Update changelog.

[youtubedl] / youtube_dl / extractor / vevo.py

diff --git a/youtube_dl/extractor/vevo.py b/youtube_dl/extractor/vevo.py
index e458ac961f701391c2fb651b1557ac7070dc6e2a..4c0de354f7d8ac467f0f027b201d785a38c77ffa 100644 (file)
--- a/youtube_dl/extractor/vevo.py
+++ b/youtube_dl/extractor/vevo.py
@@ -1,42 +1,92 @@
 from __future__ import unicode_literals
 
 import re
 from __future__ import unicode_literals
 
 import re

-import xml.etree.ElementTree
-import datetime

 
 from .common import InfoExtractor
 
 from .common import InfoExtractor

+from ..compat import (
+    compat_etree_fromstring,
+    compat_urllib_request,
+)

 from ..utils import (
 from ..utils import (

-    compat_HTTPError,

     ExtractorError,
     ExtractorError,

+    int_or_none,

 )
 
 
 class VevoIE(InfoExtractor):
     """
     Accepts urls from vevo.com or in the format 'vevo:{id}'
 )
 
 
 class VevoIE(InfoExtractor):
     """
     Accepts urls from vevo.com or in the format 'vevo:{id}'

-    (currently used by MTVIE)
+    (currently used by MTVIE and MySpaceIE)

     """
     _VALID_URL = r'''(?x)
     """
     _VALID_URL = r'''(?x)

-        (?:https?://www\.vevo\.com/watch/(?:[^/]+/[^/]+/)?|
+        (?:https?://www\.vevo\.com/watch/(?:[^/]+/(?:[^/]+/)?)?|

            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
            vevo:)
         (?P<id>[^&?#]+)'''
            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
            vevo:)
         (?P<id>[^&?#]+)'''

+

     _TESTS = [{
         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
     _TESTS = [{
         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',

-        'file': 'GB1101300280.mp4',
-        "md5": "06bea460acb744eab74a9d7dcb4bfd61",
+        "md5": "95ee28ee45e70130e3ab02b0f579ae23",

         'info_dict': {
         'info_dict': {

+            'id': 'GB1101300280',
+            'ext': 'mp4',

             "upload_date": "20130624",
             "uploader": "Hurts",
             "title": "Somebody to Die For",
             "duration": 230.12,
             "width": 1920,
             "height": 1080,
             "upload_date": "20130624",
             "uploader": "Hurts",
             "title": "Somebody to Die For",
             "duration": 230.12,
             "width": 1920,
             "height": 1080,

+            # timestamp and upload_date are often incorrect; seem to change randomly
+            'timestamp': int,
+        }
+    }, {
+        'note': 'v3 SMIL format',
+        'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',
+        'md5': 'f6ab09b034f8c22969020b042e5ac7fc',
+        'info_dict': {
+            'id': 'USUV71302923',
+            'ext': 'mp4',
+            'upload_date': '20140219',
+            'uploader': 'Cassadee Pope',
+            'title': 'I Wish I Could Break Your Heart',
+            'duration': 226.101,
+            'age_limit': 0,
+            'timestamp': int,
+        }
+    }, {
+        'note': 'Age-limited video',
+        'url': 'https://www.vevo.com/watch/justin-timberlake/tunnel-vision-explicit/USRV81300282',
+        'info_dict': {
+            'id': 'USRV81300282',
+            'ext': 'mp4',
+            'age_limit': 18,
+            'title': 'Tunnel Vision (Explicit)',
+            'uploader': 'Justin Timberlake',
+            'upload_date': 're:2013070[34]',
+            'timestamp': int,
+        },
+        'params': {
+            'skip_download': 'true',

         }
     }]
     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
 
         }
     }]
     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
 

+    def _real_initialize(self):
+        req = compat_urllib_request.Request(
+            'http://www.vevo.com/auth', data=b'')
+        webpage = self._download_webpage(
+            req, None,
+            note='Retrieving oauth token',
+            errnote='Unable to retrieve oauth token',
+            fatal=False)
+        if webpage is False:
+            self._oauth_token = None
+        else:
+            self._oauth_token = self._search_regex(
+                r'access_token":\s*"([^"]+)"',
+                webpage, 'access token', fatal=False)
+

     def _formats_from_json(self, video_info):
         last_version = {'version': -1}
         for version in video_info['videoVersions']:
     def _formats_from_json(self, video_info):
         last_version = {'version': -1}
         for version in video_info['videoVersions']:


@@ -47,7 +97,7 @@ class VevoIE(InfoExtractor):
         if last_version['version'] == -1:
             raise ExtractorError('Unable to extract last version of the video')
 
         if last_version['version'] == -1:
             raise ExtractorError('Unable to extract last version of the video')
 

-        renditions = xml.etree.ElementTree.fromstring(last_version['data'])
+        renditions = compat_etree_fromstring(last_version['data'])

         formats = []
         # Already sorted from worst to best quality
         for rend in renditions.findall('rendition'):
         formats = []
         # Already sorted from worst to best quality
         for rend in renditions.findall('rendition'):


@@ -64,7 +114,7 @@ class VevoIE(InfoExtractor):
 
     def _formats_from_smil(self, smil_xml):
         formats = []
 
     def _formats_from_smil(self, smil_xml):
         formats = []

-        smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
+        smil_doc = compat_etree_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']
         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
         for el in els:
             src = el.attrib['src']


@@ -97,35 +147,83 @@ class VevoIE(InfoExtractor):
             })
         return formats
 
             })
         return formats
 

+    def _download_api_formats(self, video_id):
+        if not self._oauth_token:
+            self._downloader.report_warning(
+                'No oauth token available, skipping API HLS download')
+            return []
+
+        api_url = 'https://apiv2.vevo.com/video/%s/streams/hls?token=%s' % (
+            video_id, self._oauth_token)
+        api_data = self._download_json(
+            api_url, video_id,
+            note='Downloading HLS formats',
+            errnote='Failed to download HLS format list', fatal=False)
+        if api_data is None:
+            return []
+
+        m3u8_url = api_data[0]['url']
+        return self._extract_m3u8_formats(
+            m3u8_url, video_id, entry_protocol='m3u8_native', ext='mp4',
+            preference=0)
+

     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
 
         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
 
         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id

-        video_info = self._download_json(json_url, video_id)['video']
+        response = self._download_json(json_url, video_id)
+        video_info = response['video']
+
+        if not video_info:
+            if 'statusMessage' in response:
+                raise ExtractorError('%s said: %s' % (self.IE_NAME, response['statusMessage']), expected=True)
+            raise ExtractorError('Unable to extract videos')

 
         formats = self._formats_from_json(video_info)
 
         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,
-                                              '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(
-                'Cannot download SMIL information, falling back to JSON ..')

 
 

-        timestamp_ms = int(self._search_regex(
-            r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
-        upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
+        is_explicit = video_info.get('isExplicit')
+        if is_explicit is True:
+            age_limit = 18
+        elif is_explicit is False:
+            age_limit = 0
+        else:
+            age_limit = None
+
+        # Download via HLS API
+        formats.extend(self._download_api_formats(video_id))
+
+        # Download SMIL
+        smil_blocks = sorted((
+            f for f in video_info['videoVersions']
+            if f['sourceType'] == 13),
+            key=lambda f: f['version'])
+        smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
+            self._SMIL_BASE_URL, video_id, video_id.lower())
+        if smil_blocks:
+            smil_url_m = self._search_regex(
+                r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
+                default=None)
+            if smil_url_m is not None:
+                smil_url = smil_url_m
+        if smil_url:
+            smil_xml = self._download_webpage(
+                smil_url, video_id, 'Downloading SMIL info', fatal=False)
+            if smil_xml:
+                formats.extend(self._formats_from_smil(smil_xml))
+
+        self._sort_formats(formats)
+        timestamp_ms = int_or_none(self._search_regex(
+            r'/Date\((\d+)\)/',
+            video_info['launchDate'], 'launch date', fatal=False))
+

         return {
             'id': video_id,
             'title': video_info['title'],
             'formats': formats,
             'thumbnail': video_info['imageUrl'],
         return {
             'id': video_id,
             'title': video_info['title'],
             'formats': formats,
             'thumbnail': video_info['imageUrl'],

-            'upload_date': upload_date.strftime('%Y%m%d'),
+            'timestamp': timestamp_ms // 1000,

             'uploader': video_info['mainArtists'][0]['artistName'],
             'duration': video_info['duration'],
             'uploader': video_info['mainArtists'][0]['artistName'],
             'duration': video_info['duration'],

+            'age_limit': age_limit,

         }
         }