Allow extractors to specify section_start/end for clips

[yt-dlp.git] / yt_dlp / extractor / tvp.py

diff --git a/yt_dlp/extractor/tvp.py b/yt_dlp/extractor/tvp.py
index 22cfbd25e0fa86562f3e8cf8658c492ec07f4f92..69168f6558bde624db53d63439307da3f3c4b0f8 100644 (file)
--- a/yt_dlp/extractor/tvp.py
+++ b/yt_dlp/extractor/tvp.py
@@ -1,6 +1,3 @@
-# coding: utf-8
-from __future__ import unicode_literals
-
 import itertools
 import random
 import re
@@ -251,6 +248,52 @@ def _real_extract(self, url):
         }
 
 
+class TVPStreamIE(InfoExtractor):
+    IE_NAME = 'tvp:stream'
+    _VALID_URL = r'(?:tvpstream:|https?://tvpstream\.vod\.tvp\.pl/(?:\?(?:[^&]+[&;])*channel_id=)?)(?P<id>\d*)'
+    _TESTS = [{
+        # untestable as "video" id changes many times across a day
+        'url': 'https://tvpstream.vod.tvp.pl/?channel_id=1455',
+        'only_matching': True,
+    }, {
+        'url': 'tvpstream:39821455',
+        'only_matching': True,
+    }, {
+        # the default stream when you provide no channel_id, most probably TVP Info
+        'url': 'tvpstream:',
+        'only_matching': True,
+    }, {
+        'url': 'https://tvpstream.vod.tvp.pl/',
+        'only_matching': True,
+    }]
+
+    _PLAYER_BOX_RE = r'<div\s[^>]*id\s*=\s*["\']?tvp_player_box["\']?[^>]+data-%s-id\s*=\s*["\']?(\d+)'
+    _BUTTON_RE = r'<div\s[^>]*data-channel-id=["\']?%s["\']?[^>]*\sdata-title=(?:"([^"]*)"|\'([^\']*)\')[^>]*\sdata-stationname=(?:"([^"]*)"|\'([^\']*)\')'
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+        channel_url = self._proto_relative_url('//tvpstream.vod.tvp.pl/?channel_id=%s' % channel_id or 'default')
+        webpage = self._download_webpage(channel_url, channel_id, 'Downloading channel webpage')
+        if not channel_id:
+            channel_id = self._search_regex(self._PLAYER_BOX_RE % 'channel',
+                                            webpage, 'default channel id')
+        video_id = self._search_regex(self._PLAYER_BOX_RE % 'video',
+                                      webpage, 'video id')
+        audition_title, station_name = self._search_regex(
+            self._BUTTON_RE % (re.escape(channel_id)), webpage,
+            'audition title and station name',
+            group=(1, 2))
+        return {
+            '_type': 'url_transparent',
+            'id': channel_id,
+            'url': 'tvp:%s' % video_id,
+            'title': audition_title,
+            'alt_title': station_name,
+            'is_live': True,
+            'ie_key': 'TVPEmbed',
+        }
+
+
 class TVPEmbedIE(InfoExtractor):
     IE_NAME = 'tvp:embed'
     IE_DESC = 'Telewizja Polska'