[ccc] improve extraction

youtube_dl/extractor/ccc.py

@@ -1,13 +1,9 @@
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 from ..utils import (
     int_or_none,
-    parse_duration,
+    parse_iso8601,
-    qualities,
-    unified_strdate,
 )
 
 
@@ -19,14 +15,14 @@ class CCCIE(InfoExtractor):
         'url': 'https://media.ccc.de/v/30C3_-_5443_-_en_-_saal_g_-_201312281830_-_introduction_to_processor_design_-_byterazor#video',
         'md5': '3a1eda8f3a29515d27f5adb967d7e740',
         'info_dict': {
-            'id': '30C3_-_5443_-_en_-_saal_g_-_201312281830_-_introduction_to_processor_design_-_byterazor',
+            'id': '1839',
             'ext': 'mp4',
             'title': 'Introduction to Processor Design',
-            'description': 'md5:80be298773966f66d56cb11260b879af',
+            'description': 'md5:df55f6d073d4ceae55aae6f2fd98a0ac',
             'thumbnail': 're:^https?://.*\.jpg$',
-            'view_count': int,
             'upload_date': '20131228',
-            'duration': 3660,
+            'timestamp': 1388188800,
+            'duration': 3710,
         }
     }, {
         'url': 'https://media.ccc.de/v/32c3-7368-shopshifting#download',
@@ -34,79 +30,48 @@ class CCCIE(InfoExtractor):
     }]
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        display_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage(url, display_id)
-
+        event_id = self._search_regex("data-id='(\d+)'", webpage, 'event id')
-        if self._downloader.params.get('prefer_free_formats'):
+        event_data = self._download_json('https://media.ccc.de/public/events/%s' % event_id, event_id)
-            preference = qualities(['mp3', 'opus', 'mp4-lq', 'webm-lq', 'h264-sd', 'mp4-sd', 'webm-sd', 'mp4', 'webm', 'mp4-hd', 'h264-hd', 'webm-hd'])
-        else:
-            preference = qualities(['opus', 'mp3', 'webm-lq', 'mp4-lq', 'webm-sd', 'h264-sd', 'mp4-sd', 'webm', 'mp4', 'webm-hd', 'mp4-hd', 'h264-hd'])
-
-        title = self._html_search_regex(
-            r'(?s)<h1>(.*?)</h1>', webpage, 'title')
-        description = self._html_search_regex(
-            r'(?s)<h3>About</h3>(.+?)<h3>',
-            webpage, 'description', fatal=False)
-        upload_date = unified_strdate(self._html_search_regex(
-            r"(?s)<span[^>]+class='[^']*fa-calendar-o'[^>]*>(.+?)</span>",
-            webpage, 'upload date', fatal=False))
-        view_count = int_or_none(self._html_search_regex(
-            r"(?s)<span class='[^']*fa-eye'></span>(.*?)</li>",
-            webpage, 'view count', fatal=False))
-        duration = parse_duration(self._html_search_regex(
-            r'(?s)<span[^>]+class=(["\']).*?fa-clock-o.*?\1[^>]*></span>(?P<duration>.+?)</li',
-            webpage, 'duration', fatal=False, group='duration'))
 
-        matches = re.finditer(r'''(?xs)
-            <(?:span|div)\s+class='label\s+filetype'>(?P<format>[^<]*)</(?:span|div)>\s*
-            <(?:span|div)\s+class='label\s+filetype'>(?P<lang>[^<]*)</(?:span|div)>\s*
-            <a\s+download\s+href='(?P<http_url>[^']+)'>\s*
-            (?:
-                .*?
-                <a\s+(?:download\s+)?href='(?P<torrent_url>[^']+\.torrent)'
-            )?''', webpage)
         formats = []
-        for m in matches:
+        for recording in event_data.get('recordings', []):
-            format = m.group('format')
+            recording_url = recording.get('recording_url')
-            format_id = self._search_regex(
+            if not recording_url:
-                r'.*/([a-z0-9_-]+)/[^/]*$',
+                continue
-                m.group('http_url'), 'format id', default=None)
+            language = recording.get('language')
-            if format_id:
+            folder = recording.get('folder')
-                format_id = m.group('lang') + '-' + format_id
+            format_id = None
-            vcodec = 'h264' if 'h264' in format_id else (
+            if language:
-                'none' if format_id in ('mp3', 'opus') else None
+                format_id = language
+            if folder:
+                if language:
+                    format_id += '-' + folder
+                else:
+                    format_id = folder
+            vcodec = 'h264' if 'h264' in folder else (
+                'none' if folder in ('mp3', 'opus') else None
             )
             formats.append({
                 'format_id': format_id,
-                'format': format,
+                'url': recording_url,
-                'language': m.group('lang'),
+                'width': int_or_none(recording.get('width')),
-                'url': m.group('http_url'),
+                'height': int_or_none(recording.get('height')),
+                'filesize': int_or_none(recording.get('size'), invscale=1024 * 1024),
+                'language': language,
                 'vcodec': vcodec,
-                'preference': preference(format_id),
             })
-
-            if m.group('torrent_url'):
-                formats.append({
-                    'format_id': 'torrent-%s' % (format if format_id is None else format_id),
-                    'format': '%s (torrent)' % format,
-                    'proto': 'torrent',
-                    'format_note': '(unsupported; will just download the .torrent file)',
-                    'vcodec': vcodec,
-                    'preference': -100 + preference(format_id),
-                    'url': m.group('torrent_url'),
-                })
         self._sort_formats(formats)
 
-        thumbnail = self._html_search_regex(
-            r"<video.*?poster='([^']+)'", webpage, 'thumbnail', fatal=False)
-
         return {
-            'id': video_id,
+            'id': event_id,
-            'title': title,
+            'display_id': display_id,
-            'description': description,
+            'title': event_data['title'],
-            'thumbnail': thumbnail,
+            'description': event_data.get('description'),
-            'view_count': view_count,
+            'thumbnail': event_data.get('thumb_url'),
-            'upload_date': upload_date,
+            'timestamp': parse_iso8601(event_data.get('date')),
-            'duration': duration,
+            'duration': int_or_none(event_data.get('length')),
+            'tags': event_data.get('tags'),
             'formats': formats,
         }