Merge pull request #1 from rg3/master

update `1st may

youtube_dl/extractor/extractors.py

@@ -941,6 +941,12 @@ from .xhamster import (
     XHamsterIE,
     XHamsterEmbedIE,
 )
+from .xiami import (
+    XiamiSongIE,
+    XiamiAlbumIE,
+    XiamiArtistIE,
+    XiamiCollectionIE
+)
 from .xminus import XMinusIE
 from .xnxx import XNXXIE
 from .xstream import XstreamIE

youtube_dl/extractor/pbs.py

@@ -512,7 +512,9 @@ class PBSIE(InfoExtractor):
         if http_url:
             for m3u8_format in m3u8_formats:
                 bitrate = self._search_regex(r'(\d+k)', m3u8_format['url'], 'bitrate', default=None)
-                if not bitrate:
+                # extract only the formats that we know that they will be available as http format.
+                # https://projects.pbs.org/confluence/display/coveapi/COVE+Video+Specifications
+                if not bitrate or bitrate not in ('400k', '800k', '1200k', '2500k'):
                     continue
                 f = m3u8_format.copy()
                 f.update({
@@ -535,6 +537,19 @@ class PBSIE(InfoExtractor):
                 'ext': 'ttml',
                 'url': closed_captions_url,
             }]
+            mobj = re.search(r'/(\d+)_Encoded\.dfxp', closed_captions_url)
+            if mobj:
+                ttml_caption_suffix, ttml_caption_id = mobj.group(0, 1)
+                ttml_caption_id = int(ttml_caption_id)
+                subtitles['en'].extend([{
+                    'url': closed_captions_url.replace(
+                        ttml_caption_suffix, '/%d_Encoded.srt' % (ttml_caption_id + 1)),
+                    'ext': 'srt',
+                }, {
+                    'url': closed_captions_url.replace(
+                        ttml_caption_suffix, '/%d_Encoded.vtt' % (ttml_caption_id + 2)),
+                    'ext': 'vtt',
+                }])
 
         # info['title'] is often incomplete (e.g. 'Full Episode', 'Episode 5', etc)
         # Try turning it to 'program - title' naming scheme if possible

youtube_dl/extractor/ted.py

@@ -27,7 +27,7 @@ class TEDIE(InfoExtractor):
         '''
     _TESTS = [{
         'url': 'http://www.ted.com/talks/dan_dennett_on_our_consciousness.html',
-        'md5': 'fc94ac279feebbce69f21c0c6ee82810',
+        'md5': '0de43ac406aa3e4ea74b66c9c7789b13',
         'info_dict': {
             'id': '102',
             'ext': 'mp4',
@@ -37,21 +37,26 @@ class TEDIE(InfoExtractor):
                             'consciousness, but that half the time our brains are '
                             'actively fooling us.'),
             'uploader': 'Dan Dennett',
-            'width': 854,
+            'width': 853,
             'duration': 1308,
         }
     }, {
         'url': 'http://www.ted.com/watch/ted-institute/ted-bcg/vishal-sikka-the-beauty-and-power-of-algorithms',
-        'md5': '226f4fb9c62380d11b7995efa4c87994',
+        'md5': 'b899ac15e345fb39534d913f7606082b',
         'info_dict': {
-            'id': 'vishal-sikka-the-beauty-and-power-of-algorithms',
+            'id': 'tSVI8ta_P4w',
             'ext': 'mp4',
             'title': 'Vishal Sikka: The beauty and power of algorithms',
             'thumbnail': 're:^https?://.+\.jpg',
-            'description': 'Adaptive, intelligent, and consistent, algorithms are emerging as the ultimate app for everything from matching consumers to products to assessing medical diagnoses. Vishal Sikka shares his appreciation for the algorithm, charting both its inherent beauty and its growing power.',
+            'description': 'md5:6261fdfe3e02f4f579cbbfc00aff73f4',
-        }
+            'upload_date': '20140122',
+            'uploader_id': 'TEDInstitute',
+            'uploader': 'TED Institute',
+        },
+        'add_ie': ['Youtube'],
     }, {
         'url': 'http://www.ted.com/talks/gabby_giffords_and_mark_kelly_be_passionate_be_courageous_be_your_best',
+        'md5': '71b3ab2f4233012dce09d515c9c39ce2',
         'info_dict': {
             'id': '1972',
             'ext': 'mp4',
@@ -102,9 +107,9 @@ class TEDIE(InfoExtractor):
     }]
 
     _NATIVE_FORMATS = {
-        'low': {'preference': 1, 'width': 320, 'height': 180},
+        'low': {'width': 320, 'height': 180},
-        'medium': {'preference': 2, 'width': 512, 'height': 288},
+        'medium': {'width': 512, 'height': 288},
-        'high': {'preference': 3, 'width': 854, 'height': 480},
+        'high': {'width': 854, 'height': 480},
     }
 
     def _extract_info(self, webpage):
@@ -171,15 +176,21 @@ class TEDIE(InfoExtractor):
                 if finfo:
                     f.update(finfo)
 
+        http_url = None
         for format_id, resources in talk_info['resources'].items():
             if format_id == 'h264':
                 for resource in resources:
+                    h264_url = resource.get('file')
+                    if not h264_url:
+                        continue
                     bitrate = int_or_none(resource.get('bitrate'))
                     formats.append({
-                        'url': resource['file'],
+                        'url': h264_url,
                         'format_id': '%s-%sk' % (format_id, bitrate),
                         'tbr': bitrate,
                     })
+                    if re.search('\d+k', h264_url):
+                        http_url = h264_url
             elif format_id == 'rtmp':
                 streamer = talk_info.get('streamer')
                 if not streamer:
@@ -195,16 +206,24 @@ class TEDIE(InfoExtractor):
                         'tbr': int_or_none(resource.get('bitrate')),
                     })
             elif format_id == 'hls':
-                hls_formats = self._extract_m3u8_formats(
+                formats.extend(self._extract_m3u8_formats(
-                    resources.get('stream'), video_name, 'mp4', m3u8_id=format_id)
+                    resources.get('stream'), video_name, 'mp4', m3u8_id=format_id, fatal=False))
-                for f in hls_formats:
+
-                    if f.get('format_id') == 'hls-meta':
+        m3u8_formats = list(filter(
-                        continue
+            lambda f: f.get('protocol') == 'm3u8' and f.get('vcodec') != 'none' and f.get('resolution') != 'multiple',
-                    if not f.get('height'):
+            formats))
-                        f['vcodec'] = 'none'
+        if http_url:
-                    else:
+            for m3u8_format in m3u8_formats:
-                        f['acodec'] = 'none'
+                bitrate = self._search_regex(r'(\d+k)', m3u8_format['url'], 'bitrate', default=None)
-                formats.extend(hls_formats)
+                if not bitrate:
+                    continue
+                f = m3u8_format.copy()
+                f.update({
+                    'url': re.sub(r'\d+k', bitrate, http_url),
+                    'format_id': m3u8_format['format_id'].replace('hls', 'http'),
+                    'protocol': 'http',
+                })
+                formats.append(f)
 
         audio_download = talk_info.get('audioDownload')
         if audio_download:
@@ -212,7 +231,6 @@ class TEDIE(InfoExtractor):
                 'url': audio_download,
                 'format_id': 'audio',
                 'vcodec': 'none',
-                'preference': -0.5,
             })
 
         self._sort_formats(formats)
@@ -254,7 +272,11 @@ class TEDIE(InfoExtractor):
 
         config_json = self._html_search_regex(
             r'"pages\.jwplayer"\s*,\s*({.+?})\s*\)\s*</script>',
-            webpage, 'config')
+            webpage, 'config', default=None)
+        if not config_json:
+            embed_url = self._search_regex(
+                r"<iframe[^>]+class='pages-video-embed__video__object'[^>]+src='([^']+)'", webpage, 'embed url')
+            return self.url_result(self._proto_relative_url(embed_url))
         config = json.loads(config_json)['config']
         video_url = config['video']['url']
         thumbnail = config.get('image', {}).get('url')

youtube_dl/extractor/vlive.py

@@ -43,7 +43,7 @@ class VLiveIE(InfoExtractor):
         status_params = self._download_json(
             'http://www.vlive.tv/video/status?videoSeq=%s' % video_id,
             video_id, 'Downloading JSON status',
-            headers={'Referer': url})
+            headers={'Referer': url.encode('utf-8')})
         status = status_params.get('status')
         air_start = status_params.get('onAirStartAt', '')
         is_live = status_params.get('isLive')

youtube_dl/extractor/xiami.py

@@ -0,0 +1,158 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..compat import compat_urllib_parse_unquote
+from ..utils import int_or_none
+
+
+class XiamiBaseIE(InfoExtractor):
+    _API_BASE_URL = 'http://www.xiami.com/song/playlist/cat/json/id'
+
+    def _extract_track(self, track, track_id=None):
+        title = track['title']
+        track_url = self._decrypt(track['location'])
+
+        subtitles = {}
+        lyrics_url = track.get('lyric_url') or track.get('lyric')
+        if lyrics_url and lyrics_url.startswith('http'):
+            subtitles['origin'] = [{'url': lyrics_url}]
+
+        return {
+            'id': track.get('song_id') or track_id,
+            'url': track_url,
+            'title': title,
+            'thumbnail': track.get('pic') or track.get('album_pic'),
+            'duration': int_or_none(track.get('length')),
+            'creator': track.get('artist', '').split(';')[0],
+            'track': title,
+            'album': track.get('album_name'),
+            'artist': track.get('artist'),
+            'subtitles': subtitles,
+        }
+
+    def _extract_tracks(self, item_id, typ=None):
+        playlist = self._download_json(
+            '%s/%s%s' % (self._API_BASE_URL, item_id, '/type/%s' % typ if typ else ''), item_id)
+        return [
+            self._extract_track(track, item_id)
+            for track in playlist['data']['trackList']]
+
+    @staticmethod
+    def _decrypt(origin):
+        n = int(origin[0])
+        origin = origin[1:]
+        short_lenth = len(origin) // n
+        long_num = len(origin) - short_lenth * n
+        l = tuple()
+        for i in range(0, n):
+            length = short_lenth
+            if i < long_num:
+                length += 1
+            l += (origin[0:length], )
+            origin = origin[length:]
+        ans = ''
+        for i in range(0, short_lenth + 1):
+            for j in range(0, n):
+                if len(l[j]) > i:
+                    ans += l[j][i]
+        return compat_urllib_parse_unquote(ans).replace('^', '0')
+
+
+class XiamiSongIE(XiamiBaseIE):
+    IE_NAME = 'xiami:song'
+    IE_DESC = '虾米音乐'
+    _VALID_URL = r'https?://(?:www\.)?xiami\.com/song/(?P<id>[0-9]+)'
+    _TESTS = [{
+        'url': 'http://www.xiami.com/song/1775610518',
+        'md5': '521dd6bea40fd5c9c69f913c232cb57e',
+        'info_dict': {
+            'id': '1775610518',
+            'ext': 'mp3',
+            'title': 'Woman',
+            'thumbnail': r're:http://img\.xiami\.net/images/album/.*\.jpg',
+            'duration': 265,
+            'creator': 'HONNE',
+            'track': 'Woman',
+            'album': 'Woman',
+            'artist': 'HONNE',
+            'subtitles': {
+                'origin': [{
+                    'ext': 'lrc',
+                }],
+            },
+        }
+    }, {
+        'url': 'http://www.xiami.com/song/1775256504',
+        'md5': '932a3abd45c6aa2b1fdbe028fcb4c4fc',
+        'info_dict': {
+            'id': '1775256504',
+            'ext': 'mp3',
+            'title': '悟空',
+            'thumbnail': r're:http://img\.xiami\.net/images/album/.*\.jpg',
+            'duration': 200,
+            'creator': '戴荃',
+            'track': '悟空',
+            'album': '悟空',
+            'artist': '戴荃',
+            'subtitles': {
+                'origin': [{
+                    'ext': 'lrc',
+                }],
+            },
+        }
+    }]
+
+    def _real_extract(self, url):
+        return self._extract_tracks(self._match_id(url))[0]
+
+
+class XiamiPlaylistBaseIE(XiamiBaseIE):
+    def _real_extract(self, url):
+        item_id = self._match_id(url)
+        return self.playlist_result(self._extract_tracks(item_id, self._TYPE), item_id)
+
+
+class XiamiAlbumIE(XiamiPlaylistBaseIE):
+    IE_NAME = 'xiami:album'
+    IE_DESC = '虾米音乐 - 专辑'
+    _VALID_URL = r'https?://(?:www\.)?xiami\.com/album/(?P<id>[0-9]+)'
+    _TYPE = '1'
+    _TESTS = [{
+        'url': 'http://www.xiami.com/album/2100300444',
+        'info_dict': {
+            'id': '2100300444',
+        },
+        'playlist_count': 10,
+    }, {
+        'url': 'http://www.xiami.com/album/512288?spm=a1z1s.6843761.1110925389.6.hhE9p9',
+        'only_matching': True,
+    }]
+
+
+class XiamiArtistIE(XiamiPlaylistBaseIE):
+    IE_NAME = 'xiami:artist'
+    IE_DESC = '虾米音乐 - 歌手'
+    _VALID_URL = r'https?://(?:www\.)?xiami\.com/artist/(?P<id>[0-9]+)'
+    _TYPE = '2'
+    _TEST = {
+        'url': 'http://www.xiami.com/artist/2132?spm=0.0.0.0.dKaScp',
+        'info_dict': {
+            'id': '2132',
+        },
+        'playlist_count': 20,
+    }
+
+
+class XiamiCollectionIE(XiamiPlaylistBaseIE):
+    IE_NAME = 'xiami:collection'
+    IE_DESC = '虾米音乐 - 精选集'
+    _VALID_URL = r'https?://(?:www\.)?xiami\.com/collect/(?P<id>[0-9]+)'
+    _TYPE = '3'
+    _TEST = {
+        'url': 'http://www.xiami.com/collect/156527391?spm=a1z1s.2943601.6856193.12.4jpBnr',
+        'info_dict': {
+            'id': '156527391',
+        },
+        'playlist_mincount': 29,
+    }