@ from .common import InfoExtractor, SearchInfoExtractor from ..utils import ( compat_urllib_parse, compat_urlparse, - determine_ext, clean_html, + int_or_none, ) 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'https?://(?:screen|movies)\.yahoo\.com/.*?-(?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', }, }, { - 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': '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': 'https://movies.yahoo.com/video/world-loves-spider-man-190819223.html', + 'md5': '410b7104aa9893b765bc22787a22f3d9', + 'info_dict': { + 'id': '516ed8e2-2c4f-339f-a211-7a8b49d30845', + 'ext': 'mp4', + 'title': 'The World Loves Spider-Man', + 'description': '''People all over the world are celebrating the release of \"The Amazing Spider-Man 2.\" We're taking a look at the enthusiastic response Spider-Man has received from viewers all over the world.''', + } + } ] def _real_extract(self, url): @@ -40,16 +54,25 @@ class YahooIE(InfoExtractor): video_id = mobj.group('id') webpage = self._download_webpage(url, video_id) - items_json = self._search_regex(r'mediaItems: ({.*?})$', - webpage, u'items', flags=re.MULTILINE) - items = json.loads(items_json) - info = items['mediaItems']['query']['results']['mediaObj'][0] - # The 'meta' field is not always in the video webpage, we request it - # from another page - long_id = info['id'] - return self._get_info(long_id, video_id) - - def _get_info(self, long_id, video_id): + items_json = self._search_regex( + r'mediaItems: ({.*?})$', 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 = \{"([^"]+)"' + ] + long_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID') + video_id = long_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 + long_id = info['id'] + return self._get_info(long_id, video_id, webpage) + + def _get_info(self, long_id, video_id, webpage): query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"' ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="US"' ' AND protocol="http"' % long_id) @@ -58,19 +81,18 @@ class YahooIE(InfoExtractor): '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) + video_id, 'Downloading video info') info = query_result['query']['results']['mediaObj'][0] meta = info['meta'] formats = [] for s in info['streams']: format_info = { - 'width': s.get('width'), - 'height': s.get('height'), - 'bitrate': s.get('bitrate'), + 'width': int_or_none(s.get('width')), + 'height': int_or_none(s.get('height')), + 'tbr': int_or_none(s.get('bitrate')), } host = s['host'] @@ -84,17 +106,16 @@ class YahooIE(InfoExtractor): else: format_url = compat_urlparse.urljoin(host, path) format_info['url'] = format_url - format_info['ext'] = determine_ext(format_url) - formats.append(format_info) - formats = sorted(formats, key=lambda f:(f['height'], f['width'])) + + self._sort_formats(formats) return { 'id': video_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), } @@ -102,57 +123,52 @@ 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', + _TESTS = [{ + 'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html', + 'md5': '67010fdf3a08d290e060a4dd96baa07b', + 'info_dict': { + 'id': '104538833', + 'ext': 'mp4', + 'title': 'China Moses Is Crazy About the Blues', + 'description': '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) + long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, 'long id') + return self._get_info(long_id, video_id, webpage) 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, + }