[meipai] Fix regular videos extraction and improve (closes #10718)

youtube_dl/extractor/meipai.py

@@ -1,15 +1,19 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-from ..utils import parse_iso8601
 from .common import InfoExtractor
+from ..utils import (
+    int_or_none,
+    parse_duration,
+    unified_timestamp,
+)
 
 
 class MeipaiIE(InfoExtractor):
     IE_DESC = '美拍'
     _VALID_URL = r'https?://(?:www\.)?meipai.com/media/(?P<id>[0-9]+)'
-    _TESTS = [
-        {
+    _TESTS = [{
+        # regular uploaded video
         'url': 'http://www.meipai.com/media/531697625',
         'md5': 'e3e9600f9e55a302daecc90825854b4f',
         'info_dict': {
@@ -18,27 +22,15 @@ class MeipaiIE(InfoExtractor):
             'title': '#葉子##阿桑##余姿昀##超級女聲#',
             'description': '#葉子##阿桑##余姿昀##超級女聲#',
             'thumbnail': 're:^https?://.*\.jpg$',
+            'duration': 152,
+            'timestamp': 1465492420,
+            'upload_date': '20160609',
+            'view_count': 35511,
             'creator': '她她-TATA',
             'tags': ['葉子', '阿桑', '余姿昀', '超級女聲'],
-                'release_date': 1465492420,
-            }
-        },
-        {
-            'url': 'http://www.meipai.com/media/576409659',
-            'md5': '2e807c16ebe67b8b6b3c8dcacbc32f48',
-            'info_dict': {
-                'id': '576409659',
-                'ext': 'mp4',
-                'title': '#失語者##蔡健雅##吉他彈唱#',
-                'description': '#失語者##蔡健雅##吉他彈唱#',
-                'thumbnail': 're:^https?://.*\.jpg$',
-                'creator': '她她-TATA',
-                'tags': ['失語者', '蔡健雅', '吉他彈唱'],
-                'release_date': 1472534847,
         }
-        },
+    }, {
         # record of live streaming
-        {
         'url': 'http://www.meipai.com/media/585526361',
         'md5': 'ff7d6afdbc6143342408223d4f5fb99a',
         'info_dict': {
@@ -47,53 +39,66 @@ class MeipaiIE(InfoExtractor):
             'title': '姿昀和善願 練歌練琴啦😁😁😁',
             'description': '姿昀和善願 練歌練琴啦😁😁😁',
             'thumbnail': 're:^https?://.*\.jpg$',
+            'duration': 5975,
+            'timestamp': 1474311799,
+            'upload_date': '20160919',
+            'view_count': 1215,
             'creator': '她她-TATA',
-                'release_date': 1474311799,
         }
-        },
-    ]
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 
-        title = self._og_search_title(webpage, default=None)
-        if title is None:
-            # fall back to text used in title
-            title = self._html_search_regex(
-                r'<title[^>]*>(.+)</title>', webpage, 'title')
+        title = self._og_search_title(
+            webpage, default=None) or self._html_search_regex(
+            r'<title[^>]*>([^<]+)</title>', webpage, 'title')
 
-        release_date = self._og_search_property(
-            'video:release_date', webpage, 'release date', fatal=False)
-        release_date = parse_iso8601(release_date)
+        formats = []
+
+        # recorded playback of live streaming
+        m3u8_url = self._html_search_regex(
+            r'file:\s*encodeURIComponent\((["\'])(?P<url>(?:(?!\1).)+)\1\)',
+            webpage, 'm3u8 url', group='url', default=None)
+        if m3u8_url:
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                m3u8_id='hls', fatal=False))
+
+        if not formats:
+            # regular uploaded video
+            video_url = self._search_regex(
+                r'data-video=(["\'])(?P<url>(?:(?!\1).)+)\1', webpage, 'video url',
+                group='url', default=None)
+            if video_url:
+                formats.append({
+                    'url': video_url,
+                    'format_id': 'http',
+                })
+
+        timestamp = unified_timestamp(self._og_search_property(
+            'video:release_date', webpage, 'release date', fatal=False))
 
         tags = self._og_search_property(
             'video:tag', webpage, 'tags', default='').split(',')
 
-        info = {
+        view_count = int_or_none(self._html_search_meta(
+            'interactionCount', webpage, 'view count'))
+        duration = parse_duration(self._html_search_meta(
+            'duration', webpage, 'duration'))
+        creator = self._og_search_property(
+            'video:director', webpage, 'creator', fatal=False)
+
+        return {
             'id': video_id,
             'title': title,
-            'thumbnail': self._og_search_thumbnail(webpage),
             'description': self._og_search_description(webpage),
-            'release_date': release_date,
-            'creator': self._og_search_property(
-                'video:director', webpage, 'creator', fatal=False),
+            'thumbnail': self._og_search_thumbnail(webpage),
+            'duration': duration,
+            'timestamp': timestamp,
+            'view_count': view_count,
+            'creator': creator,
             'tags': tags,
+            'formats': formats,
         }
-
-        keywords = self._html_search_meta(
-            'keywords', webpage, 'keywords', default=[])
-
-        if '直播回放' in keywords:
-            # recorded playback of live streaming
-            m3u8_url = self._html_search_regex(
-                r'file:\s*encodeURIComponent\(["\'](.+)["\']\)',
-                webpage,
-                'm3u8_url')
-            info['formats'] = self._extract_m3u8_formats(
-                m3u8_url, video_id, 'mp4', 'm3u8_native')
-        else:
-            # regular uploaded video
-            info['url'] = self._og_search_video_url(webpage)
-
-        return info