[joj] Rewrite and add support for generic embeds (closes #13268)

youtube_dl/extractor/generic.py

@@ -91,6 +91,7 @@ from .anvato import AnvatoIE
 from .washingtonpost import WashingtonPostIE
 from .wistia import WistiaIE
 from .mediaset import MediasetIE
+from .joj import JojIE
 
 
 class GenericIE(InfoExtractor):
@@ -1770,6 +1771,16 @@ class GenericIE(InfoExtractor):
             },
             'add_ie': [MediasetIE.ie_key()],
         },
+        {
+            # JOJ.sk embeds
+            'url': 'https://www.noviny.sk/slovensko/238543-slovenskom-sa-prehnala-vlna-silnych-burok',
+            'info_dict': {
+                'id': '238543-slovenskom-sa-prehnala-vlna-silnych-burok',
+                'title': 'Slovenskom sa prehnala vlna silných búrok',
+            },
+            'playlist_mincount': 5,
+            'add_ie': [JojIE.ie_key()],
+        },
         {
             # AMP embed (see https://www.ampproject.org/docs/reference/components/amp-video)
             'url': 'https://tvrain.ru/amp/418921/',
@@ -2722,6 +2733,12 @@ class GenericIE(InfoExtractor):
             return self.playlist_from_matches(
                 mediaset_urls, video_id, video_title, ie=MediasetIE.ie_key())
 
+        # Look for JOJ.sk embeds
+        joj_urls = JojIE._extract_urls(webpage)
+        if joj_urls:
+            return self.playlist_from_matches(
+                joj_urls, video_id, video_title, ie=JojIE.ie_key())
+
         def merge_dicts(dict1, dict2):
             merged = {}
             for k, v in dict1.items():

youtube_dl/extractor/joj.py

@@ -1,56 +1,100 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-from .common import InfoExtractor
 import re
 
+from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import (
+    int_or_none,
+    js_to_json,
+    try_get,
+)
+
 
 class JojIE(InfoExtractor):
-    _VALID_URL = r'https?://[a-z0-9]+\.joj\.sk/([^/]+/)*(?P<title_query>(?P<release_date>[0-9]{4}(-[0-9]{2}){2}).*)' # noqa
+    _VALID_URL = r'''(?x)
+                    (?:
+                        joj:|
+                        https?://media\.joj\.sk/embed/
+                    )
+                    (?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})
+                '''
     _TESTS = [{
-        'url': 'https://www.joj.sk/nove-byvanie/archiv/2017-05-28-nove-byvanie', # noqa
+        'url': 'https://media.joj.sk/embed/a388ec4c-6019-4a4a-9312-b1bee194e932',
         'info_dict': {
             'id': 'a388ec4c-6019-4a4a-9312-b1bee194e932',
             'ext': 'mp4',
-            'title': 'Nové Bývanie',
-            'release_date': '20170528'
+            'title': 'NOVÉ BÝVANIE',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'duration': 3118,
         }
     }, {
-        'url': 'http://nasi.joj.sk/epizody/2016-09-06-stari-rodicia',
-        'info_dict': {
-            'id': 'f18b2c5f-9ea8-4941-a164-a814c53306ad',
-            'ext': 'mp4',
-            'title': 'Starí Rodičia',
-            'release_date': '20160906'
-        }
+        'url': 'joj:a388ec4c-6019-4a4a-9312-b1bee194e932',
+        'only_matching': True,
     }]
 
-    media_src_url = 'http://n16.joj.sk/storage/'
-    xml_source_url = 'https://media.joj.sk/services/Video.php?clip='
+    @staticmethod
+    def _extract_urls(webpage):
+        return re.findall(
+            r'<iframe\b[^>]+\bsrc=["\'](?P<url>(?:https?:)?//media\.joj\.sk/embed/[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})',
+            webpage)
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        release_date = mobj.group('release_date').replace('-', '')
-        webpage = self._download_webpage(url, 'id')
-        video_id = self._html_search_regex(
-            r'https?://([a-z0-9]+\.)joj\.sk/embed/(?P<video_id>[a-f0-9\-]+)',
-            webpage, 'id', group='video_id')
-        xml_playlist_url = self.xml_source_url + video_id
-        xml_playlist_et = self._download_xml(xml_playlist_url, 'XML playlist')
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(
+            'https://media.joj.sk/embed/%s' % video_id, video_id)
+
+        title = self._search_regex(
+            (r'videoTitle\s*:\s*(["\'])(?P<title>(?:(?!\1).)+)\1',
+             r'<title>(?P<title>[^<]+)'), webpage, 'title',
+            default=None, group='title') or self._og_search_title(webpage)
+
+        bitrates = self._parse_json(
+            self._search_regex(
+                r'(?s)bitrates\s*=\s*({.+?});', webpage, 'bitrates',
+                default='{}'),
+            video_id, transform_source=js_to_json, fatal=False)
+
         formats = []
-        for file_el in xml_playlist_et.findall('files/file'):
-            try:
-                height = int(file_el.attrib['id'].replace('p', ''))
-            except ValueError:
-                height = 0
-            formats.append({'height': height,
-                            'url': self.media_src_url + file_el.attrib['path'].replace(  # noqa
-                                'dat/', '', 1)})
+        for format_url in try_get(bitrates, lambda x: x['mp4'], list) or []:
+            if isinstance(format_url, compat_str):
+                height = self._search_regex(
+                    r'(\d+)[pP]\.', format_url, 'height', default=None)
+                formats.append({
+                    'url': format_url,
+                    'format_id': '%sp' % height if height else None,
+                    'height': int(height),
+                })
+        if not formats:
+            playlist = self._download_xml(
+                'https://media.joj.sk/services/Video.php?clip=%s' % video_id,
+                video_id)
+            for file_el in playlist.findall('./files/file'):
+                path = file_el.get('path')
+                if not path:
+                    continue
+                format_id = file_el.get('id') or file_el.get('label')
+                formats.append({
+                    'url': 'http://n16.joj.sk/storage/%s' % path.replace(
+                        'dat/', '', 1),
+                    'format_id': format_id,
+                    'height': int_or_none(self._search_regex(
+                        r'(\d+)[pP]', format_id or path, 'height',
+                        default=None)),
+                })
         self._sort_formats(formats)
 
+        thumbnail = self._og_search_thumbnail(webpage)
+
+        duration = int_or_none(self._search_regex(
+            r'videoDuration\s*:\s*(\d+)', webpage, 'duration', fatal=False))
+
         return {
             'id': video_id,
-            'title': self._og_search_title(webpage).title(),
+            'title': title,
+            'thumbnail': thumbnail,
+            'duration': duration,
             'formats': formats,
-            'release_date': release_date
         }