add an extractor for tv.sohu.com

2 files changed, 98 insertions, 0 deletions

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index c20172a53..3a08d676f 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -55,6 +55,7 @@ from .redtube import RedTubeIE
 from .ringtv import RingTVIE
 from .roxwel import RoxwelIE
 from .sina import SinaIE
+from .sohu import SohuIE
 from .soundcloud import SoundcloudIE, SoundcloudSetIE
 from .spiegel import SpiegelIE
 from .stanfordoc import StanfordOpenClassroomIE
diff --git a/youtube_dl/extractor/sohu.py b/youtube_dl/extractor/sohu.py
new file mode 100644
index 000000000..830814221
--- /dev/null
+++ b/youtube_dl/extractor/sohu.py
@@ -0,0 +1,97 @@
+# encoding: utf-8
+
+import re
+import json
+import time
+import logging
+import urllib2
+
+from .common import InfoExtractor
+from ..utils import compat_urllib_request
+
+
+class SohuIE(InfoExtractor):
+    _VALID_URL = r'https?://tv\.sohu\.com/\d+?/n(?P<id>\d+)\.shtml.*?'
+
+    _TEST = {
+        u'url': u'http://tv.sohu.com/20130724/n382479172.shtml#super',
+        u'file': u'382479172.flv',
+        u'md5': u'cc84eed6b6fbf0f2f9a8d3cb9da1939b',
+        u'info_dict': {
+            u'title': u'The Illest - Far East Movement Riff Raff',
+        },
+    }
+
+    def _clearn_html(self, string):
+        tags = re.findall(r'<.+?>', string)
+        for t in tags:
+            string = string.replace(t, ' ')
+        for i in range(2):
+            spaces = re.findall(r'\s+', string)
+            for s in spaces:
+                string = string.replace(s, ' ')
+        string = string.strip()
+        return string
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+        webpage = self._download_webpage(url, video_id)
+        pattern = r'<h1 id="video-title">\n*?(.+?)\n*?</h1>'
+        compiled = re.compile(pattern, re.DOTALL)
+        title = self._search_regex(compiled, webpage, u'video title').strip('\t\n')
+        title = self._clearn_html(title)
+        pattern = re.compile(r'var vid="(\d+)"')
+        result = re.search(pattern, webpage)
+        if not result:
+            logging.info('[Sohu] could not get vid')
+            return None
+        vid = result.group(1)
+        logging.info('vid: %s' % vid)
+        base_url_1 = 'http://hot.vrs.sohu.com/vrs_flash.action?vid='
+        url_1 = base_url_1 + vid
+        logging.info('json url: %s' % url_1)
+        json_1 = json.loads(urllib2.urlopen(url_1).read())
+        # get the highest definition video vid and json infomation.
+        vids = []
+        qualities = ('oriVid', 'superVid', 'highVid', 'norVid')
+        for vid_name in qualities:
+            vids.append(json_1['data'][vid_name])
+        clearest_vid = 0
+        for i, v in enumerate(vids):
+            if v != 0:
+                clearest_vid = v
+                logging.info('quality definition: %s' % qualities[i][:-3])
+                break
+        if not clearest_vid:
+            logging.warning('could not find valid clearest_vid')
+            return None
+        if vid != clearest_vid:
+            url_1 = '%s%d' % (base_url_1, clearest_vid)
+            logging.info('highest definition json url: %s' % url_1)
+            json_1 = json.loads(urllib2.urlopen(url_1).read())
+        allot = json_1['allot']
+        prot = json_1['prot']
+        clipsURL = json_1['data']['clipsURL']
+        su = json_1['data']['su']
+        num_of_parts = json_1['data']['totalBlocks']
+        logging.info('Total parts: %d' % num_of_parts)
+        base_url_3 = 'http://allot/?prot=prot&file=clipsURL[i]&new=su[i]'
+        files_info = []
+        for i in range(num_of_parts):
+            middle_url = 'http://%s/?prot=%s&file=%s&new=%s' % (allot, prot, clipsURL[i], su[i])
+            logging.info('middle url part %d: %s' % (i, middle_url))
+            middle_info = urllib2.urlopen(middle_url).read().split('|')
+            middle_part_1 = middle_info[0]
+            download_url = '%s%s?key=%s' % (middle_info[0], su[i], middle_info[3])
+
+            info = {
+                'id': '%s_part%02d' % (video_id, i + 1),
+                'title': title,
+                'url': download_url,
+                'ext': 'mp4',
+            }
+            files_info.append(info)
+            time.sleep(1)
+
+        return files_info