[picarto] Improve extraction (closes #6205, closes #12514, closes #15276, closes #15551)

youtube_dl/extractor/extractors.py

@@ -816,8 +816,8 @@
 from .phoenix import PhoenixIE
 from .photobucket import PhotobucketIE
 from .picarto import (
-    PicartoVodIE,
     PicartoIE,
+    PicartoVodIE,
 )
 from .piksel import PikselIE
 from .pinkbike import PinkbikeIE

youtube_dl/extractor/picarto.py

@@ -1,12 +1,21 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import time
+
 from .common import InfoExtractor
-from ..utils import ExtractorError, js_to_json, urlencode_postdata
+from ..compat import compat_str
+from ..utils import (
+    ExtractorError,
+    js_to_json,
+    try_get,
+    update_url_query,
+    urlencode_postdata,
+)
 
 
 class PicartoIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www.)?picarto\.tv/(?P<id>[a-zA-Z0-9]+)[^/]*$'
+    _VALID_URL = r'https?://(?:www.)?picarto\.tv/(?P<id>[a-zA-Z0-9]+)'
     _TEST = {
         'url': 'https://picarto.tv/Setz',
         'info_dict': {
@@ -16,72 +25,141 @@ class PicartoIE(InfoExtractor):
             'timestamp': int,
             'is_live': True
         },
-        'params': {
+        'skip': 'Stream is offline',
-            'skip_download': True
-        }
     }
 
+    @classmethod
+    def suitable(cls, url):
+        return False if PicartoVodIE.suitable(url) else super(PicartoIE, cls).suitable(url)
+
     def _real_extract(self, url):
         channel_id = self._match_id(url)
         stream_page = self._download_webpage(url, channel_id)
 
-        if 'This channel does not exist.' in stream_page:
+        if '>This channel does not exist' in stream_page:
-            raise ExtractorError('Channel does not exist', expected=True)
+            raise ExtractorError(
+                'Channel %s does not exist' % channel_id, expected=True)
 
-        player_settings_js = self._html_search_regex(
+        player = self._parse_json(
-            r'(?s)playerSettings\[1\]\s*=\s*(\{.+?\}\n)', stream_page, 'player-settings')
+            self._search_regex(
-        player_settings = self._parse_json(player_settings_js, channel_id,
+                r'(?s)playerSettings\[\d+\]\s*=\s*(\{.+?\}\s*\n)', stream_page,
-            transform_source=js_to_json)
+                'player settings'),
-        if not player_settings.get('online'):
+            channel_id, transform_source=js_to_json)
+
+        if player.get('online') is False:
             raise ExtractorError('Stream is offline', expected=True)
 
-        cdn_data = self._download_json('https://picarto.tv/process/channel', channel_id,
+        cdn_data = self._download_json(
+            'https://picarto.tv/process/channel', channel_id,
             data=urlencode_postdata({'loadbalancinginfo': channel_id}),
-            note='Fetching load balancer info')
+            note='Downloading load balancing info')
-        edge = [edge['ep'] for edge in cdn_data['edges'] if edge['id'] == cdn_data['preferedEdge']][0]
 
-        formats = self._extract_m3u8_formats('https://%s/hls/%s/index.m3u8' % (edge, channel_id),
+        def get_event(key):
-            channel_id, 'mp4')
+            return try_get(player, lambda x: x['event'][key], compat_str) or ''
-        formats.append({'url': 'https://%s/mp4/%s.mp4' % (edge, channel_id)})
+
+        params = {
+            'token': player.get('token') or '',
+            'ticket': get_event('ticket'),
+            'con': int(time.time() * 1000),
+            'type': get_event('ticket'),
+            'scope': get_event('scope'),
+        }
+
+        prefered_edge = cdn_data.get('preferedEdge')
+        default_tech = player.get('defaultTech')
+
+        formats = []
+
+        for edge in cdn_data['edges']:
+            edge_ep = edge.get('ep')
+            if not edge_ep or not isinstance(edge_ep, compat_str):
+                continue
+            edge_id = edge.get('id')
+            for tech in cdn_data['techs']:
+                tech_label = tech.get('label')
+                tech_type = tech.get('type')
+                preference = 0
+                if edge_id == prefered_edge:
+                    preference += 1
+                if tech_type == default_tech:
+                    preference += 1
+                format_id = []
+                if edge_id:
+                    format_id.append(edge_id)
+                if tech_type == 'application/x-mpegurl' or tech_label == 'HLS':
+                    format_id.append('hls')
+                    formats.extend(self._extract_m3u8_formats(
+                        update_url_query(
+                            'https://%s/hls/%s/index.m3u8'
+                            % (edge_ep, channel_id), params),
+                        channel_id, 'mp4', preference=preference,
+                        m3u8_id='-'.join(format_id), fatal=False))
+                    continue
+                elif tech_type == 'video/mp4' or tech_label == 'MP4':
+                    format_id.append('mp4')
+                    formats.append({
+                        'url': update_url_query(
+                            'https://%s/mp4/%s.mp4' % (edge_ep, channel_id),
+                            params),
+                        'format_id': '-'.join(format_id),
+                        'preference': preference,
+                    })
+                else:
+                    # rtmp format does not seem to work
+                    continue
         self._sort_formats(formats)
 
+        mature = player.get('mature')
+        if mature is None:
+            age_limit = None
+        else:
+            age_limit = 18 if mature is True else 0
+
         return {
             'id': channel_id,
-            'formats': formats,
-            'ext': 'mp4',
             'title': self._live_title(channel_id),
             'is_live': True,
-            'thumbnail': player_settings.get('vodThumb'),
+            'thumbnail': player.get('vodThumb'),
-            'age_limit': 18 if player_settings.get('mature') else None,
+            'age_limit': age_limit,
+            'formats': formats,
         }
 
 
 class PicartoVodIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www.)?picarto\.tv/videopopout/(?P<id>[a-zA-Z0-9_\-\.]+).flv'
+    _VALID_URL = r'https?://(?:www.)?picarto\.tv/videopopout/(?P<id>[^/?#&]+)'
-    _TEST = {
+    _TESTS = [{
-        'url': 'https://picarto.tv/videopopout/Carrot_2018.01.11.07.55.12.flv',
+        'url': 'https://picarto.tv/videopopout/ArtofZod_2017.12.12.00.13.23.flv',
-        'md5': '80765b67813053ff31d4df2bd5e900ce',
+        'md5': '3ab45ba4352c52ee841a28fb73f2d9ca',
         'info_dict': {
-            'id': 'Carrot_2018.01.11.07.55.12',
+            'id': 'ArtofZod_2017.12.12.00.13.23.flv',
             'ext': 'mp4',
-            'title': 'Carrot_2018.01.11.07.55.12',
+            'title': 'ArtofZod_2017.12.12.00.13.23.flv',
-            'thumbnail': r're:^https?://.*\.jpg$'
+            'thumbnail': r're:^https?://.*\.jpg'
-        }
+        },
-    }
+    }, {
+        'url': 'https://picarto.tv/videopopout/Plague',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
+
         webpage = self._download_webpage(url, video_id)
 
-        vod_info_js = self._html_search_regex(r'(?s)"#vod-player",\s*(\{.+?\})\)',
+        vod_info = self._parse_json(
-            webpage, video_id)
+            self._search_regex(
-        vod_info = self._parse_json(vod_info_js, video_id, transform_source=js_to_json)
+                r'(?s)#vod-player["\']\s*,\s*(\{.+?\})\s*\)', webpage,
+                video_id),
+            video_id, transform_source=js_to_json)
+
+        formats = self._extract_m3u8_formats(
+            vod_info['vod'], video_id, 'mp4', entry_protocol='m3u8_native',
+            m3u8_id='hls')
+        self._sort_formats(formats)
 
         return {
             'id': video_id,
             'title': video_id,
-            'ext': 'mp4',
-            'protocol': 'm3u8',
-            'url': vod_info['vod'],
             'thumbnail': vod_info.get('vodThumb'),
+            'formats': formats,
         }