[youtube] Fix extraction.

[youtube-dl] / youtube_dl / extractor / udn.py

diff --git a/youtube_dl/extractor/udn.py b/youtube_dl/extractor/udn.py
index ee35b7227372c0ddc128dfc694577578f9fc6009..2c8e5c7b41eba5220981a215a566fc8fc565320e 100644 (file)
--- a/youtube_dl/extractor/udn.py
+++ b/youtube_dl/extractor/udn.py
@@ -1,11 +1,13 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import json
+import re
+
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
+    int_or_none,
     js_to_json,
-    ExtractorError,
 )
 from ..compat import compat_urlparse
 
@@ -16,13 +18,17 @@ class UDNEmbedIE(InfoExtractor):
     _VALID_URL = r'https?:' + _PROTOCOL_RELATIVE_VALID_URL
     _TESTS = [{
         'url': 'http://video.udn.com/embed/news/300040',
-        'md5': 'de06b4c90b042c128395a88f0384817e',
         'info_dict': {
             'id': '300040',
             'ext': 'mp4',
             'title': '生物老師男變女 全校挺"做自己"',
-            'thumbnail': 're:^https?://.*\.jpg$',
-        }
+            'thumbnail': r're:^https?://.*\.jpg$',
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+        'expected_warnings': ['Failed to parse JSON Expecting value'],
     }, {
         'url': 'https://video.udn.com/embed/news/300040',
         'only_matching': True,
@@ -37,40 +43,60 @@ class UDNEmbedIE(InfoExtractor):
 
         page = self._download_webpage(url, video_id)
 
-        options = json.loads(js_to_json(self._html_search_regex(
-            r'var options\s*=\s*([^;]+);', page, 'video urls dictionary')))
-
-        video_urls = options['video']
+        options_str = self._html_search_regex(
+            r'var\s+options\s*=\s*([^;]+);', page, 'options')
+        trans_options_str = js_to_json(options_str)
+        options = self._parse_json(trans_options_str, 'options', fatal=False) or {}
+        if options:
+            video_urls = options['video']
+            title = options['title']
+            poster = options.get('poster')
+        else:
+            video_urls = self._parse_json(self._html_search_regex(
+                r'"video"\s*:\s*({.+?})\s*,', trans_options_str, 'video urls'), 'video urls')
+            title = self._html_search_regex(
+                r"title\s*:\s*'(.+?)'\s*,", options_str, 'title')
+            poster = self._html_search_regex(
+                r"poster\s*:\s*'(.+?)'\s*,", options_str, 'poster', default=None)
 
         if video_urls.get('youtube'):
             return self.url_result(video_urls.get('youtube'), 'Youtube')
 
-        try:
-            del video_urls['youtube']
-        except KeyError:
-            pass
+        formats = []
+        for video_type, api_url in video_urls.items():
+            if not api_url:
+                continue
 
-        formats = [{
-            'url': self._download_webpage(
+            video_url = self._download_webpage(
                 compat_urlparse.urljoin(url, api_url), video_id,
-                'retrieve url for %s video' % video_type),
-            'format_id': video_type,
-            'preference': 0 if video_type == 'mp4' else -1,
-        } for video_type, api_url in video_urls.items() if api_url]
+                note='retrieve url for %s video' % video_type)
 
-        if not formats:
-            raise ExtractorError('No videos found', expected=True)
+            ext = determine_ext(video_url)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    video_url, video_id, ext='mp4', m3u8_id='hls'))
+            elif ext == 'f4m':
+                formats.extend(self._extract_f4m_formats(
+                    video_url, video_id, f4m_id='hds'))
+            else:
+                mobj = re.search(r'_(?P<height>\d+)p_(?P<tbr>\d+)\.mp4', video_url)
+                a_format = {
+                    'url': video_url,
+                    # video_type may be 'mp4', which confuses YoutubeDL
+                    'format_id': 'http-' + video_type,
+                }
+                if mobj:
+                    a_format.update({
+                        'height': int_or_none(mobj.group('height')),
+                        'tbr': int_or_none(mobj.group('tbr')),
+                    })
+                formats.append(a_format)
 
         self._sort_formats(formats)
 
-        thumbnail = None
-
-        if options.get('gallery') and len(options['gallery']):
-            thumbnail = options['gallery'][0].get('original')
-
         return {
             'id': video_id,
             'formats': formats,
-            'title': options['title'],
-            'thumbnail': thumbnail
+            'title': title,
+            'thumbnail': poster,
         }