Merge tag 'upstream/2014.01.17.2'

[youtubedl] / youtube_dl / extractor / vevo.py

diff --git a/youtube_dl/extractor/vevo.py b/youtube_dl/extractor/vevo.py
index aa88e1a92e587a9d68075ac6a5196e35ce7857af..a4b26a26f4132840c57700fad96785dfb390a8db 100644 (file)
--- a/youtube_dl/extractor/vevo.py
+++ b/youtube_dl/extractor/vevo.py
@@ -1,44 +1,133 @@
 import re
 import json
 import re
 import json

+import xml.etree.ElementTree
+import datetime

 
 from .common import InfoExtractor
 from ..utils import (
 
 from .common import InfoExtractor
 from ..utils import (

-    unified_strdate,
+    compat_HTTPError,

     ExtractorError,
 )
 
     ExtractorError,
 )
 

+

 class VevoIE(InfoExtractor):
     """
 class VevoIE(InfoExtractor):
     """

-    Accecps urls from vevo.com or in the format 'vevo:{id}'
+    Accepts urls from vevo.com or in the format 'vevo:{id}'

     (currently used by MTVIE)
     """
     (currently used by MTVIE)
     """

-    _VALID_URL = r'((http://www.vevo.com/watch/.*?/.*?/)|(vevo:))(?P<id>.*)$'
+    _VALID_URL = r'''(?x)
+        (?:https?://www\.vevo\.com/watch/(?:[^/]+/[^/]+/)?|
+           https?://cache\.vevo\.com/m/html/embed\.html\?video=|
+           https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
+           vevo:)
+        (?P<id>[^&?#]+)'''
+    _TESTS = [{
+        u'url': u'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
+        u'file': u'GB1101300280.mp4',
+        u"md5": u"06bea460acb744eab74a9d7dcb4bfd61",
+        u'info_dict': {
+            u"upload_date": u"20130624",
+            u"uploader": u"Hurts",
+            u"title": u"Somebody to Die For",
+            u"duration": 230.12,
+            u"width": 1920,
+            u"height": 1080,
+        }
+    }]
+    _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
+
+    def _formats_from_json(self, video_info):
+        last_version = {'version': -1}
+        for version in video_info['videoVersions']:
+            # These are the HTTP downloads, other types are for different manifests
+            if version['sourceType'] == 2:
+                if version['version'] > last_version['version']:
+                    last_version = version
+        if last_version['version'] == -1:
+            raise ExtractorError(u'Unable to extract last version of the video')
+
+        renditions = xml.etree.ElementTree.fromstring(last_version['data'])
+        formats = []
+        # Already sorted from worst to best quality
+        for rend in renditions.findall('rendition'):
+            attr = rend.attrib
+            format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
+            formats.append({
+                'url': attr['url'],
+                'format_id': attr['name'],
+                'format_note': format_note,
+                'height': int(attr['frameheight']),
+                'width': int(attr['frameWidth']),
+            })
+        return formats
+
+    def _formats_from_smil(self, smil_xml):
+        formats = []
+        smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
+        els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
+        for el in els:
+            src = el.attrib['src']
+            m = re.match(r'''(?xi)
+                (?P<ext>[a-z0-9]+):
+                (?P<path>
+                    [/a-z0-9]+     # The directory and main part of the URL
+                    _(?P<cbr>[0-9]+)k
+                    _(?P<width>[0-9]+)x(?P<height>[0-9]+)
+                    _(?P<vcodec>[a-z0-9]+)
+                    _(?P<vbr>[0-9]+)
+                    _(?P<acodec>[a-z0-9]+)
+                    _(?P<abr>[0-9]+)
+                    \.[a-z0-9]+  # File extension
+                )''', src)
+            if not m:
+                continue
+
+            format_url = self._SMIL_BASE_URL + m.group('path')
+            formats.append({
+                'url': format_url,
+                'format_id': u'SMIL_' + m.group('cbr'),
+                'vcodec': m.group('vcodec'),
+                'acodec': m.group('acodec'),
+                'vbr': int(m.group('vbr')),
+                'abr': int(m.group('abr')),
+                'ext': m.group('ext'),
+                'width': int(m.group('width')),
+                'height': int(m.group('height')),
+            })
+        return formats

 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
 
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
 

-        json_url = 'http://www.vevo.com/data/video/%s' % video_id
-        base_url = 'http://smil.lvl3.vevo.com'
-        videos_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (base_url, video_id, video_id.lower())
+        json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id

         info_json = self._download_webpage(json_url, video_id, u'Downloading json info')
         info_json = self._download_webpage(json_url, video_id, u'Downloading json info')

-        links_webpage = self._download_webpage(videos_url, video_id, u'Downloading videos urls')
-
-        self.report_extraction(video_id)
-        video_info = json.loads(info_json)
-        m_urls = list(re.finditer(r'<video src="(?P<ext>.*?):(?P<url>.*?)"', links_webpage))
-        if m_urls is None or len(m_urls) == 0:
-            raise ExtractorError(u'Unable to extract video url')
-        # They are sorted from worst to best quality
-        m_url = m_urls[-1]
-        video_url = base_url + m_url.group('url')
-        ext = m_url.group('ext')
-
-        return {'url': video_url,
-                'ext': ext,
-                'id': video_id,
-                'title': video_info['title'],
-                'thumbnail': video_info['img'],
-                'upload_date': video_info['launchDate'].replace('/',''),
-                'uploader': video_info['Artists'][0]['title'],
-                }
+        video_info = json.loads(info_json)['video']
+
+        formats = self._formats_from_json(video_info)
+        try:
+            smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
+                self._SMIL_BASE_URL, video_id, video_id.lower())
+            smil_xml = self._download_webpage(smil_url, video_id,
+                                              u'Downloading SMIL info')
+            formats.extend(self._formats_from_smil(smil_xml))
+        except ExtractorError as ee:
+            if not isinstance(ee.cause, compat_HTTPError):
+                raise
+            self._downloader.report_warning(
+                u'Cannot download SMIL information, falling back to JSON ..')
+
+        timestamp_ms = int(self._search_regex(
+            r'/Date\((\d+)\)/', video_info['launchDate'], u'launch date'))
+        upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
+        info = {
+            'id': video_id,
+            'title': video_info['title'],
+            'formats': formats,
+            'thumbnail': video_info['imageUrl'],
+            'upload_date': upload_date.strftime('%Y%m%d'),
+            'uploader': video_info['mainArtists'][0]['artistName'],
+            'duration': video_info['duration'],
+        }
+
+        return info