[youtube] Fix extraction (closes #20758, closes #20759, closes #20761, closes #20762...

[youtube-dl] / youtube_dl / extractor / chaturbate.py

diff --git a/youtube_dl/extractor/chaturbate.py b/youtube_dl/extractor/chaturbate.py
index 1c2f065dfb134f3050680ae9b3a6a4ba419933b5..e2b828d8afdada257b6b4f458022947522180642 100644 (file)
--- a/youtube_dl/extractor/chaturbate.py
+++ b/youtube_dl/extractor/chaturbate.py
@@ -31,10 +31,18 @@ class ChaturbateIE(InfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage(
+            url, video_id, headers=self.geo_verification_headers())
 
-        m3u8_urls = re.findall(
-            r'var hlsSource.+? = (["\'])(?P<url>http.+?\.m3u8)', webpage)
+        m3u8_urls = []
+
+        for m in re.finditer(
+                r'(["\'])(?P<url>http.+?\.m3u8.*?)\1', webpage):
+            m3u8_fast_url, m3u8_no_fast_url = m.group('url'), m.group(
+                'url').replace('_fast', '')
+            for m3u8_url in (m3u8_fast_url, m3u8_no_fast_url):
+                if m3u8_url not in m3u8_urls:
+                    m3u8_urls.append(m3u8_url)
 
         if not m3u8_urls:
             error = self._search_regex(
@@ -51,8 +59,12 @@ class ChaturbateIE(InfoExtractor):
 
         formats = []
         for m3u8_url in m3u8_urls:
-            formats.append(self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4')[0])
-
+            m3u8_id = 'fast' if '_fast' in m3u8_url else 'slow'
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, video_id, ext='mp4',
+                # ffmpeg skips segments for fast m3u8
+                preference=-10 if m3u8_id == 'fast' else None,
+                m3u8_id=m3u8_id, fatal=False, live=True))
         self._sort_formats(formats)
 
         return {