X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/f46044c66663049e286c20ee015db99d47d9dd8a..8b4fae8ce16f284d2b7a5bb2ee099e9ecaf0c0d2:/youtube_dl/extractor/letv.py

diff --git a/youtube_dl/extractor/letv.py b/youtube_dl/extractor/letv.py
index 583ce35..9665ece 100644
--- a/youtube_dl/extractor/letv.py
+++ b/youtube_dl/extractor/letv.py
@@ -4,33 +4,43 @@ from __future__ import unicode_literals
 import datetime
 import re
 import time
+import base64
+import hashlib
 
 from .common import InfoExtractor
 from ..compat import (
-    compat_urlparse,
     compat_urllib_parse,
+    compat_ord,
+    compat_str,
 )
 from ..utils import (
     determine_ext,
     ExtractorError,
     parse_iso8601,
+    sanitized_Request,
+    int_or_none,
+    str_or_none,
+    encode_data_uri,
+    url_basename,
 )
 
 
 class LetvIE(InfoExtractor):
+    IE_DESC = '乐视网'
     _VALID_URL = r'http://www\.letv\.com/ptv/vplay/(?P<id>\d+).html'
 
     _TESTS = [{
         'url': 'http://www.letv.com/ptv/vplay/22005890.html',
-        'md5': 'cab23bd68d5a8db9be31c9a222c1e8df',
+        'md5': 'edadcfe5406976f42f9f266057ee5e40',
         'info_dict': {
             'id': '22005890',
             'ext': 'mp4',
             'title': '第87届奥斯卡颁奖礼完美落幕 《鸟人》成最大赢家',
-            'timestamp': 1424747397,
-            'upload_date': '20150224',
             'description': 'md5:a9cb175fd753e2962176b7beca21a47c',
-        }
+        },
+        'params': {
+            'hls_prefer_native': True,
+        },
     }, {
         'url': 'http://www.letv.com/ptv/vplay/1415246.html',
         'info_dict': {
@@ -39,12 +49,24 @@ class LetvIE(InfoExtractor):
             'title': '美人天下01',
             'description': 'md5:f88573d9d7225ada1359eaf0dbf8bcda',
         },
-        'expected_warnings': [
-            'publish time'
-        ]
+        'params': {
+            'hls_prefer_native': True,
+        },
+    }, {
+        'note': 'This video is available only in Mainland China, thus a proxy is needed',
+        'url': 'http://www.letv.com/ptv/vplay/1118082.html',
+        'md5': '2424c74948a62e5f31988438979c5ad1',
+        'info_dict': {
+            'id': '1118082',
+            'ext': 'mp4',
+            'title': '与龙共舞 完整版',
+            'description': 'md5:7506a5eeb1722bb9d4068f85024e3986',
+        },
+        'params': {
+            'hls_prefer_native': True,
+        },
+        'skip': 'Only available in China',
     }]
-    # http://www.letv.com/ptv/vplay/1118082.html
-    # This video is available only in Mainland China
 
     @staticmethod
     def urshift(val, n):
@@ -65,6 +87,27 @@ class LetvIE(InfoExtractor):
         _loc3_ = self.ror(_loc3_, _loc2_ % 17)
         return _loc3_
 
+    # see M3U8Encryption class in KLetvPlayer.swf
+    @staticmethod
+    def decrypt_m3u8(encrypted_data):
+        if encrypted_data[:5].decode('utf-8').lower() != 'vc_01':
+            return encrypted_data
+        encrypted_data = encrypted_data[5:]
+
+        _loc4_ = bytearray()
+        while encrypted_data:
+            b = compat_ord(encrypted_data[0])
+            _loc4_.extend([b // 16, b & 0x0f])
+            encrypted_data = encrypted_data[1:]
+        idx = len(_loc4_) - 11
+        _loc4_ = _loc4_[idx:] + _loc4_[:idx]
+        _loc7_ = bytearray()
+        while _loc4_:
+            _loc7_.append(_loc4_[0] * 16 + _loc4_[1])
+            _loc4_ = _loc4_[2:]
+
+        return bytes(_loc7_)
+
     def _real_extract(self, url):
         media_id = self._match_id(url)
         page = self._download_webpage(url, media_id)
@@ -76,9 +119,16 @@ class LetvIE(InfoExtractor):
             'tkey': self.calc_time_key(int(time.time())),
             'domain': 'www.letv.com'
         }
+        play_json_req = sanitized_Request(
+            'http://api.letv.com/mms/out/video/playJson?' + compat_urllib_parse.urlencode(params)
+        )
+        cn_verification_proxy = self._downloader.params.get('cn_verification_proxy')
+        if cn_verification_proxy:
+            play_json_req.add_header('Ytdl-request-proxy', cn_verification_proxy)
+
         play_json = self._download_json(
-            'http://api.letv.com/mms/out/video/playJson?' + compat_urllib_parse.urlencode(params),
-            media_id, 'playJson data')
+            play_json_req,
+            media_id, 'Downloading playJson data')
 
         # Check for errors
         playstatus = play_json['playstatus']
@@ -99,31 +149,37 @@ class LetvIE(InfoExtractor):
         for format_id in formats:
             if format_id in dispatch:
                 media_url = playurl['domain'][0] + dispatch[format_id][0]
-
-                # Mimic what flvxz.com do
-                url_parts = list(compat_urlparse.urlparse(media_url))
-                qs = dict(compat_urlparse.parse_qs(url_parts[4]))
-                qs.update({
-                    'platid': '14',
-                    'splatid': '1401',
-                    'tss': 'no',
-                    'retry': 1
+                media_url += '&' + compat_urllib_parse.urlencode({
+                    'm3v': 1,
+                    'format': 1,
+                    'expect': 3,
+                    'rateid': format_id,
                 })
-                url_parts[4] = compat_urllib_parse.urlencode(qs)
-                media_url = compat_urlparse.urlunparse(url_parts)
+
+                nodes_data = self._download_json(
+                    media_url, media_id,
+                    'Download JSON metadata for format %s' % format_id)
+
+                req = self._request_webpage(
+                    nodes_data['nodelist'][0]['location'], media_id,
+                    note='Downloading m3u8 information for format %s' % format_id)
+
+                m3u8_data = self.decrypt_m3u8(req.read())
 
                 url_info_dict = {
-                    'url': media_url,
-                    'ext': determine_ext(dispatch[format_id][1])
+                    'url': encode_data_uri(m3u8_data, 'application/vnd.apple.mpegurl'),
+                    'ext': determine_ext(dispatch[format_id][1]),
+                    'format_id': format_id,
+                    'protocol': 'm3u8',
                 }
 
                 if format_id[-1:] == 'p':
-                    url_info_dict['height'] = format_id[:-1]
+                    url_info_dict['height'] = int_or_none(format_id[:-1])
 
                 urls.append(url_info_dict)
 
         publish_time = parse_iso8601(self._html_search_regex(
-            r'发布时间&nbsp;([^<>]+) ', page, 'publish time', fatal=False),
+            r'发布时间&nbsp;([^<>]+) ', page, 'publish time', default=None),
             delimiter=' ', timezone=datetime.timedelta(hours=8))
         description = self._html_search_meta('description', page, fatal=False)
 
@@ -188,3 +244,113 @@ class LetvPlaylistIE(LetvTvIE):
         },
         'playlist_mincount': 7
     }]
+
+
+class LetvCloudIE(InfoExtractor):
+    IE_DESC = '乐视云'
+    _VALID_URL = r'https?://yuntv\.letv\.com/bcloud.html\?.+'
+
+    _TESTS = [{
+        'url': 'http://yuntv.letv.com/bcloud.html?uu=p7jnfw5hw9&vu=467623dedf',
+        'md5': '26450599afd64c513bc77030ad15db44',
+        'info_dict': {
+            'id': 'p7jnfw5hw9_467623dedf',
+            'ext': 'mp4',
+            'title': 'Video p7jnfw5hw9_467623dedf',
+        },
+    }, {
+        'url': 'http://yuntv.letv.com/bcloud.html?uu=p7jnfw5hw9&vu=ec93197892&pu=2c7cd40209&auto_play=1&gpcflag=1&width=640&height=360',
+        'md5': 'e03d9cc8d9c13191e1caf277e42dbd31',
+        'info_dict': {
+            'id': 'p7jnfw5hw9_ec93197892',
+            'ext': 'mp4',
+            'title': 'Video p7jnfw5hw9_ec93197892',
+        },
+    }, {
+        'url': 'http://yuntv.letv.com/bcloud.html?uu=p7jnfw5hw9&vu=187060b6fd',
+        'md5': 'cb988699a776b22d4a41b9d43acfb3ac',
+        'info_dict': {
+            'id': 'p7jnfw5hw9_187060b6fd',
+            'ext': 'mp4',
+            'title': 'Video p7jnfw5hw9_187060b6fd',
+        },
+    }]
+
+    @staticmethod
+    def sign_data(obj):
+        if obj['cf'] == 'flash':
+            salt = '2f9d6924b33a165a6d8b5d3d42f4f987'
+            items = ['cf', 'format', 'ran', 'uu', 'ver', 'vu']
+        elif obj['cf'] == 'html5':
+            salt = 'fbeh5player12c43eccf2bec3300344'
+            items = ['cf', 'ran', 'uu', 'bver', 'vu']
+        input_data = ''.join([item + obj[item] for item in items]) + salt
+        obj['sign'] = hashlib.md5(input_data.encode('utf-8')).hexdigest()
+
+    def _get_formats(self, cf, uu, vu, media_id):
+        def get_play_json(cf, timestamp):
+            data = {
+                'cf': cf,
+                'ver': '2.2',
+                'bver': 'firefox44.0',
+                'format': 'json',
+                'uu': uu,
+                'vu': vu,
+                'ran': compat_str(timestamp),
+            }
+            self.sign_data(data)
+            return self._download_json(
+                'http://api.letvcloud.com/gpc.php?' + compat_urllib_parse.urlencode(data),
+                media_id, 'Downloading playJson data for type %s' % cf)
+
+        play_json = get_play_json(cf, time.time())
+        # The server time may be different from local time
+        if play_json.get('code') == 10071:
+            play_json = get_play_json(cf, play_json['timestamp'])
+
+        if not play_json.get('data'):
+            if play_json.get('message'):
+                raise ExtractorError('Letv cloud said: %s' % play_json['message'], expected=True)
+            elif play_json.get('code'):
+                raise ExtractorError('Letv cloud returned error %d' % play_json['code'], expected=True)
+            else:
+                raise ExtractorError('Letv cloud returned an unknwon error')
+
+        def b64decode(s):
+            return base64.b64decode(s.encode('utf-8')).decode('utf-8')
+
+        formats = []
+        for media in play_json['data']['video_info']['media'].values():
+            play_url = media['play_url']
+            url = b64decode(play_url['main_url'])
+            decoded_url = b64decode(url_basename(url))
+            formats.append({
+                'url': url,
+                'ext': determine_ext(decoded_url),
+                'format_id': int_or_none(play_url.get('vtype')),
+                'format_note': str_or_none(play_url.get('definition')),
+                'width': int_or_none(play_url.get('vwidth')),
+                'height': int_or_none(play_url.get('vheight')),
+            })
+
+        return formats
+
+    def _real_extract(self, url):
+        uu_mobj = re.search('uu=([\w]+)', url)
+        vu_mobj = re.search('vu=([\w]+)', url)
+
+        if not uu_mobj or not vu_mobj:
+            raise ExtractorError('Invalid URL: %s' % url, expected=True)
+
+        uu = uu_mobj.group(1)
+        vu = vu_mobj.group(1)
+        media_id = uu + '_' + vu
+
+        formats = self._get_formats('flash', uu, vu, media_id) + self._get_formats('html5', uu, vu, media_id)
+        self._sort_formats(formats)
+
+        return {
+            'id': media_id,
+            'title': 'Video %s' % media_id,
+            'formats': formats,
+        }