[limelight] Fix python 2.6, simplify, make more robust (Closes #6734)
authorSergey M․ <dstftw@gmail.com>
Sun, 4 Oct 2015 14:41:57 +0000 (20:41 +0600)
committerSergey M․ <dstftw@gmail.com>
Sun, 4 Oct 2015 14:42:35 +0000 (20:42 +0600)
youtube_dl/extractor/__init__.py
youtube_dl/extractor/limelight.py

index 20cc3660cebd6e23d77d396a9cf06b4263eae53c..3ace1cc2ca7a375820d423f495d57c4e76aae018 100644 (file)
@@ -296,9 +296,9 @@ from .lifenews import (
     LifeEmbedIE,
 )
 from .limelight import (
-    LimeLightMediaIE,
-    LimeLightChannelIE,
-    LimeLightChannelListIE,
+    LimelightMediaIE,
+    LimelightChannelIE,
+    LimelightChannelListIE,
 )
 from .liveleak import LiveLeakIE
 from .livestream import (
index dcfc215c790f78723acfb65cb5e33b51d8bf8085..599d8413dde4e9c51ae74f16329f5ca0c28bf50d 100644 (file)
@@ -5,65 +5,105 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
-    int_or_none,
     determine_ext,
+    float_or_none,
+    int_or_none,
 )
 
 
-class LimeLightBaseIE(InfoExtractor):
+class LimelightBaseIE(InfoExtractor):
+    _PLAYLIST_SERVICE_URL = 'http://production-ps.lvp.llnw.net/r/PlaylistService/%s/%s/%s'
+    _API_URL = 'http://api.video.limelight.com/rest/organizations/%s/%s/%s/%s.json'
 
-    def get_playlist_service(self, id, method):
-        return self._download_json(self.PLAYLIST_SERVICE_URL % (id, method), id)
+    def _call_playlist_service(self, item_id, method, fatal=True):
+        return self._download_json(
+            self._PLAYLIST_SERVICE_URL % (self._PLAYLIST_SERVICE_PATH, item_id, method),
+            item_id, 'Downloading PlaylistService %s JSON' % method, fatal=fatal)
 
-    def get_api(self, orgId, id, method):
-        return self._download_json(self.API_URL % (orgId, id, method), id)
+    def _call_api(self, organization_id, item_id, method):
+        return self._download_json(
+            self._API_URL % (organization_id, self._API_PATH, item_id, method),
+            item_id, 'Downloading API %s JSON' % method)
 
-    def process_data(self, mobileUrls, streams, properties):
+    def _extract(self, item_id, pc_method, mobile_method, meta_method):
+        pc = self._call_playlist_service(item_id, pc_method)
+        metadata = self._call_api(pc['orgId'], item_id, meta_method)
+        mobile = self._call_playlist_service(item_id, mobile_method, fatal=False)
+        return pc, mobile, metadata
+
+    def _extract_info(self, streams, mobile_urls, properties):
         video_id = properties['media_id']
         formats = []
 
-        for mobileUrl in mobileUrls:
-            if '.m3u8' in mobileUrl['mobileUrl']:
-                formats.extend(self._extract_m3u8_formats(mobileUrl['mobileUrl'], video_id))
-            else:
-                formats.append({'url': mobileUrl['mobileUrl']})
-
         for stream in streams:
-            if '.f4m' in stream['url']:
-                formats.extend(self._extract_f4m_formats(stream['url'], video_id))
+            stream_url = stream.get('url')
+            if not stream_url:
+                continue
+            if '.f4m' in stream_url:
+                formats.extend(self._extract_f4m_formats(stream_url, video_id))
             else:
                 fmt = {
-                    'url': stream.get('url'),
-                    'abr': stream.get('audioBitRate'),
-                    'vbr': stream.get('videoBitRate'),
-                    'fps': stream.get('videoFrameRate'),
-                    'width': stream.get('videoWidthInPixels'),
-                    'height': stream.get('videoHeightInPixels'),
-                    'ext': determine_ext(stream.get('url'))
+                    'url': stream_url,
+                    'abr': float_or_none(stream.get('audioBitRate')),
+                    'vbr': float_or_none(stream.get('videoBitRate')),
+                    'fps': float_or_none(stream.get('videoFrameRate')),
+                    'width': int_or_none(stream.get('videoWidthInPixels')),
+                    'height': int_or_none(stream.get('videoHeightInPixels')),
+                    'ext': determine_ext(stream_url)
                 }
-                rtmp = re.search(r'^(?P<url>rtmp://[^/]+/(?P<app>.+))/(?P<playpath>mp4:.+)$', stream['url'])
+                rtmp = re.search(r'^(?P<url>rtmpe?://[^/]+/(?P<app>.+))/(?P<playpath>mp4:.+)$', stream_url)
                 if rtmp:
+                    format_id = 'rtmp'
+                    if stream.get('videoBitRate'):
+                        format_id += '-%d' % int_or_none(stream['videoBitRate'])
                     fmt.update({
                         'url': rtmp.group('url'),
                         'play_path': rtmp.group('playpath'),
                         'app': rtmp.group('app'),
+                        'ext': 'flv',
+                        'format_id': format_id,
                     })
                 formats.append(fmt)
 
+        for mobile_url in mobile_urls:
+            media_url = mobile_url.get('mobileUrl')
+            if not media_url:
+                continue
+            format_id = mobile_url.get('targetMediaPlatform')
+            if determine_ext(media_url) == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    media_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    preference=-1, m3u8_id=format_id))
+            else:
+                formats.append({
+                    'url': media_url,
+                    'format_id': format_id,
+                    'preference': -1,
+                })
+
         self._sort_formats(formats)
 
         title = properties['title']
         description = properties.get('description')
-        timestamp = properties.get('create_date')
-        duration = int_or_none(properties.get('duration_in_milliseconds'))
-        filesize = properties.get('total_storage_in_bytes')
+        timestamp = int_or_none(properties.get('publish_date') or properties.get('create_date'))
+        duration = float_or_none(properties.get('duration_in_milliseconds'), 1000)
+        filesize = int_or_none(properties.get('total_storage_in_bytes'))
         categories = [properties.get('category')]
+        tags = properties.get('tags', [])
         thumbnails = [{
-            'url': thumbnail.get('url'),
+            'url': thumbnail['url'],
             'width': int_or_none(thumbnail.get('width')),
             'height': int_or_none(thumbnail.get('height')),
-        } for thumbnail in properties.get('thumbnails')]
-        subtitles = {caption.get('language_code'): [{'url': caption.get('url')}] for caption in properties.get('captions')}
+        } for thumbnail in properties.get('thumbnails', []) if thumbnail.get('url')]
+
+        subtitles = {}
+        for caption in properties.get('captions', {}):
+            lang = caption.get('language_code')
+            subtitles_url = caption.get('url')
+            if lang and subtitles_url:
+                subtitles[lang] = [{
+                    'url': subtitles_url,
+                }]
 
         return {
             'id': video_id,
@@ -74,44 +114,50 @@ class LimeLightBaseIE(InfoExtractor):
             'duration': duration,
             'filesize': filesize,
             'categories': categories,
+            'tags': tags,
             'thumbnails': thumbnails,
             'subtitles': subtitles,
         }
 
 
-class LimeLightMediaIE(LimeLightBaseIE):
+class LimelightMediaIE(LimelightBaseIE):
     IE_NAME = 'limelight'
-    _VALID_URL = r'http://link\.videoplatform\.limelight\.com/media/?.*mediaId=(?P<id>[a-z0-9]{32})'
+    _VALID_URL = r'(?:limelight:media:|http://link\.videoplatform\.limelight\.com/media/\??\bmediaId=)(?P<id>[a-z0-9]{32})'
     _TEST = {
         'url': 'http://link.videoplatform.limelight.com/media/?mediaId=3ffd040b522b4485b6d84effc750cd86',
-        'md5': '3213605088be599705677ef785db6972',
         'info_dict': {
             'id': '3ffd040b522b4485b6d84effc750cd86',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 'HaP and the HB Prince Trailer',
             'description': 'As Harry Potter begins his 6th year at Hogwarts School of Witchcraft and Wizardry, he discovers an old book marked mysteriously "This book is the property of the Half-Blood Prince" and begins to learn more about Lord Voldemort\'s dark past.',
             'thumbnail': 're:^https?://.*\.jpeg$',
-            'duration': 144230,
+            'duration': 144.23,
             'timestamp': 1244136834,
-            "upload_date": "20090604",
-        }
+            'upload_date': '20090604',
+        },
+        'params': {
+            # rtmp download
+            'skip_download': True,
+        },
     }
-    PLAYLIST_SERVICE_URL = 'http://production-ps.lvp.llnw.net/r/PlaylistService/media/%s/%s'
-    API_URL = 'http://api.video.limelight.com/rest/organizations/%s/media/%s/%s.json'
+    _PLAYLIST_SERVICE_PATH = 'media'
+    _API_PATH = 'media'
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        mobile_json_data = self.get_playlist_service(video_id, 'getMobilePlaylistByMediaId')
-        pc_json_data = self.get_playlist_service(video_id, 'getPlaylistByMediaId')
-        properties = self.get_api(pc_json_data['orgId'], video_id, 'properties')
+        pc, mobile, metadata = self._extract(
+            video_id, 'getPlaylistByMediaId', 'getMobilePlaylistByMediaId', 'properties')
 
-        return self.process_data(mobile_json_data['mediaList'][0]['mobileUrls'], pc_json_data['playlistItems'][0]['streams'], properties)
+        return self._extract_info(
+            pc['playlistItems'][0].get('streams', []),
+            mobile['mediaList'][0].get('mobileUrls', []) if mobile else [],
+            metadata)
 
 
-class LimeLightChannelIE(LimeLightBaseIE):
+class LimelightChannelIE(LimelightBaseIE):
     IE_NAME = 'limelight:channel'
-    _VALID_URL = r'http://link\.videoplatform\.limelight\.com/media/?.*channelId=(?P<id>[a-z0-9]{32})'
+    _VALID_URL = r'(?:limelight:channel:|http://link\.videoplatform\.limelight\.com/media/\??\bchannelId=)(?P<id>[a-z0-9]{32})'
     _TEST = {
         'url': 'http://link.videoplatform.limelight.com/media/?channelId=ab6a524c379342f9b23642917020c082',
         'info_dict': {
@@ -120,31 +166,29 @@ class LimeLightChannelIE(LimeLightBaseIE):
         },
         'playlist_mincount': 3,
     }
-    PLAYLIST_SERVICE_URL = 'http://production-ps.lvp.llnw.net/r/PlaylistService/channel/%s/%s'
-    API_URL = 'http://api.video.limelight.com/rest/organizations/%s/channels/%s/%s.json'
+    _PLAYLIST_SERVICE_PATH = 'channel'
+    _API_PATH = 'channels'
 
     def _real_extract(self, url):
         channel_id = self._match_id(url)
 
-        mobile_json_data = self.get_playlist_service(channel_id, 'getMobilePlaylistWithNItemsByChannelId?begin=0&count=-1')
-        pc_json_data = self.get_playlist_service(channel_id, 'getPlaylistByChannelId')
-        medias = self.get_api(pc_json_data['orgId'], channel_id, 'media')
+        pc, mobile, medias = self._extract(
+            channel_id, 'getPlaylistByChannelId',
+            'getMobilePlaylistWithNItemsByChannelId?begin=0&count=-1', 'media')
 
-        entries = []
-        for i in range(len(medias['media_list'])):
-            entries.append(self.process_data(mobile_json_data['mediaList'][i]['mobileUrls'], pc_json_data['playlistItems'][i]['streams'], medias['media_list'][i]))
+        entries = [
+            self._extract_info(
+                pc['playlistItems'][i].get('streams', []),
+                mobile['mediaList'][i].get('mobileUrls', []) if mobile else [],
+                medias['media_list'][i])
+            for i in range(len(medias['media_list']))]
 
-        return {
-            'id': channel_id,
-            'title': pc_json_data['title'],
-            'entries': entries,
-            '_type': 'playlist',
-        }
+        return self.playlist_result(entries, channel_id, pc['title'])
 
 
-class LimeLightChannelListIE(LimeLightBaseIE):
+class LimelightChannelListIE(LimelightBaseIE):
     IE_NAME = 'limelight:channel_list'
-    _VALID_URL = r'http://link\.videoplatform\.limelight\.com/media/?.*channelListId=(?P<id>[a-z0-9]{32})'
+    _VALID_URL = r'(?:limelight:channel_list:|http://link\.videoplatform\.limelight\.com/media/\?.*?\bchannelListId=)(?P<id>[a-z0-9]{32})'
     _TEST = {
         'url': 'http://link.videoplatform.limelight.com/media/?channelListId=301b117890c4465c8179ede21fd92e2b',
         'info_dict': {
@@ -153,24 +197,15 @@ class LimeLightChannelListIE(LimeLightBaseIE):
         },
         'playlist_mincount': 2,
     }
-    PLAYLIST_SERVICE_URL = 'http://production-ps.lvp.llnw.net/r/PlaylistService/channel_list/%s/%s'
+    _PLAYLIST_SERVICE_PATH = 'channel_list'
 
     def _real_extract(self, url):
         channel_list_id = self._match_id(url)
 
-        json_data = self.get_playlist_service(channel_list_id, 'getMobileChannelListById')
+        channel_list = self._call_playlist_service(channel_list_id, 'getMobileChannelListById')
 
-        entries = []
-        for channel in json_data['channelList']:
-            entries.append({
-                'url': 'http://link.videoplatform.limelight.com/media/?channelId=%s' % channel['id'],
-                '_type': 'url',
-                'ie_key': 'LimeLightChannel',
-            })
+        entries = [
+            self.url_result('limelight:channel:%s' % channel['id'], 'LimelightChannel')
+            for channel in channel_list['channelList']]
 
-        return {
-            'id': channel_list_id,
-            'title': json_data['title'],
-            'entries': entries,
-            '_type': 'playlist',
-        }
+        return self.playlist_result(entries, channel_list_id, channel_list['title'])