Merge branch 'lecture2go' of https://github.com/nichdu/youtube-dl into nichdu-lecture2go
[youtube-dl] / youtube_dl / extractor / rtlnl.py
index a3ca79f2ccfd2e00c09a4f9b2a9503fa85669b65..e0c530d64a97500eaa1b653599f92c9420518c77 100644 (file)
@@ -1,16 +1,25 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
-from ..utils import parse_duration
+from ..utils import (
+    int_or_none,
+    parse_duration,
+)
 
 
-class RtlXlIE(InfoExtractor):
-    IE_NAME = 'rtlxl.nl'
-    _VALID_URL = r'https?://(www\.)?rtlxl\.nl/#!/[^/]+/(?P<uuid>[^/?]+)'
+class RtlNlIE(InfoExtractor):
+    IE_NAME = 'rtl.nl'
+    IE_DESC = 'rtl.nl and rtlxl.nl'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?
+        (?:
+            rtlxl\.nl/\#!/[^/]+/|
+            rtl\.nl/system/videoplayer/(?:[^/]+/)+(?:video_)?embed\.html\b.+?\buuid=
+        )
+        (?P<id>[0-9a-f-]+)'''
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.rtlxl.nl/#!/rtl-nieuws-132237/6e4203a6-0a5e-3596-8424-c599a59e0677',
         'md5': 'cc16baa36a6c169391f0764fa6b16654',
         'info_dict': {
@@ -22,25 +31,63 @@ class RtlXlIE(InfoExtractor):
             'upload_date': '20140814',
             'duration': 576.880,
         },
-    }
+    }, {
+        'url': 'http://www.rtl.nl/system/videoplayer/derden/rtlnieuws/video_embed.html#uuid=84ae5571-ac25-4225-ae0c-ef8d9efb2aed/autoplay=false',
+        'md5': 'dea7474214af1271d91ef332fb8be7ea',
+        'info_dict': {
+            'id': '84ae5571-ac25-4225-ae0c-ef8d9efb2aed',
+            'ext': 'mp4',
+            'timestamp': 1424039400,
+            'title': 'RTL Nieuws - Nieuwe beelden Kopenhagen: chaos direct na aanslag',
+            'thumbnail': 're:^https?://screenshots\.rtl\.nl/system/thumb/sz=[0-9]+x[0-9]+/uuid=84ae5571-ac25-4225-ae0c-ef8d9efb2aed$',
+            'upload_date': '20150215',
+            'description': 'Er zijn nieuwe beelden vrijgegeven die vlak na de aanslag in Kopenhagen zijn gemaakt. Op de video is goed te zien hoe omstanders zich bekommeren om één van de slachtoffers, terwijl de eerste agenten ter plaatse komen.',
+        }
+    }, {
+        # empty synopsis and missing episodes (see https://github.com/rg3/youtube-dl/issues/6275)
+        'url': 'http://www.rtl.nl/system/videoplayer/derden/rtlnieuws/video_embed.html#uuid=f536aac0-1dc3-4314-920e-3bd1c5b3811a/autoplay=false',
+        'info_dict': {
+            'id': 'f536aac0-1dc3-4314-920e-3bd1c5b3811a',
+            'ext': 'mp4',
+            'title': 'RTL Nieuws - Meer beelden van overval juwelier',
+            'thumbnail': 're:^https?://screenshots\.rtl\.nl/system/thumb/sz=[0-9]+x[0-9]+/uuid=f536aac0-1dc3-4314-920e-3bd1c5b3811a$',
+            'timestamp': 1437233400,
+            'upload_date': '20150718',
+            'duration': 30.474,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        # encrypted m3u8 streams, georestricted
+        'url': 'http://www.rtlxl.nl/#!/afl-2-257632/52a74543-c504-4cde-8aa8-ec66fe8d68a7',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.rtl.nl/system/videoplayer/derden/embed.html#!/uuid=bb0353b0-d6a4-1dad-90e9-18fe75b8d1f0',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        uuid = mobj.group('uuid')
-
+        uuid = self._match_id(url)
         info = self._download_json(
-            'http://www.rtl.nl/system/s4m/vfd/version=2/uuid=%s/fmt=flash/' % uuid,
+            'http://www.rtl.nl/system/s4m/vfd/version=2/uuid=%s/fmt=adaptive/' % uuid,
             uuid)
 
         material = info['material'][0]
-        episode_info = info['episodes'][0]
+        title = info['abstracts'][0]['name']
+        subtitle = material.get('title')
+        if subtitle:
+            title += ' - %s' % subtitle
+        description = material.get('synopsis')
 
-        progname = info['abstracts'][0]['name']
-        subtitle = material['title'] or info['episodes'][0]['name']
+        meta = info.get('meta', {})
 
         # Use unencrypted m3u8 streams (See https://github.com/rg3/youtube-dl/issues/4118)
-        videopath = material['videopath'].replace('.f4m', '.m3u8')
-        m3u8_url = 'http://manifest.us.rtl.nl' + videopath
+        # NB: nowadays, recent ffmpeg and avconv can handle these encrypted streams, so
+        # this adaptive -> flash workaround is not required in general, but it also
+        # allows bypassing georestriction therefore is retained for now.
+        videopath = material['videopath'].replace('/adaptive/', '/flash/')
+        m3u8_url = meta.get('videohost', 'http://manifest.us.rtl.nl') + videopath
 
         formats = self._extract_m3u8_formats(m3u8_url, uuid, ext='mp4')
 
@@ -58,14 +105,29 @@ class RtlXlIE(InfoExtractor):
                 'quality': 0,
             }
         ])
-
         self._sort_formats(formats)
 
+        thumbnails = []
+
+        for p in ('poster_base_url', '"thumb_base_url"'):
+            if not meta.get(p):
+                continue
+
+            thumbnails.append({
+                'url': self._proto_relative_url(meta[p] + uuid),
+                'width': int_or_none(self._search_regex(
+                    r'/sz=([0-9]+)', meta[p], 'thumbnail width', fatal=False)),
+                'height': int_or_none(self._search_regex(
+                    r'/sz=[0-9]+x([0-9]+)',
+                    meta[p], 'thumbnail height', fatal=False))
+            })
+
         return {
             'id': uuid,
-            'title': '%s - %s' % (progname, subtitle),
+            'title': title,
             'formats': formats,
             'timestamp': material['original_date'],
-            'description': episode_info['synopsis'],
+            'description': description,
             'duration': parse_duration(material.get('duration')),
+            'thumbnails': thumbnails,
         }