]> gitweb @ CieloNegro.org - youtube-dl.git/commitdiff
[le] extract http formats
authorRemita Amine <remitamine@gmail.com>
Fri, 8 Jul 2016 14:35:20 +0000 (15:35 +0100)
committerRemita Amine <remitamine@gmail.com>
Fri, 8 Jul 2016 14:35:20 +0000 (15:35 +0100)
youtube_dl/extractor/leeco.py

index 171b705c70444b0b705f1ef2896469eaae7a0034..e9cc9aa5983967861b08a2d9ee79297ae3a1726e 100644 (file)
@@ -23,6 +23,7 @@ from ..utils import (
     str_or_none,
     url_basename,
     urshift,
+    update_url_query,
 )
 
 
@@ -89,6 +90,10 @@ class LeIE(InfoExtractor):
         _loc3_ = self.ror(_loc3_, _loc2_ % 17)
         return _loc3_
 
+    # reversed from http://jstatic.letvcdn.com/sdk/player.js
+    def get_mms_key(self, time):
+        return self.ror(time, 8) ^ 185025305
+
     # see M3U8Encryption class in KLetvPlayer.swf
     @staticmethod
     def decrypt_m3u8(encrypted_data):
@@ -109,23 +114,7 @@ class LeIE(InfoExtractor):
 
         return bytes(_loc7_)
 
-    def _real_extract(self, url):
-        media_id = self._match_id(url)
-        page = self._download_webpage(url, media_id)
-        params = {
-            'id': media_id,
-            'platid': 1,
-            'splatid': 101,
-            'format': 1,
-            'tkey': self.calc_time_key(int(time.time())),
-            'domain': 'www.le.com'
-        }
-
-        play_json = self._download_json(
-            'http://api.le.com/mms/out/video/playJson',
-            media_id, 'Downloading playJson data', query=params,
-            headers=self.geo_verification_headers())
-
+    def _check_errors(self, play_json):
         # Check for errors
         playstatus = play_json['playstatus']
         if playstatus['status'] == 0:
@@ -136,43 +125,99 @@ class LeIE(InfoExtractor):
                 msg = 'Generic error. flag = %d' % flag
             raise ExtractorError(msg, expected=True)
 
-        playurl = play_json['playurl']
+    def _real_extract(self, url):
+        media_id = self._match_id(url)
+        page = self._download_webpage(url, media_id)
+
+        play_json_h5 = self._download_json(
+            'http://api.le.com/mms/out/video/playJsonH5',
+            media_id, 'Downloading html5 playJson data', query={
+                'id': media_id,
+                'platid': 3,
+                'splatid': 304,
+                'format': 1,
+                'tkey': self.get_mms_key(int(time.time())),
+                'domain': 'www.le.com',
+                'tss': 'no',
+            },
+            headers=self.geo_verification_headers())
+        self._check_errors(play_json_h5)
 
-        formats = ['350', '1000', '1300', '720p', '1080p']
-        dispatch = playurl['dispatch']
+        play_json_flash = self._download_json(
+            'http://api.le.com/mms/out/video/playJson',
+            media_id, 'Downloading flash playJson data', query={
+                'id': media_id,
+                'platid': 1,
+                'splatid': 101,
+                'format': 1,
+                'tkey': self.calc_time_key(int(time.time())),
+                'domain': 'www.le.com',
+            },
+            headers=self.geo_verification_headers())
+        self._check_errors(play_json_flash)
 
-        urls = []
-        for format_id in formats:
-            if format_id in dispatch:
-                media_url = playurl['domain'][0] + dispatch[format_id][0]
-                media_url += '&' + compat_urllib_parse_urlencode({
-                    'm3v': 1,
+        def get_h5_urls(media_url, format_id):
+            location = self._download_json(
+                media_url, media_id,
+                'Download JSON metadata for format %s' % format_id, query={
                     'format': 1,
                     'expect': 3,
-                    'rateid': format_id,
-                })
+                    'tss': 'no',
+                })['location']
 
-                nodes_data = self._download_json(
-                    media_url, media_id,
-                    'Download JSON metadata for format %s' % format_id)
+            return {
+                'http': update_url_query(location, {'tss': 'no'}),
+                'hls': update_url_query(location, {'tss': 'ios'}),
+            }
 
-                req = self._request_webpage(
-                    nodes_data['nodelist'][0]['location'], media_id,
-                    note='Downloading m3u8 information for format %s' % format_id)
+        def get_flash_urls(media_url, format_id):
+            media_url += '&' + compat_urllib_parse_urlencode({
+                'm3v': 1,
+                'format': 1,
+                'expect': 3,
+                'rateid': format_id,
+            })
 
-                m3u8_data = self.decrypt_m3u8(req.read())
+            nodes_data = self._download_json(
+                media_url, media_id,
+                'Download JSON metadata for format %s' % format_id)
 
-                url_info_dict = {
-                    'url': encode_data_uri(m3u8_data, 'application/vnd.apple.mpegurl'),
-                    'ext': determine_ext(dispatch[format_id][1]),
-                    'format_id': format_id,
-                    'protocol': 'm3u8',
-                }
+            req = self._request_webpage(
+                nodes_data['nodelist'][0]['location'], media_id,
+                note='Downloading m3u8 information for format %s' % format_id)
 
-                if format_id[-1:] == 'p':
-                    url_info_dict['height'] = int_or_none(format_id[:-1])
+            m3u8_data = self.decrypt_m3u8(req.read())
 
-                urls.append(url_info_dict)
+            return {
+                'hls': encode_data_uri(m3u8_data, 'application/vnd.apple.mpegurl'),
+            }
+
+        extracted_formats = []
+        formats = []
+        for play_json, get_urls in ((play_json_h5, get_h5_urls), (play_json_flash, get_flash_urls)):
+            playurl = play_json['playurl']
+            play_domain = playurl['domain'][0]
+
+            for format_id, format_data in playurl.get('dispatch', []).items():
+                if format_id in extracted_formats:
+                    continue
+                extracted_formats.append(format_id)
+
+                media_url = play_domain + format_data[0]
+                for protocol, format_url in get_urls(media_url, format_id).items():
+                    f = {
+                        'url': format_url,
+                        'ext': determine_ext(format_data[1]),
+                        'format_id': '%s-%s' % (protocol, format_id),
+                        'protocol': 'm3u8_native' if protocol == 'hls' else 'http',
+                        'quality': int_or_none(format_id),
+                    }
+
+                    if format_id[-1:] == 'p':
+                        f['height'] = int_or_none(format_id[:-1])
+
+                    formats.append(f)
+        self._sort_formats(formats, ('height', 'quality', 'format_id'))
 
         publish_time = parse_iso8601(self._html_search_regex(
             r'发布时间&nbsp;([^<>]+) ', page, 'publish time', default=None),
@@ -181,7 +226,7 @@ class LeIE(InfoExtractor):
 
         return {
             'id': media_id,
-            'formats': urls,
+            'formats': formats,
             'title': playurl['title'],
             'thumbnail': playurl['pic'],
             'description': description,