[wsj] improve extraction

youtube_dl/extractor/wsj.py

@@ -4,16 +4,22 @@
 from .common import InfoExtractor
 from ..utils import (
     int_or_none,
+    float_or_none,
     unified_strdate,
 )
 
 
 class WSJIE(InfoExtractor):
-    _VALID_URL = r'https?://video-api\.wsj\.com/api-video/player/iframe\.html\?guid=(?P<id>[a-zA-Z0-9-]+)'
+    _VALID_URL = r'''(?x)https?://
+        (?:
+            video-api\.wsj\.com/api-video/player/iframe\.html\?guid=|
+            (?:www\.)?wsj\.com/video/[^/]+/
+        )
+        (?P<id>[a-zA-Z0-9-]+)'''
     IE_DESC = 'Wall Street Journal'
-    _TEST = {
+    _TESTS = [{
         'url': 'http://video-api.wsj.com/api-video/player/iframe.html?guid=1BD01A4C-BFE8-40A5-A42F-8A8AF9898B1A',
-        'md5': '9747d7a6ebc2f4df64b981e1dde9efa9',
+        'md5': 'e230a5bb249075e40793b655a54a02e4',
         'info_dict': {
             'id': '1BD01A4C-BFE8-40A5-A42F-8A8AF9898B1A',
             'ext': 'mp4',
@@ -24,65 +30,60 @@ class WSJIE(InfoExtractor):
             'duration': 90,
             'title': 'Bills Coach Rex Ryan Updates His Old Jets Tattoo',
         },
-    }
+    }, {
+        'url': 'http://www.wsj.com/video/can-alphabet-build-a-smarter-city/359DDAA8-9AC1-489C-82E6-0429C1E430E0.html',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        bitrates = [128, 174, 264, 320, 464, 664, 1264]
         api_url = (
             'http://video-api.wsj.com/api-video/find_all_videos.asp?'
-            'type=guid&count=1&query=%s&'
+            'type=guid&count=1&query=%s&fields=type,hls,videoMP4List,'
-            'fields=hls,adZone,thumbnailList,guid,state,secondsUntilStartTime,'
+            'thumbnailList,author,description,name,duration,videoURL,'
-            'author,description,name,linkURL,videoStillURL,duration,videoURL,'
+            'titletag,formattedCreationDate,keywords,editor' % video_id)
-            'adCategory,catastrophic,linkShortURL,doctypeID,youtubeID,'
-            'titletag,rssURL,wsj-section,wsj-subsection,allthingsd-section,'
-            'allthingsd-subsection,sm-section,sm-subsection,provider,'
-            'formattedCreationDate,keywords,keywordsOmniture,column,editor,'
-            'emailURL,emailPartnerID,showName,omnitureProgramName,'
-            'omnitureVideoFormat,linkRelativeURL,touchCastID,'
-            'omniturePublishDate,%s') % (
-                video_id, ','.join('video%dkMP4Url' % br for br in bitrates))
         info = self._download_json(api_url, video_id)['items'][0]
-
-        # Thumbnails are conveniently in the correct format already
-        thumbnails = info.get('thumbnailList')
-        creator = info.get('author')
-        uploader_id = info.get('editor')
-        categories = info.get('keywords')
-        duration = int_or_none(info.get('duration'))
-        upload_date = unified_strdate(
-            info.get('formattedCreationDate'), day_first=False)
         title = info.get('name', info.get('titletag'))
 
-        formats = [{
+        formats = []
-            'format_id': 'f4m',
+
-            'format_note': 'f4m (meta URL)',
+        f4m_url = info.get('videoURL')
-            'url': info['videoURL'],
+        if f4m_url:
-        }]
+            formats.extend(self._extract_f4m_formats(
-        if info.get('hls'):
+                f4m_url, video_id, f4m_id='hds', fatal=False))
+
+        m3u8_url = info.get('hls')
+        if m3u8_url:
             formats.extend(self._extract_m3u8_formats(
                 info['hls'], video_id, ext='mp4',
-                preference=0, entry_protocol='m3u8_native'))
+                entry_protocol='m3u8_native', m3u8_id='hls', fatal=False))
-        for br in bitrates:
+
-            field = 'video%dkMP4Url' % br
+        for v in info.get('videoMP4List', []):
-            if info.get(field):
+            mp4_url = v.get('url')
-                formats.append({
+            if not mp4_url:
-                    'format_id': 'mp4-%d' % br,
+                continue
-                    'container': 'mp4',
+            tbr = int_or_none(v.get('bitrate'))
-                    'tbr': br,
+            formats.append({
-                    'url': info[field],
+                'url': mp4_url,
-                })
+                'format_id': 'http' + ('-%d' % tbr if tbr else ''),
+                'tbr': tbr,
+                'width': int_or_none(v.get('width')),
+                'height': int_or_none(v.get('height')),
+                'fps': float_or_none(v.get('fps')),
+            })
         self._sort_formats(formats)
 
         return {
             'id': video_id,
             'formats': formats,
-            'thumbnails': thumbnails,
+            # Thumbnails are conveniently in the correct format already
-            'creator': creator,
+            'thumbnails': info.get('thumbnailList'),
-            'uploader_id': uploader_id,
+            'creator': info.get('author'),
-            'duration': duration,
+            'uploader_id': info.get('editor'),
-            'upload_date': upload_date,
+            'duration': int_or_none(info.get('duration')),
+            'upload_date': unified_strdate(info.get(
+                'formattedCreationDate'), day_first=False),
             'title': title,
-            'categories': categories,
+            'categories': info.get('keywords'),
         }