]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/sohu.py
debian/control: Mark compliancy with policy 4.1.3 (no changes needed).
[youtubedl] / youtube_dl / extractor / sohu.py
index bebcafb62bea3fe2e180818256475262177f902f..a62ed84f1e34b75a721914d68a101ad2a0bb8983 100644 (file)
@@ -1,24 +1,88 @@
-# encoding: utf-8
+# coding: utf-8
+from __future__ import unicode_literals
 
-import json
 import re
 
 from .common import InfoExtractor
-from ..utils import ExtractorError
+from ..compat import (
+    compat_str,
+    compat_urllib_parse_urlencode,
+)
+from ..utils import (
+    ExtractorError,
+    int_or_none,
+    try_get,
+)
 
 
 class SohuIE(InfoExtractor):
     _VALID_URL = r'https?://(?P<mytv>my\.)?tv\.sohu\.com/.+?/(?(mytv)|n)(?P<id>\d+)\.shtml.*?'
 
-    _TEST = {
-        u'url': u'http://tv.sohu.com/20130724/n382479172.shtml#super',
-        u'file': u'382479172.mp4',
-        u'md5': u'bde8d9a6ffd82c63a1eefaef4eeefec7',
-        u'info_dict': {
-            u'title': u'MV:Far East Movement《The Illest》',
+    # Sohu videos give different MD5 sums on Travis CI and my machine
+    _TESTS = [{
+        'note': 'This video is available only in Mainland China',
+        'url': 'http://tv.sohu.com/20130724/n382479172.shtml#super',
+        'info_dict': {
+            'id': '382479172',
+            'ext': 'mp4',
+            'title': 'MV:Far East Movement《The Illest》',
         },
-        u'skip': u'Only available from China',
-    }
+        'skip': 'On available in China',
+    }, {
+        'url': 'http://tv.sohu.com/20150305/n409385080.shtml',
+        'info_dict': {
+            'id': '409385080',
+            'ext': 'mp4',
+            'title': '《2015湖南卫视羊年元宵晚会》唐嫣《花好月圆》',
+        }
+    }, {
+        'url': 'http://my.tv.sohu.com/us/232799889/78693464.shtml',
+        'info_dict': {
+            'id': '78693464',
+            'ext': 'mp4',
+            'title': '【爱范品】第31期:MWC见不到的奇葩手机',
+        }
+    }, {
+        'note': 'Multipart video',
+        'url': 'http://my.tv.sohu.com/pl/8384802/78910339.shtml',
+        'info_dict': {
+            'id': '78910339',
+            'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
+        },
+        'playlist': [{
+            'info_dict': {
+                'id': '78910339_part1',
+                'ext': 'mp4',
+                'duration': 294,
+                'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
+            }
+        }, {
+            'info_dict': {
+                'id': '78910339_part2',
+                'ext': 'mp4',
+                'duration': 300,
+                'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
+            }
+        }, {
+            'info_dict': {
+                'id': '78910339_part3',
+                'ext': 'mp4',
+                'duration': 150,
+                'title': '【神探苍实战秘籍】第13期 战争之影 赫卡里姆',
+            }
+        }]
+    }, {
+        'note': 'Video with title containing dash',
+        'url': 'http://my.tv.sohu.com/us/249884221/78932792.shtml',
+        'info_dict': {
+            'id': '78932792',
+            'ext': 'mp4',
+            'title': 'youtube-dl testing video',
+        },
+        'params': {
+            'skip_download': True
+        }
+    }]
 
     def _real_extract(self, url):
 
@@ -26,70 +90,113 @@ class SohuIE(InfoExtractor):
             if mytv:
                 base_data_url = 'http://my.tv.sohu.com/play/videonew.do?vid='
             else:
-                base_data_url = u'http://hot.vrs.sohu.com/vrs_flash.action?vid='
-            data_url = base_data_url + str(vid_id)
-            data_json = self._download_webpage(
-                data_url, video_id,
-                note=u'Downloading JSON data for ' + str(vid_id))
-            return json.loads(data_json)
+                base_data_url = 'http://hot.vrs.sohu.com/vrs_flash.action?vid='
+
+            return self._download_json(
+                base_data_url + vid_id, video_id,
+                'Downloading JSON data for %s' % vid_id,
+                headers=self.geo_verification_headers())
 
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('id')
         mytv = mobj.group('mytv') is not None
 
         webpage = self._download_webpage(url, video_id)
-        raw_title = self._html_search_regex(r'(?s)<title>(.+?)</title>',
-                                            webpage, u'video title')
-        title = raw_title.partition('-')[0].strip()
-
-        vid = self._html_search_regex(r'var vid ?= ?["\'](\d+)["\']', webpage,
-                                      u'video path')
-        data = _fetch_data(vid, mytv)
-
-        QUALITIES = ('ori', 'super', 'high', 'nor')
-        vid_ids = [data['data'][q + 'Vid']
-                   for q in QUALITIES
-                   if data['data'][q + 'Vid'] != 0]
-        if not vid_ids:
-            raise ExtractorError(u'No formats available for this video')
-
-        # For now, we just pick the highest available quality
-        vid_id = vid_ids[-1]
-
-        format_data = data if vid == vid_id else _fetch_data(vid_id, mytv)
-        part_count = format_data['data']['totalBlocks']
-        allot = format_data['allot']
-        prot = format_data['prot']
-        clipsURL = format_data['data']['clipsURL']
-        su = format_data['data']['su']
 
-        playlist = []
-        for i in range(part_count):
-            part_url = ('http://%s/?prot=%s&file=%s&new=%s' %
-                        (allot, prot, clipsURL[i], su[i]))
-            part_str = self._download_webpage(
-                part_url, video_id,
-                note=u'Downloading part %d of %d' % (i+1, part_count))
+        title = re.sub(r' - 搜狐视频$', '', self._og_search_title(webpage))
+
+        vid = self._html_search_regex(
+            r'var vid ?= ?["\'](\d+)["\']',
+            webpage, 'video path')
+        vid_data = _fetch_data(vid, mytv)
+        if vid_data['play'] != 1:
+            if vid_data.get('status') == 12:
+                raise ExtractorError(
+                    '%s said: There\'s something wrong in the video.' % self.IE_NAME,
+                    expected=True)
+            else:
+                self.raise_geo_restricted(
+                    '%s said: The video is only licensed to users in Mainland China.' % self.IE_NAME)
 
-            part_info = part_str.split('|')
-            video_url = '%s%s?key=%s' % (part_info[0], su[i], part_info[3])
+        formats_json = {}
+        for format_id in ('nor', 'high', 'super', 'ori', 'h2644k', 'h2654k'):
+            vid_id = vid_data['data'].get('%sVid' % format_id)
+            if not vid_id:
+                continue
+            vid_id = compat_str(vid_id)
+            formats_json[format_id] = vid_data if vid == vid_id else _fetch_data(vid_id, mytv)
 
-            video_info = {
-                'id': '%s_part%02d' % (video_id, i + 1),
+        part_count = vid_data['data']['totalBlocks']
+
+        playlist = []
+        for i in range(part_count):
+            formats = []
+            for format_id, format_data in formats_json.items():
+                allot = format_data['allot']
+
+                data = format_data['data']
+                clips_url = data['clipsURL']
+                su = data['su']
+
+                video_url = 'newflv.sohu.ccgslb.net'
+                cdnId = None
+                retries = 0
+
+                while 'newflv.sohu.ccgslb.net' in video_url:
+                    params = {
+                        'prot': 9,
+                        'file': clips_url[i],
+                        'new': su[i],
+                        'prod': 'flash',
+                        'rb': 1,
+                    }
+
+                    if cdnId is not None:
+                        params['idc'] = cdnId
+
+                    download_note = 'Downloading %s video URL part %d of %d' % (
+                        format_id, i + 1, part_count)
+
+                    if retries > 0:
+                        download_note += ' (retry #%d)' % retries
+                    part_info = self._parse_json(self._download_webpage(
+                        'http://%s/?%s' % (allot, compat_urllib_parse_urlencode(params)),
+                        video_id, download_note), video_id)
+
+                    video_url = part_info['url']
+                    cdnId = part_info.get('nid')
+
+                    retries += 1
+                    if retries > 5:
+                        raise ExtractorError('Failed to get video URL')
+
+                formats.append({
+                    'url': video_url,
+                    'format_id': format_id,
+                    'filesize': int_or_none(
+                        try_get(data, lambda x: x['clipsBytes'][i])),
+                    'width': int_or_none(data.get('width')),
+                    'height': int_or_none(data.get('height')),
+                    'fps': int_or_none(data.get('fps')),
+                })
+            self._sort_formats(formats)
+
+            playlist.append({
+                'id': '%s_part%d' % (video_id, i + 1),
                 'title': title,
-                'url': video_url,
-                'ext': 'mp4',
-            }
-            playlist.append(video_info)
+                'duration': vid_data['data']['clipsDuration'][i],
+                'formats': formats,
+            })
 
         if len(playlist) == 1:
             info = playlist[0]
             info['id'] = video_id
         else:
             info = {
-                '_type': 'playlist',
+                '_type': 'multi_video',
                 'entries': playlist,
                 'id': video_id,
+                'title': title,
             }
 
         return info