[sohu] Modernize and extract all formats and more metadata (Closes #4409, closes #2056, closes #3009)

youtube_dl/extractor/sohu.py

@@ -1,11 +1,10 @@
 # encoding: utf-8
 from __future__ import unicode_literals
 
-import json
 import re
 
 from .common import InfoExtractor
-from ..utils import ExtractorError
+from .common import compat_str
 
 
 class SohuIE(InfoExtractor):
@@ -29,60 +28,73 @@ class SohuIE(InfoExtractor):
                 base_data_url = 'http://my.tv.sohu.com/play/videonew.do?vid='
             else:
                 base_data_url = 'http://hot.vrs.sohu.com/vrs_flash.action?vid='
-            data_url = base_data_url + str(vid_id)
+
-            data_json = self._download_webpage(
+            return self._download_json(
-                data_url, video_id,
+                base_data_url + vid_id, video_id,
-                note='Downloading JSON data for ' + str(vid_id))
+                'Downloading JSON data for %s' % vid_id)
-            return json.loads(data_json)
 
         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>',
+        raw_title = self._html_search_regex(
-                                            webpage, 'video title')
+            r'(?s)<title>(.+?)</title>',
+            webpage, 'video title')
         title = raw_title.partition('-')[0].strip()
 
-        vid = self._html_search_regex(r'var vid ?= ?["\'](\d+)["\']', webpage,
+        vid = self._html_search_regex(
-                                      'video path')
+            r'var vid ?= ?["\'](\d+)["\']',
-        data = _fetch_data(vid, mytv)
+            webpage, 'video path')
+        vid_data = _fetch_data(vid, mytv)
 
-        QUALITIES = ('ori', 'super', 'high', 'nor')
+        formats_json = {}
-        vid_ids = [data['data'][q + 'Vid']
+        for format_id in ('nor', 'high', 'super', 'ori', 'h2644k', 'h2654k'):
-                   for q in QUALITIES
+            vid_id = vid_data['data'].get('%sVid' % format_id)
-                   if data['data'][q + 'Vid'] != 0]
+            if not vid_id:
-        if not vid_ids:
+                continue
-            raise ExtractorError('No formats available for this video')
+            vid_id = compat_str(vid_id)
+            formats_json[format_id] = vid_data if vid == vid_id else _fetch_data(vid_id, mytv)
 
-        # For now, we just pick the highest available quality
+        part_count = vid_data['data']['totalBlocks']
-        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' %
+            formats = []
-                        (allot, prot, clipsURL[i], su[i]))
+            for format_id, format_data in formats_json.items():
-            part_str = self._download_webpage(
+                allot = format_data['allot']
-                part_url, video_id,
+                prot = format_data['prot']
-                note='Downloading part %d of %d' % (i + 1, part_count))
 
-            part_info = part_str.split('|')
+                data = format_data['data']
-            video_url = '%s%s?key=%s' % (part_info[0], su[i], part_info[3])
+                clips_url = data['clipsURL']
+                su = data['su']
 
-            video_info = {
+                part_str = self._download_webpage(
-                'id': '%s_part%02d' % (video_id, i + 1),
+                    'http://%s/?prot=%s&file=%s&new=%s' %
+                    (allot, prot, clips_url[i], su[i]),
+                    video_id,
+                    'Downloading %s video URL part %d of %d'
+                    % (format_id, i + 1, part_count))
+
+                part_info = part_str.split('|')
+                video_url = '%s%s?key=%s' % (part_info[0], su[i], part_info[3])
+
+                formats.append({
+                    'url': video_url,
+                    'format_id': format_id,
+                    'filesize': data['clipsBytes'][i],
+                    'width': data['width'],
+                    'height': data['height'],
+                    'fps': data['fps'],
+                })
+            self._sort_formats(formats)
+
+            playlist.append({
+                'id': '%s_part%d' % (video_id, i + 1),
                 'title': title,
-                'url': video_url,
+                'duration': vid_data['data']['clipsDuration'][i],
-                'ext': 'mp4',
+                'formats': formats,
-            }
+            })
-            playlist.append(video_info)
 
         if len(playlist) == 1:
             info = playlist[0]