[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/rtp.py b/youtube_dl/extractor/rtp.py
index 7736cabbac6b855be739b8c7521a073ea630f32f..02986f4426dd1ec63f56a3bf26b0ade4b60de15c 100644 (file)
--- a/youtube_dl/extractor/rtp.py
+++ b/youtube_dl/extractor/rtp.py
@@ -1,25 +1,24 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import json
-
 from .common import InfoExtractor
-from ..utils import js_to_json
+from ..utils import (
+    determine_ext,
+    js_to_json,
+)
 
 
 class RTPIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?rtp\.pt/play/p(?P<program_id>[0-9]+)/(?P<id>[^/?#]+)/?'
     _TESTS = [{
         'url': 'http://www.rtp.pt/play/p405/e174042/paixoes-cruzadas',
+        'md5': 'e736ce0c665e459ddb818546220b4ef8',
         'info_dict': {
             'id': 'e174042',
             'ext': 'mp3',
             'title': 'Paixões Cruzadas',
             'description': 'As paixões musicais de António Cartaxo e António Macedo',
-            'thumbnail': 're:^https?://.*\.jpg',
-        },
-        'params': {
-            'skip_download': True,  # RTMP download
+            'thumbnail': r're:^https?://.*\.jpg',
         },
     }, {
         'url': 'http://www.rtp.pt/play/p831/a-quimica-das-coisas',
@@ -32,29 +31,36 @@ class RTPIE(InfoExtractor):
         webpage = self._download_webpage(url, video_id)
         title = self._html_search_meta(
             'twitter:title', webpage, display_name='title', fatal=True)
-        description = self._html_search_meta('description', webpage)
-        thumbnail = self._og_search_thumbnail(webpage)
-
-        player_config = self._search_regex(
-            r'(?s)RTPPLAY\.player\.newPlayer\(\s*(\{.*?\})\s*\)', webpage, 'player config')
-        config = json.loads(js_to_json(player_config))
 
-        path, ext = config.get('file').rsplit('.', 1)
-        formats = [{
-            'app': config.get('application'),
-            'play_path': '{ext:s}:{path:s}'.format(ext=ext, path=path),
-            'page_url': url,
-            'url': 'rtmp://{streamer:s}/{application:s}'.format(**config),
-            'rtmp_live': config.get('live', False),
-            'ext': ext,
-            'vcodec': config.get('type') == 'audio' and 'none' or None,
-            'player_url': 'http://programas.rtp.pt/play/player.swf?v3',
-        }]
+        config = self._parse_json(self._search_regex(
+            r'(?s)RTPPlayer\(({.+?})\);', webpage,
+            'player config'), video_id, js_to_json)
+        file_url = config['file']
+        ext = determine_ext(file_url)
+        if ext == 'm3u8':
+            file_key = config.get('fileKey')
+            formats = self._extract_m3u8_formats(
+                file_url, video_id, 'mp4', 'm3u8_native',
+                m3u8_id='hls', fatal=file_key)
+            if file_key:
+                formats.append({
+                    'url': 'https://cdn-ondemand.rtp.pt' + file_key,
+                    'preference': 1,
+                })
+            self._sort_formats(formats)
+        else:
+            formats = [{
+                'url': file_url,
+                'ext': ext,
+            }]
+        if config.get('mediaType') == 'audio':
+            for f in formats:
+                f['vcodec'] = 'none'
 
         return {
             'id': video_id,
             'title': title,
             'formats': formats,
-            'description': description,
-            'thumbnail': thumbnail,
+            'description': self._html_search_meta(['description', 'twitter:description'], webpage),
+            'thumbnail': config.get('poster') or self._og_search_thumbnail(webpage),
         }