Merge tag 'upstream/2014.12.01'

[youtubedl] / youtube_dl / extractor / yahoo.py

diff --git a/youtube_dl/extractor/yahoo.py b/youtube_dl/extractor/yahoo.py
index 4b3aec9d1f5a8ce91f676d585e5f98dd3ba126aa..0fdb122436d9b7e158a706cd1048630e252d5666 100644 (file)
--- a/youtube_dl/extractor/yahoo.py
+++ b/youtube_dl/extractor/yahoo.py
@@ -1,113 +1,249 @@
-import datetime
+# coding: utf-8
+from __future__ import unicode_literals
+

 import itertools
 import json
 import re
 
 from .common import InfoExtractor, SearchInfoExtractor
 from ..utils import (
 import itertools
 import json
 import re
 
 from .common import InfoExtractor, SearchInfoExtractor
 from ..utils import (

-    compat_urllib_parse,
-

     ExtractorError,
     ExtractorError,

+    compat_urllib_parse,
+    compat_urlparse,
+    clean_html,
+    int_or_none,

 )
 
 )
 

+

 class YahooIE(InfoExtractor):
 class YahooIE(InfoExtractor):

-    """Information extractor for screen.yahoo.com."""
-    _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 = [
+        {
+            '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,
+            },
+        },
+        {
+            '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)
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)

-        if mobj is None:
-            raise ExtractorError(u'Invalid URL: %s' % url)
-        video_id = mobj.group('id')
-        webpage = self._download_webpage(url, video_id)
-        m_id = re.search(r'YUI\.namespace\("Media"\)\.CONTENT_ID = "(?P<new_id>.+?)";', webpage)
-
-        if m_id is None: 
-            # TODO: Check which url parameters are required
-            info_url = 'http://cosmos.bcst.yahoo.com/rest/v2/pops;lmsoverride=1;outputformat=mrss;cb=974419660;id=%s;rd=news.yahoo.com;datacontext=mdb;lg=KCa2IihxG3qE60vQ7HtyUy' % video_id
-            webpage = self._download_webpage(info_url, video_id, u'Downloading info webpage')
-            info_re = r'''<title><!\[CDATA\[(?P<title>.*?)\]\]></title>.*
-                        <description><!\[CDATA\[(?P<description>.*?)\]\]></description>.*
-                        <media:pubStart><!\[CDATA\[(?P<date>.*?)\ .*\]\]></media:pubStart>.*
-                        <media:content\ medium="image"\ url="(?P<thumb>.*?)"\ name="LARGETHUMB"
-                        '''
-            self.report_extraction(video_id)
-            m_info = re.search(info_re, webpage, re.VERBOSE|re.DOTALL)
-            if m_info is None:
-                raise ExtractorError(u'Unable to extract video info')
-            video_title = m_info.group('title')
-            video_description = m_info.group('description')
-            video_thumb = m_info.group('thumb')
-            video_date = m_info.group('date')
-            video_date = datetime.datetime.strptime(video_date, '%m/%d/%Y').strftime('%Y%m%d')
-    
-            # TODO: Find a way to get mp4 videos
-            rest_url = 'http://cosmos.bcst.yahoo.com/rest/v2/pops;element=stream;outputformat=mrss;id=%s;lmsoverride=1;bw=375;dynamicstream=1;cb=83521105;tech=flv,mp4;rd=news.yahoo.com;datacontext=mdb;lg=KCa2IihxG3qE60vQ7HtyUy' % video_id
-            webpage = self._download_webpage(rest_url, video_id, u'Downloading video url webpage')
-            m_rest = re.search(r'<media:content url="(?P<url>.*?)" path="(?P<path>.*?)"', webpage)
-            video_url = m_rest.group('url')
-            video_path = m_rest.group('path')
-            if m_rest is None:
-                raise ExtractorError(u'Unable to extract video url')
-
-        else: # We have to use a different method if another id is defined
-            long_id = m_id.group('new_id')
-            info_url = 'http://video.query.yahoo.com/v1/public/yql?q=SELECT%20*%20FROM%20yahoo.media.video.streams%20WHERE%20id%3D%22' + long_id + '%22%20AND%20format%3D%22mp4%2Cflv%22%20AND%20protocol%3D%22rtmp%2Chttp%22%20AND%20plrs%3D%2286Gj0vCaSzV_Iuf6hNylf2%22%20AND%20acctid%3D%22389%22%20AND%20plidl%3D%22%22%20AND%20pspid%3D%22792700001%22%20AND%20offnetwork%3D%22false%22%20AND%20site%3D%22ivy%22%20AND%20lang%3D%22en-US%22%20AND%20region%3D%22US%22%20AND%20override%3D%22none%22%3B&env=prod&format=json&callback=YUI.Env.JSONP.yui_3_8_1_1_1368368376830_335'
-            webpage = self._download_webpage(info_url, video_id, u'Downloading info json')
-            json_str = re.search(r'YUI.Env.JSONP.yui.*?\((.*?)\);', webpage).group(1)
-            info = json.loads(json_str)
-            res = info[u'query'][u'results'][u'mediaObj'][0]
-            stream = res[u'streams'][0]
-            video_path = stream[u'path']
-            video_url = stream[u'host']
-            meta = res[u'meta']
-            video_title = meta[u'title']
-            video_description = meta[u'description']
-            video_thumb = meta[u'thumbnail']
-            video_date = None # I can't find it
-
-        info_dict = {
-                     'id': video_id,
-                     'url': video_url,
-                     'play_path': video_path,
-                     'title':video_title,
-                     'description': video_description,
-                     'thumbnail': video_thumb,
-                     'upload_date': video_date,
-                     'ext': 'flv',
-                     }
-        return info_dict
+        display_id = mobj.group('display_id')
+        url = mobj.group('url')
+        host = mobj.group('host')
+        webpage = self._download_webpage(url, display_id)

 
 

-class YahooSearchIE(SearchInfoExtractor):
-    """Information Extractor for Yahoo! Video search queries."""
+        # 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*"([^"]+)"',
+            ]
+            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="%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]
+        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 = {
+                'width': int_or_none(s.get('width')),
+                'height': int_or_none(s.get('height')),
+                'tbr': int_or_none(s.get('bitrate')),
+            }
+
+            host = s['host']
+            path = s['path']
+            if host.startswith('rtmp'):
+                format_info.update({
+                    'url': host,
+                    'play_path': path,
+                    'ext': 'flv',
+                })
+            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'] if meta.get('thumbnail') else self._og_search_thumbnail(webpage),
+            'duration': int_or_none(meta.get('duration')),
+        }
+
+
+class YahooSearchIE(SearchInfoExtractor):
+    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,
+        }