[polskieradio] Improve extraction (Closes #9813)

diff --git a/youtube_dl/extractor/polskieradio.py b/youtube_dl/extractor/polskieradio.py
index 5d4b116d8ceac39867269b869456e1f6851b8e3a..f5adff08ffd91bc01bffd57be3814e75415d553f 100644 (file)
--- a/youtube_dl/extractor/polskieradio.py
+++ b/youtube_dl/extractor/polskieradio.py
@@ -1,74 +1,91 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-from .common import InfoExtractor
-from ..utils import int_or_none
+import re
 
-import calendar
-from datetime import datetime
+from .common import InfoExtractor
+from ..compat import (
+    compat_str,
+    compat_urllib_parse_unquote,
+)
+from ..utils import (
+    int_or_none,
+    strip_or_none,
+    unified_timestamp,
+)
 
 
 class PolskieRadioIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?polskieradio\.pl/[0-9]+/[0-9]+/Artykul/(?P<id>[0-9]+),.+'
+    _VALID_URL = r'https?://(?:www\.)?polskieradio\.pl/\d+/\d+/Artykul/(?P<id>[0-9]+)'
     _TESTS = [{
         'url': 'http://www.polskieradio.pl/7/5102/Artykul/1587943,Prof-Andrzej-Nowak-o-historii-nie-da-sie-myslec-beznamietnie',
-        'md5': '2984ee6ce9046d91fc233bc1a864a09a',
         'info_dict': {
             'id': '1587943',
-            'ext': 'mp3',
             'title': 'Prof. Andrzej Nowak: o historii nie da się myśleć beznamiętnie',
             'description': 'md5:12f954edbf3120c5e7075e17bf9fc5c5',
-            'release_date': '20160227',
-            'upload_date': '20160227',
-            'timestamp': 1456594200,
-            'duration': 2364
-        }
+        },
+        'playlist': [{
+            'md5': '2984ee6ce9046d91fc233bc1a864a09a',
+            'info_dict': {
+                'id': '1540576',
+                'ext': 'mp3',
+                'title': 'md5:d4623290d4ac983bf924061c75c23a0d',
+                'timestamp': 1456594200,
+                'upload_date': '20160227',
+                'duration': 2364,
+            },
+        }],
     }, {
-        'url': 'http://polskieradio.pl/9/305/Artykul/1632955,Bardzo-popularne-slowo-remis',
-        'md5': '68a393e25b942c1a76872f56d303a31a',
+        'url': 'http://www.polskieradio.pl/265/5217/Artykul/1635803,Euro-2016-nie-ma-miejsca-na-blad-Polacy-graja-ze-Szwajcaria-o-cwiercfinal',
         'info_dict': {
-            'id': '1632955',
-            'ext': 'mp3',
-            'title': 'Bardzo popularne słowo: remis',
-            'description': 'md5:3b58dfae614100abc0f175a0b26d5680',
-            'release_date': '20160617',
-            'upload_date': '20160617',
-            'timestamp': 1466184900,
-            'duration': 393
-        }
+            'id': '1635803',
+            'title': 'Euro 2016: nie ma miejsca na błąd. Polacy grają ze Szwajcarią o ćwierćfinał',
+            'description': 'md5:01cb7d0cad58664095d72b51a1ebada2',
+        },
+        'playlist_mincount': 12,
+    }, {
+        'url': 'http://polskieradio.pl/9/305/Artykul/1632955,Bardzo-popularne-slowo-remis',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.polskieradio.pl/7/5102/Artykul/1587943',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
-        metadata_string = self._html_search_regex(r'<span class="play pr-media-play" data-media=(\{.+\})>', webpage, 'metadata')
-        metadata = self._parse_json(metadata_string, video_id)
+        playlist_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, playlist_id)
+
+        content = self._search_regex(
+            r'(?s)<div[^>]+class="audio atarticle"[^>]*>(.+?)<script>',
+            webpage, 'content')
+
+        timestamp = unified_timestamp(self._html_search_regex(
+            r'(?s)<span[^>]+id="datetime2"[^>]*>(.+?)</span>',
+            webpage, 'timestamp', fatal=False))
 
-        title = self._og_search_title(webpage)
-        if title is not None:
-            title = title.strip()
+        entries = []
 
-        description = self._og_search_description(webpage)
-        if description is not None:
-            description = description.strip()
+        media_urls = set()
 
-        release_date = self._html_search_regex(r'Data emisji:[^0-9]+([0-9]{1,2}\.[0-9]{2}\.[0-9]{4})', webpage, 'release date', fatal=False)
-        if release_date is not None:
-            release_date = datetime.strptime(release_date, '%d.%m.%Y').strftime('%Y%m%d')
+        for data_media in re.findall(r'<[^>]+data-media=({[^>]+})', content):
+            media = self._parse_json(data_media, playlist_id, fatal=False)
+            if not media.get('file') or not media.get('desc'):
+                continue
+            media_url = self._proto_relative_url(media['file'], 'http:')
+            if media_url in media_urls:
+                continue
+            media_urls.add(media_url)
+            entries.append({
+                'id': compat_str(media['id']),
+                'url': media_url,
+                'title': compat_urllib_parse_unquote(media['desc']),
+                'duration': int_or_none(media.get('length')),
+                'vcodec': 'none' if media.get('provider') == 'audio' else None,
+                'timestamp': timestamp,
+            })
 
-        upload_datetime = self._html_search_regex(r'<span id="datetime2" class="time">\s+(.+)\s+</span>', webpage, 'release time', fatal=False)
-        if upload_datetime is not None:
-            timestamp = calendar.timegm(datetime.strptime(upload_datetime, '%d.%m.%Y %H:%M').timetuple())
-        else:
-            timestamp = None
+        title = self._og_search_title(webpage).strip()
+        description = strip_or_none(self._og_search_description(webpage))
 
-        return {
-            'id': video_id,
-            'title': title,
-            'description': description,
-            'display_id': metadata.get('id'),
-            'duration': int_or_none(metadata.get('length')),
-            'url': self._proto_relative_url(metadata.get('file'), 'http:'),
-            'release_date': release_date,
-            'timestamp': timestamp
-        }
+        return self.playlist_result(entries, playlist_id, title, description)