]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/vevo.py
debian/copyright: Fix missing-license-paragraph-in-dep5-copyright.
[youtubedl] / youtube_dl / extractor / vevo.py
index 3f6020f74ec9eeefbddafc184d3f48cf5e436adb..c912c3cbe7ae42b221816b50ec6a97139cb13d55 100644 (file)
@@ -1,11 +1,11 @@
+from __future__ import unicode_literals
+
 import re
-import json
 import xml.etree.ElementTree
-import datetime
 
 from .common import InfoExtractor
 from ..utils import (
-    compat_HTTPError,
+    compat_urllib_request,
     ExtractorError,
 )
 
@@ -13,24 +13,77 @@ from ..utils import (
 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'((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,
-            u"width": 1920,
-            u"height": 1080,
+        'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
+        "md5": "95ee28ee45e70130e3ab02b0f579ae23",
+        'info_dict': {
+            'id': 'GB1101300280',
+            'ext': 'mp4',
+            "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/'
 
+    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']:
@@ -39,7 +92,7 @@ class VevoIE(InfoExtractor):
                 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')
+            raise ExtractorError('Unable to extract last version of the video')
 
         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
         formats = []
@@ -78,50 +131,76 @@ class VevoIE(InfoExtractor):
                 continue
 
             format_url = self._SMIL_BASE_URL + m.group('path')
-            format_note = ('%(vcodec)s@%(vbr)4sk, %(acodec)s@%(abr)3sk' %
-                           m.groupdict())
             formats.append({
                 'url': format_url,
-                'format_id': u'SMIL_' + m.group('cbr'),
-                'format_note': format_note,
+                'format_id': '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 _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
-        info_json = self._download_webpage(json_url, video_id, u'Downloading json info')
-        video_info = json.loads(info_json)['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)
-        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 ..')
 
+        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))
+
+        self._sort_formats(formats)
         timestamp_ms = int(self._search_regex(
-            r'/Date\((\d+)\)/', video_info['launchDate'], u'launch date'))
-        upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
-        info = {
+            r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
+
+        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'],
+            'age_limit': age_limit,
         }
-
-        return info