]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/ooyala.py
Merge tag 'upstream/2016.12.01'
[youtubedl] / youtube_dl / extractor / ooyala.py
index 2044e107eba9808bbde802e8468bf6b009841fb8..c2807d0f61b2ab5134944bd0c79b2030df80d3a1 100644 (file)
 from __future__ import unicode_literals
 import re
-import json
+import base64
 
 from .common import InfoExtractor
 from ..utils import (
-    unescapeHTML,
+    int_or_none,
+    float_or_none,
     ExtractorError,
+    unsmuggle_url,
+    determine_ext,
 )
+from ..compat import compat_urllib_parse_urlencode
 
 
-class OoyalaIE(InfoExtractor):
+class OoyalaBaseIE(InfoExtractor):
+    _PLAYER_BASE = 'http://player.ooyala.com/'
+    _CONTENT_TREE_BASE = _PLAYER_BASE + 'player_api/v1/content_tree/'
+    _AUTHORIZATION_URL_TEMPLATE = _PLAYER_BASE + 'sas/player_api/v2/authorization/embed_code/%s/%s?'
+
+    def _extract(self, content_tree_url, video_id, domain='example.org', supportedformats=None):
+        content_tree = self._download_json(content_tree_url, video_id)['content_tree']
+        metadata = content_tree[list(content_tree)[0]]
+        embed_code = metadata['embed_code']
+        pcode = metadata.get('asset_pcode') or embed_code
+        title = metadata['title']
+
+        auth_data = self._download_json(
+            self._AUTHORIZATION_URL_TEMPLATE % (pcode, embed_code) +
+            compat_urllib_parse_urlencode({
+                'domain': domain,
+                'supportedFormats': supportedformats or 'mp4,rtmp,m3u8,hds',
+            }), video_id)
+
+        cur_auth_data = auth_data['authorization_data'][embed_code]
+
+        urls = []
+        formats = []
+        if cur_auth_data['authorized']:
+            for stream in cur_auth_data['streams']:
+                s_url = base64.b64decode(
+                    stream['url']['data'].encode('ascii')).decode('utf-8')
+                if s_url in urls:
+                    continue
+                urls.append(s_url)
+                ext = determine_ext(s_url, None)
+                delivery_type = stream['delivery_type']
+                if delivery_type == 'hls' or ext == 'm3u8':
+                    formats.extend(self._extract_m3u8_formats(
+                        re.sub(r'/ip(?:ad|hone)/', '/all/', s_url), embed_code, 'mp4', 'm3u8_native',
+                        m3u8_id='hls', fatal=False))
+                elif delivery_type == 'hds' or ext == 'f4m':
+                    formats.extend(self._extract_f4m_formats(
+                        s_url + '?hdcore=3.7.0', embed_code, f4m_id='hds', fatal=False))
+                elif ext == 'smil':
+                    formats.extend(self._extract_smil_formats(
+                        s_url, embed_code, fatal=False))
+                else:
+                    formats.append({
+                        'url': s_url,
+                        'ext': ext or stream.get('delivery_type'),
+                        'vcodec': stream.get('video_codec'),
+                        'format_id': delivery_type,
+                        'width': int_or_none(stream.get('width')),
+                        'height': int_or_none(stream.get('height')),
+                        'abr': int_or_none(stream.get('audio_bitrate')),
+                        'vbr': int_or_none(stream.get('video_bitrate')),
+                        'fps': float_or_none(stream.get('framerate')),
+                    })
+        else:
+            raise ExtractorError('%s said: %s' % (
+                self.IE_NAME, cur_auth_data['message']), expected=True)
+        self._sort_formats(formats)
+
+        subtitles = {}
+        for lang, sub in metadata.get('closed_captions_vtt', {}).get('captions', {}).items():
+            sub_url = sub.get('url')
+            if not sub_url:
+                continue
+            subtitles[lang] = [{
+                'url': sub_url,
+            }]
+
+        return {
+            'id': embed_code,
+            'title': title,
+            'description': metadata.get('description'),
+            'thumbnail': metadata.get('thumbnail_image') or metadata.get('promo_image'),
+            'duration': float_or_none(metadata.get('duration'), 1000),
+            'subtitles': subtitles,
+            'formats': formats,
+        }
+
+
+class OoyalaIE(OoyalaBaseIE):
     _VALID_URL = r'(?:ooyala:|https?://.+?\.ooyala\.com/.*?(?:embedCode|ec)=)(?P<id>.+?)(&|$)'
 
     _TESTS = [
         {
             # From http://it.slashdot.org/story/13/04/25/178216/recovering-data-from-broken-hard-drives-and-ssds-video
             'url': 'http://player.ooyala.com/player.js?embedCode=pxczE2YjpfHfn1f3M-ykG_AmJRRn0PD8',
-            'md5': '3f5cceb3a7bf461d6c29dc466cf8033c',
             'info_dict': {
                 'id': 'pxczE2YjpfHfn1f3M-ykG_AmJRRn0PD8',
                 'ext': 'mp4',
                 'title': 'Explaining Data Recovery from Hard Drives and SSDs',
                 'description': 'How badly damaged does a drive have to be to defeat Russell and his crew? Apparently, smashed to bits.',
+                'duration': 853.386,
             },
+            # The video in the original webpage now uses PlayWire
+            'skip': 'Ooyala said: movie expired',
         }, {
             # Only available for ipad
             'url': 'http://player.ooyala.com/player.js?embedCode=x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0',
-            'md5': '4b9754921fddb68106e48c142e2a01e6',
             'info_dict': {
                 'id': 'x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0',
                 'ext': 'mp4',
                 'title': 'Simulation Overview - Levels of Simulation',
-                'description': '',
+                'duration': 194.948,
             },
         },
+        {
+            # Information available only through SAS api
+            # From http://community.plm.automation.siemens.com/t5/News-NX-Manufacturing/Tool-Path-Divide/ba-p/4187
+            'url': 'http://player.ooyala.com/player.js?embedCode=FiOG81ZTrvckcchQxmalf4aQj590qTEx',
+            'md5': 'a84001441b35ea492bc03736e59e7935',
+            'info_dict': {
+                'id': 'FiOG81ZTrvckcchQxmalf4aQj590qTEx',
+                'ext': 'mp4',
+                'title': 'Divide Tool Path.mp4',
+                'duration': 204.405,
+            }
+        }
     ]
 
     @staticmethod
@@ -43,58 +139,49 @@ class OoyalaIE(InfoExtractor):
     @classmethod
     def _build_url_result(cls, embed_code):
         return cls.url_result(cls._url_for_embed_code(embed_code),
-            ie=cls.ie_key())
+                              ie=cls.ie_key())
 
-    def _extract_result(self, info, more_info):
-        return {
-            'id': info['embedCode'],
+    def _real_extract(self, url):
+        url, smuggled_data = unsmuggle_url(url, {})
+        embed_code = self._match_id(url)
+        domain = smuggled_data.get('domain')
+        supportedformats = smuggled_data.get('supportedformats')
+        content_tree_url = self._CONTENT_TREE_BASE + 'embed_code/%s/%s' % (embed_code, embed_code)
+        return self._extract(content_tree_url, embed_code, domain, supportedformats)
+
+
+class OoyalaExternalIE(OoyalaBaseIE):
+    _VALID_URL = r'''(?x)
+                    (?:
+                        ooyalaexternal:|
+                        https?://.+?\.ooyala\.com/.*?\bexternalId=
+                    )
+                    (?P<partner_id>[^:]+)
+                    :
+                    (?P<id>.+)
+                    (?:
+                        :|
+                        .*?&pcode=
+                    )
+                    (?P<pcode>.+?)
+                    (?:&|$)
+                    '''
+
+    _TEST = {
+        'url': 'https://player.ooyala.com/player.js?externalId=espn:10365079&pcode=1kNG061cgaoolOncv54OAO1ceO-I&adSetCode=91cDU6NuXTGKz3OdjOxFdAgJVtQcKJnI&callback=handleEvents&hasModuleParams=1&height=968&playerBrandingId=7af3bd04449c444c964f347f11873075&targetReplaceId=videoPlayer&width=1656&wmode=opaque&allowScriptAccess=always',
+        'info_dict': {
+            'id': 'FkYWtmazr6Ed8xmvILvKLWjd4QvYZpzG',
             'ext': 'mp4',
-            'title': unescapeHTML(info['title']),
-            'url': info.get('ipad_url') or info['url'],
-            'description': unescapeHTML(more_info['description']),
-            'thumbnail': more_info['promo'],
-        }
+            'title': 'dm_140128_30for30Shorts___JudgingJewellv2',
+            'duration': 1302.0,
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+    }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        embedCode = mobj.group('id')
-        player_url = 'http://player.ooyala.com/player.js?embedCode=%s' % embedCode
-        player = self._download_webpage(player_url, embedCode)
-        mobile_url = self._search_regex(r'mobile_player_url="(.+?)&device="',
-                                        player, 'mobile player url')
-        # Looks like some videos are only available for particular devices
-        # (e.g. http://player.ooyala.com/player.js?embedCode=x1b3lqZDq9y_7kMyC2Op5qo-p077tXD0
-        # is only available for ipad)
-        # Working around with fetching URLs for all the devices found starting with 'unknown'
-        # until we succeed or eventually fail for each device.
-        devices = re.findall(r'device\s*=\s*"([^"]+)";', player)
-        devices.remove('unknown')
-        devices.insert(0, 'unknown')
-        for device in devices:
-            mobile_player = self._download_webpage(
-                '%s&device=%s' % (mobile_url, device), embedCode,
-                'Downloading mobile player JS for %s device' % device)
-            videos_info = self._search_regex(
-                r'var streams=window.oo_testEnv\?\[\]:eval\("\((\[{.*?}\])\)"\);',
-                mobile_player, 'info', fatal=False, default=None)
-            if videos_info:
-                break
-        if not videos_info:
-            raise ExtractorError('Unable to extract info')
-        videos_info = videos_info.replace('\\"', '"')
-        videos_more_info = self._search_regex(
-            r'eval\("\(({.*?\\"promo\\".*?})\)"', mobile_player, 'more info').replace('\\"', '"')
-        videos_info = json.loads(videos_info)
-        videos_more_info = json.loads(videos_more_info)
-
-        if videos_more_info.get('lineup'):
-            videos = [self._extract_result(info, more_info) for (info, more_info) in zip(videos_info, videos_more_info['lineup'])]
-            return {
-                '_type': 'playlist',
-                'id': embedCode,
-                'title': unescapeHTML(videos_more_info['title']),
-                'entries': videos,
-            }
-        else:
-            return self._extract_result(videos_info[0], videos_more_info)
-        
+        partner_id, video_id, pcode = re.match(self._VALID_URL, url).groups()
+        content_tree_url = self._CONTENT_TREE_BASE + 'external_id/%s/%s:%s' % (pcode, partner_id, video_id)
+        return self._extract(content_tree_url, video_id)