X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Famp.py;h=7ff098cfa0c9789122dab4787f877a1a9bf45352;hb=HEAD;hp=b573b928082b0b9f8f479d976f91a490cc671c1e;hpb=3793090b1b1c1e3462b80dd3045a3573545cfb29;p=youtube-dl

diff --git a/youtube_dl/extractor/amp.py b/youtube_dl/extractor/amp.py
index b573b9280..7ff098cfa 100644
--- a/youtube_dl/extractor/amp.py
+++ b/youtube_dl/extractor/amp.py
@@ -3,82 +3,100 @@ from __future__ import unicode_literals
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
+    ExtractorError,
     int_or_none,
+    mimetype2ext,
     parse_iso8601,
+    url_or_none,
 )
 
 
 class AMPIE(InfoExtractor):
-    def _get_media_node(self, item, name, default=None):
-        media_name = 'media-%s' % name
-        media_group = item.get('media-group') or item
-        return media_group.get(media_name) or item.get(media_name) or item.get(name, default)
-
     # parse Akamai Adaptive Media Player feed
     def _extract_feed_info(self, url):
-        item = self._download_json(
-            url, None,
-            'Downloading Akamai AMP feed',
-            'Unable to download Akamai AMP feed'
-            )['channel']['item']
+        feed = self._download_json(
+            url, None, 'Downloading Akamai AMP feed',
+            'Unable to download Akamai AMP feed')
+        item = feed.get('channel', {}).get('item')
+        if not item:
+            raise ExtractorError('%s said: %s' % (self.IE_NAME, feed['error']))
 
         video_id = item['guid']
-        
+
+        def get_media_node(name, default=None):
+            media_name = 'media-%s' % name
+            media_group = item.get('media-group') or item
+            return media_group.get(media_name) or item.get(media_name) or item.get(name, default)
+
         thumbnails = []
-        media_thumbnail = self._get_media_node(item, 'thumbnail')
+        media_thumbnail = get_media_node('thumbnail')
         if media_thumbnail:
             if isinstance(media_thumbnail, dict):
                 media_thumbnail = [media_thumbnail]
             for thumbnail_data in media_thumbnail:
-                thumbnail = thumbnail_data['@attributes']
+                thumbnail = thumbnail_data.get('@attributes', {})
+                thumbnail_url = url_or_none(thumbnail.get('url'))
+                if not thumbnail_url:
+                    continue
                 thumbnails.append({
-                    'url': self._proto_relative_url(thumbnail['url'], 'http:'),
+                    'url': self._proto_relative_url(thumbnail_url, 'http:'),
                     'width': int_or_none(thumbnail.get('width')),
                     'height': int_or_none(thumbnail.get('height')),
                 })
 
         subtitles = {}
-        media_subtitle = self._get_media_node(item, 'subTitle')
+        media_subtitle = get_media_node('subTitle')
         if media_subtitle:
             if isinstance(media_subtitle, dict):
                 media_subtitle = [media_subtitle]
             for subtitle_data in media_subtitle:
-                subtitle = subtitle_data['@attributes']
-                lang = subtitle.get('lang') or 'en'
-                subtitles[lang] = [{'url': subtitle['href']}]
+                subtitle = subtitle_data.get('@attributes', {})
+                subtitle_href = url_or_none(subtitle.get('href'))
+                if not subtitle_href:
+                    continue
+                subtitles.setdefault(subtitle.get('lang') or 'en', []).append({
+                    'url': subtitle_href,
+                    'ext': mimetype2ext(subtitle.get('type')) or determine_ext(subtitle_href),
+                })
 
         formats = []
-        media_content = self._get_media_node(item, 'content')
+        media_content = get_media_node('content')
         if isinstance(media_content, dict):
             media_content = [media_content]
         for media_data in media_content:
-            media = media_data['@attributes']
-            media_type = media['type']
-            if media_type == 'video/f4m':
-                f4m_formats = self._extract_f4m_formats(media['url'] + '?hdcore=3.4.0&plugin=aasp-3.4.0.132.124', video_id, f4m_id='hds', fatal=False)
-                if f4m_formats:
-                    formats.extend(f4m_formats)
-            elif media_type == 'application/x-mpegURL':
-                m3u8_formats = self._extract_m3u8_formats(media['url'], video_id, m3u8_id='hls', fatal=False)
-                if m3u8_formats:
-                    formats.extend(m3u8_formats)
+            media = media_data.get('@attributes', {})
+            media_url = url_or_none(media.get('url'))
+            if not media_url:
+                continue
+            ext = mimetype2ext(media.get('type')) or determine_ext(media_url)
+            if ext == 'f4m':
+                formats.extend(self._extract_f4m_formats(
+                    media_url + '?hdcore=3.4.0&plugin=aasp-3.4.0.132.124',
+                    video_id, f4m_id='hds', fatal=False))
+            elif ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    media_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
             else:
                 formats.append({
-                    'format_id': media_data['media-category']['@attributes']['label'],
-                    'url': media['url'],
-                    'preference': 1,
-                    'vbr': int_or_none(media.get('bitrate')),
+                    'format_id': media_data.get('media-category', {}).get('@attributes', {}).get('label'),
+                    'url': media_url,
+                    'tbr': int_or_none(media.get('bitrate')),
                     'filesize': int_or_none(media.get('fileSize')),
+                    'ext': ext,
                 })
 
         self._sort_formats(formats)
 
+        timestamp = parse_iso8601(item.get('pubDate'), ' ') or parse_iso8601(item.get('dc-date'))
+
         return {
             'id': video_id,
-            'title': self._get_media_node(item, 'title'),
-            'description': self._get_media_node(item, 'description'),
+            'title': get_media_node('title'),
+            'description': get_media_node('description'),
             'thumbnails': thumbnails,
-            'timestamp': parse_iso8601(item.get('pubDate'), ' '),
+            'timestamp': timestamp,
             'duration': int_or_none(media_content[0].get('@attributes', {}).get('duration')),
+            'subtitles': subtitles,
             'formats': formats,
         }