[zdf:channel] Simplify (#4427)

youtube_dl/extractor/zdf.py

@@ -1,12 +1,14 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import functools
 import re
 
 from .common import InfoExtractor
 from ..utils import (
     int_or_none,
     unified_strdate,
+    OnDemandPagedList,
 )
 
 
@@ -86,28 +88,8 @@ def extract_from_xml_url(ie, video_id, xml_url):
     }
 
 
-def extract_channel_from_xml_url(ie, channel_id, xml_url):
-    doc = ie._download_xml(
-        xml_url, channel_id,
-        note='Downloading channel info',
-        errnote='Failed to download channel info')
-
-    title = doc.find('.//information/title').text
-    description = doc.find('.//information/detail').text
-    assets = [{'id': asset.find('./details/assetId').text,
-               'type': asset.find('./type').text,
-               } for asset in doc.findall('.//teasers/teaser')]
-
-    return {
-        'id': channel_id,
-        'title': title,
-        'description': description,
-        'assets': assets,
-    }
-
-
 class ZDFIE(InfoExtractor):
-    _VALID_URL = r'^https?://www\.zdf\.de/ZDFmediathek(?P<hash>#)?/(.*beitrag/(?:video/)?)(?P<id>[0-9]+)(?:/[^/?]+)?(?:\?.*)?'
+    _VALID_URL = r'(?:zdf:|zdf:video:|https?://www\.zdf\.de/ZDFmediathek(?:#)?/(.*beitrag/(?:video/)?))(?P<id>[0-9]+)(?:/[^/?]+)?(?:\?.*)?'
 
     _TEST = {
         'url': 'http://www.zdf.de/ZDFmediathek/beitrag/video/2037704/ZDFspezial---Ende-des-Machtpokers--?bc=sts;stt',
@@ -124,67 +106,54 @@ class ZDFIE(InfoExtractor):
         'skip': 'Videos on ZDF.de are depublicised in short order',
     }
 
-    def _extract_video(self, video_id):
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
         xml_url = 'http://www.zdf.de/ZDFmediathek/xmlservice/web/beitragsDetails?ak=web&id=%s' % video_id
         return extract_from_xml_url(self, video_id, xml_url)
 
-    def _real_extract(self, url):
-        return self._extract_video(self._match_id(url))
-
-
-class ZDFChannelIE(ZDFIE):
-    _VALID_URL = r'^https?://www\.zdf\.de/ZDFmediathek(?P<hash>#)?/(.*kanaluebersicht/)(?P<id>[0-9]+)'
 
+class ZDFChannelIE(InfoExtractor):
+    _VALID_URL = r'(?:zdf:topic:|https?://www\.zdf\.de/ZDFmediathek(?:#)?/.*kanaluebersicht/)(?P<id>[0-9]+)'
     _TEST = {
         'url': 'http://www.zdf.de/ZDFmediathek#/kanaluebersicht/1586442/sendung/Titanic',
         'info_dict': {
             'id': '1586442',
-            'title': 'Titanic',
-            'description': 'md5:444c048cfe3fdc2561be7de4bcbf1d04',
         },
-        'playlist_count': 3,
+        'playlist_count': 4,
     }
+    _PAGE_SIZE = 50
 
-    def _extract_channel(self, channel_id):
+    def _fetch_page(self, channel_id, page):
-        def load_chunks(channel_id, chunk_length):
+        offset = page * self._PAGE_SIZE
-            offset = 0
+        xml_url = (
-            while True:
+            'http://www.zdf.de/ZDFmediathek/xmlservice/web/aktuellste?ak=web&offset=%d&maxLength=%d&id=%s'
-                url = ('http://www.zdf.de/ZDFmediathek/xmlservice/web/aktuellste?ak=web&offset=%d&maxLength=%d&id=%s'
+            % (offset, self._PAGE_SIZE, channel_id))
-                       % (offset, chunk_length, channel_id))
+        doc = self._download_xml(
-                result = extract_channel_from_xml_url(self, channel_id, url)
+            xml_url, channel_id,
-                yield result
+            note='Downloading channel info',
-                if len(result['assets']) < chunk_length:
+            errnote='Failed to download channel info')
-                    return
-                offset += chunk_length
 
-        def load_channel(channel_id):
+        title = doc.find('.//information/title').text
-            chunks = list(load_chunks(channel_id, 50))  # The server rejects higher values
+        description = doc.find('.//information/detail').text
-            assets = [asset for chunk in chunks for asset in chunk['assets']]
+        for asset in doc.findall('.//teasers/teaser'):
-            video_ids = [asset['id'] for asset in
+            a_type = asset.find('./type').text
-                         filter(lambda asset: asset['type'] == 'video',
+            a_id = asset.find('./details/assetId').text
-                                assets)]
+            if a_type not in ('video', 'topic'):
-            topic_ids = [asset['id'] for asset in
+                continue
-                         filter(lambda asset: asset['type'] == 'thema',
+            yield {
-                                assets)]
+                '_type': 'url',
-            if topic_ids:
+                'playlist_title': title,
-                video_ids = reduce(list.__add__,
+                'playlist_description': description,
-                                   [load_channel(topic_id)['video_ids']
+                'url': 'zdf:%s:%s' % (a_type, a_id),
-                                    for topic_id in topic_ids],
+            }
-                                   video_ids)
-
-            result = chunks[0]
-            result['video_ids'] = video_ids
-            return result
-
-        channel = load_channel(channel_id)
-        return {
-            '_type': 'playlist',
-            'id': channel['id'],
-            'title': channel['title'],
-            'description': channel['description'],
-            'entries': [self._extract_video(video_id)
-                        for video_id in channel['video_ids']],
-        }
 
     def _real_extract(self, url):
-        return self._extract_channel(self._match_id(url))
+        channel_id = self._match_id(url)
+        entries = OnDemandPagedList(
+            functools.partial(self._fetch_page, channel_id), self._PAGE_SIZE)
+
+        return {
+            '_type': 'playlist',
+            'id': channel_id,
+            'entries': entries,
+        }