[crunchyroll] Fix and improve extraction (closes #25096, closes #25060)

youtube_dl/extractor/crunchyroll.py

@@ -13,6 +13,7 @@ from ..compat import (
     compat_b64decode,
     compat_etree_Element,
     compat_etree_fromstring,
+    compat_str,
     compat_urllib_parse_urlencode,
     compat_urllib_request,
     compat_urlparse,
@@ -25,9 +26,9 @@ from ..utils import (
     intlist_to_bytes,
     int_or_none,
     lowercase_escape,
+    merge_dicts,
     remove_end,
     sanitized_Request,
-    unified_strdate,
     urlencode_postdata,
     xpath_text,
 )
@@ -136,6 +137,7 @@ class CrunchyrollIE(CrunchyrollBaseIE, VRVIE):
             # rtmp
             'skip_download': True,
         },
+        'skip': 'Video gone',
     }, {
         'url': 'http://www.crunchyroll.com/media-589804/culture-japan-1',
         'info_dict': {
@@ -157,11 +159,12 @@ class CrunchyrollIE(CrunchyrollBaseIE, VRVIE):
         'info_dict': {
             'id': '702409',
             'ext': 'mp4',
-            'title': 'Re:ZERO -Starting Life in Another World- Episode 5 – The Morning of Our Promise Is Still Distant',
-            'description': 'md5:97664de1ab24bbf77a9c01918cb7dca9',
+            'title': compat_str,
+            'description': compat_str,
             'thumbnail': r're:^https?://.*\.jpg$',
-            'uploader': 'TV TOKYO',
-            'upload_date': '20160508',
+            'uploader': 'Re:Zero Partners',
+            'timestamp': 1462098900,
+            'upload_date': '20160501',
         },
         'params': {
             # m3u8 download
@@ -172,12 +175,13 @@ class CrunchyrollIE(CrunchyrollBaseIE, VRVIE):
         'info_dict': {
             'id': '727589',
             'ext': 'mp4',
-            'title': "KONOSUBA -God's blessing on this wonderful world! 2 Episode 1 – Give Me Deliverance From This Judicial Injustice!",
-            'description': 'md5:cbcf05e528124b0f3a0a419fc805ea7d',
+            'title': compat_str,
+            'description': compat_str,
             'thumbnail': r're:^https?://.*\.jpg$',
             'uploader': 'Kadokawa Pictures Inc.',
-            'upload_date': '20170118',
-            'series': "KONOSUBA -God's blessing on this wonderful world!",
+            'timestamp': 1484130900,
+            'upload_date': '20170111',
+            'series': compat_str,
             'season': "KONOSUBA -God's blessing on this wonderful world! 2",
             'season_number': 2,
             'episode': 'Give Me Deliverance From This Judicial Injustice!',
@@ -200,10 +204,11 @@ class CrunchyrollIE(CrunchyrollBaseIE, VRVIE):
         'info_dict': {
             'id': '535080',
             'ext': 'mp4',
-            'title': '11eyes Episode 1 – Red Night ~ Piros éjszaka',
-            'description': 'Kakeru and Yuka are thrown into an alternate nightmarish world they call "Red Night".',
+            'title': compat_str,
+            'description': compat_str,
             'uploader': 'Marvelous AQL Inc.',
-            'upload_date': '20091021',
+            'timestamp': 1255512600,
+            'upload_date': '20091014',
         },
         'params': {
             # Just test metadata extraction
@@ -224,15 +229,17 @@ class CrunchyrollIE(CrunchyrollBaseIE, VRVIE):
             # just test metadata extraction
             'skip_download': True,
         },
+        'skip': 'Video gone',
     }, {
         # A video with a vastly different season name compared to the series name
         'url': 'http://www.crunchyroll.com/nyarko-san-another-crawling-chaos/episode-1-test-590532',
         'info_dict': {
             'id': '590532',
             'ext': 'mp4',
-            'title': 'Haiyoru! Nyaruani (ONA) Episode 1 – Test',
-            'description': 'Mahiro and Nyaruko talk about official certification.',
+            'title': compat_str,
+            'description': compat_str,
             'uploader': 'TV TOKYO',
+            'timestamp': 1330956000,
             'upload_date': '20120305',
             'series': 'Nyarko-san: Another Crawling Chaos',
             'season': 'Haiyoru! Nyaruani (ONA)',
@@ -442,23 +449,21 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
             webpage, 'language', default=None, group='lang')
 
         video_title = self._html_search_regex(
-            r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>',
-            webpage, 'video_title')
+            (r'(?s)<h1[^>]*>((?:(?!<h1).)*?<(?:span[^>]+itemprop=["\']title["\']|meta[^>]+itemprop=["\']position["\'])[^>]*>(?:(?!<h1).)+?)</h1>',
+             r'<title>(.+?),\s+-\s+.+? Crunchyroll'),
+            webpage, 'video_title', default=None)
+        if not video_title:
+            video_title = re.sub(r'^Watch\s+', '', self._og_search_description(webpage))
         video_title = re.sub(r' {2,}', ' ', video_title)
         video_description = (self._parse_json(self._html_search_regex(
             r'<script[^>]*>\s*.+?\[media_id=%s\].+?({.+?"description"\s*:.+?})\);' % video_id,
             webpage, 'description', default='{}'), video_id) or media_metadata).get('description')
         if video_description:
             video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
-        video_upload_date = self._html_search_regex(
-            [r'<div>Availability for free users:(.+?)</div>', r'<div>[^<>]+<span>\s*(.+?\d{4})\s*</span></div>'],
-            webpage, 'video_upload_date', fatal=False, flags=re.DOTALL)
-        if video_upload_date:
-            video_upload_date = unified_strdate(video_upload_date)
         video_uploader = self._html_search_regex(
             # try looking for both an uploader that's a link and one that's not
             [r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', r'<div>\s*Publisher:\s*<span>\s*(.+?)\s*</span>\s*</div>'],
-            webpage, 'video_uploader', fatal=False)
+            webpage, 'video_uploader', default=False)
 
         formats = []
         for stream in media.get('streams', []):
@@ -611,14 +616,15 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
             r'(?s)<h\d[^>]+id=["\']showmedia_about_episode_num[^>]+>.+?</h\d>\s*<h4>\s*Season (\d+)',
             webpage, 'season number', default=None))
 
-        return {
+        info = self._search_json_ld(webpage, video_id, default={})
+
+        return merge_dicts({
             'id': video_id,
             'title': video_title,
             'description': video_description,
             'duration': duration,
             'thumbnail': thumbnail,
             'uploader': video_uploader,
-            'upload_date': video_upload_date,
             'series': series,
             'season': season,
             'season_number': season_number,
@@ -626,7 +632,7 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
             'episode_number': episode_number,
             'subtitles': subtitles,
             'formats': formats,
-        }
+        }, info)
 
 
 class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):