youtube-dl

commit 8f4a2124a914207912bf9fc37e593210e8dd423b
parent 481888294d19ef52075e531ce26588d97b3d16d2
Author: Yen Chi Hsuan <yan12125@gmail.com>
Date:   Fri, 26 Feb 2016 14:26:07 +0800

[vidzi] Fix extraction

Diffstat:
M
youtube_dl/extractor/jwplatform.py
 | 
6
+++---
M
youtube_dl/extractor/vidzi.py
 | 
40
++++++++++++++++++++++++++++------------

2 files changed, 31 insertions(+), 15 deletions(-)
diff --git a/youtube_dl/extractor/jwplatform.py b/youtube_dl/extractor/jwplatform.py
@@ -8,7 +8,7 @@ from ..utils import int_or_none
 
 
 class JWPlatformBaseIE(InfoExtractor):
-    def _parse_jwplayer_data(self, jwplayer_data, video_id):
+    def _parse_jwplayer_data(self, jwplayer_data, video_id, require_title=True):
         video_data = jwplayer_data['playlist'][0]
         subtitles = {}
         for track in video_data['tracks']:
@@ -19,7 +19,7 @@ class JWPlatformBaseIE(InfoExtractor):
         for source in video_data['sources']:
             source_url = self._proto_relative_url(source['file'])
             source_type = source.get('type') or ''
-            if source_type == 'application/vnd.apple.mpegurl':
+            if source_type in ('application/vnd.apple.mpegurl', 'hls'):
                 formats.extend(self._extract_m3u8_formats(
                     source_url, video_id, 'mp4', 'm3u8_native', fatal=False))
             elif source_type.startswith('audio'):
@@ -37,7 +37,7 @@ class JWPlatformBaseIE(InfoExtractor):
 
         return {
             'id': video_id,
-            'title': video_data['title'],
+            'title': video_data['title'] if require_title else video_data.get('title'),
             'description': video_data.get('description'),
             'thumbnail': self._proto_relative_url(video_data.get('image')),
             'timestamp': int_or_none(video_data.get('pubdate')),
diff --git a/youtube_dl/extractor/vidzi.py b/youtube_dl/extractor/vidzi.py
@@ -1,11 +1,16 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-from .common import InfoExtractor
-from ..utils import smuggle_url
+import re
 
+from .jwplatform import JWPlatformBaseIE
+from ..utils import (
+    base36,
+    js_to_json,
+)
 
-class VidziIE(InfoExtractor):
+
+class VidziIE(JWPlatformBaseIE):
     _VALID_URL = r'https?://(?:www\.)?vidzi\.tv/(?P<id>\w+)'
     _TEST = {
         'url': 'http://vidzi.tv/cghql9yq6emu.html',
@@ -14,7 +19,6 @@ class VidziIE(InfoExtractor):
             'id': 'cghql9yq6emu',
             'ext': 'mp4',
             'title': 'youtube-dl test video  1\\\\2\'3/4<5\\\\6ä7↭',
-            'uploader': 'vidzi.tv',
         },
         'params': {
             # m3u8 download
@@ -29,11 +33,23 @@ class VidziIE(InfoExtractor):
         title = self._html_search_regex(
             r'(?s)<h2 class="video-title">(.*?)</h2>', webpage, 'title')
 
-        # Vidzi now uses jwplayer, which can be handled by GenericIE
-        return {
-            '_type': 'url_transparent',
-            'id': video_id,
-            'title': title,
-            'url': smuggle_url(url, {'to_generic': True}),
-            'ie_key': 'Generic',
-        }
+        mobj = re.search(r"}\('(.+)',36,(\d+),'([^']+)'\.split\('\|'\)", webpage)
+        code, count, symbols = mobj.groups()
+
+        count = int(count)
+        symbols = symbols.split('|')
+
+        while count:
+            count -= 1
+            if symbols[count]:
+                code = re.sub(r'\b%s\b' % base36(count), symbols[count], code)
+
+        code = code.replace('\\\'', '\'')
+        jwplayer_data = self._parse_json(
+            self._search_regex(r'setup\(([^)]+)\)', code, 'jwplayer data'),
+            video_id, transform_source=js_to_json)
+
+        info_dict = self._parse_jwplayer_data(jwplayer_data, video_id, require_title=False)
+        info_dict['title'] = title
+
+        return info_dict