]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/yahoo.py
Initiate new release.
[youtubedl] / youtube_dl / extractor / yahoo.py
index 5c9c361b9ee5658d307a7759040b855a3e794cf1..97dbac4cce53d7fe956b074fddbe40993fd5681f 100644 (file)
+# coding: utf-8
+from __future__ import unicode_literals
+
 import itertools
 import json
 import re
 
 from .common import InfoExtractor, SearchInfoExtractor
 import itertools
 import json
 import re
 
 from .common import InfoExtractor, SearchInfoExtractor
-from ..utils import (
+from ..compat import (
     compat_urllib_parse,
     compat_urlparse,
     compat_urllib_parse,
     compat_urlparse,
-    determine_ext,
+)
+from ..utils import (
     clean_html,
     clean_html,
+    unescapeHTML,
+    ExtractorError,
+    int_or_none,
 )
 
 
 class YahooIE(InfoExtractor):
 )
 
 
 class YahooIE(InfoExtractor):
-    IE_DESC = u'Yahoo screen'
-    _VALID_URL = r'http://screen\.yahoo\.com/.*?-(?P<id>\d*?)\.html'
+    IE_DESC = 'Yahoo screen and movies'
+    _VALID_URL = r'(?P<url>(?P<host>https?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P<display_id>.+?)-(?P<id>[0-9]+)(?:-[a-z]+)?\.html)'
     _TESTS = [
         {
     _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',
+            '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/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': 'cac903b3-fcf4-3c14-b632-643ab541712f',
+                'ext': 'mp4',
+                '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,
+            }
+        },
+        {
+            '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/hackers-sony-more-trouble-well-154609075.html',
+            'md5': '226a895aae7e21b0129e2a2006fe9690',
+            'info_dict': {
+                'id': 'e624c4bc-3389-34de-9dfc-025f74943409',
+                'ext': 'mp4',
+                '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',
+            '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://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,
+        }
     ]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
     ]
 
     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):
-        query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
-                 ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="US"'
-                 ' AND protocol="http"' % long_id)
+        display_id = mobj.group('display_id')
+        page_id = mobj.group('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'<iframe[^>]+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*"([^"]+)"',
+                r'%s[^}]*"ccm_id"\s*:\s*"([^"]+)"' % re.escape(page_id),
+            ]
+            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')
         data = compat_urllib_parse.urlencode({
         data = compat_urllib_parse.urlencode({
-            'q': query,
-            'env': 'prod',
-            'format': 'json',
+            'protocol': 'http',
+            'region': region,
         })
         })
-        query_result_json = self._download_webpage(
-            'http://video.query.yahoo.com/v1/public/yql?' + data,
-            video_id, u'Downloading video info')
-        query_result = json.loads(query_result_json)
+        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')
+
         info = query_result['query']['results']['mediaObj'][0]
         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']:
             format_info = {
 
         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']
             }
 
             host = s['host']
@@ -84,75 +214,48 @@ class YahooIE(InfoExtractor):
             else:
                 format_url = compat_urlparse.urljoin(host, path)
                 format_info['url'] = format_url
             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.append(format_info)
-        formats = sorted(formats, key=lambda f:(f['height'], f['width']))
+
+        self._sort_formats(formats)
 
         return {
             'id': video_id,
 
         return {
             'id': video_id,
-            'title': meta['title'],
+            'display_id': display_id,
+            'title': unescapeHTML(meta['title']),
             'formats': formats,
             'description': clean_html(meta['description']),
             '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<id>\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):
 class YahooSearchIE(SearchInfoExtractor):
-    IE_DESC = u'Yahoo screen search'
+    IE_DESC = 'Yahoo screen search'
     _MAX_RESULTS = 1000
     _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"""
     _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):
 
             for (i, r) in enumerate(results):
-                if (pagenum * 30) +i >= n:
+                if (pagenum * 30) + i >= n:
                     break
                 mobj = re.search(r'(?P<url>screen\.yahoo\.com/.*?-\d*?\.html)"', r)
                 e = self.url_result('http://' + mobj.group('url'), 'Yahoo')
                     break
                 mobj = re.search(r'(?P<url>screen\.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
 
                 break
 
-        return res
+        return {
+            '_type': 'playlist',
+            'id': query,
+            'entries': entries,
+        }