[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/chaturbate.py b/youtube_dl/extractor/chaturbate.py
index 1c2f065dfb134f3050680ae9b3a6a4ba419933b5..a459dcb8d5fbedc00f068da023392f0898d687a2 100644 (file)
--- a/youtube_dl/extractor/chaturbate.py
+++ b/youtube_dl/extractor/chaturbate.py
@@ -3,11 +3,15 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..utils import ExtractorError
+from ..utils import (
+    ExtractorError,
+    lowercase_escape,
+    url_or_none,
+)
 
 
 class ChaturbateIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:[^/]+\.)?chaturbate\.com/(?P<id>[^/?#]+)'
+    _VALID_URL = r'https?://(?:[^/]+\.)?chaturbate\.com/(?:fullvideo/?\?.*?\bb=)?(?P<id>[^/?&#]+)'
     _TESTS = [{
         'url': 'https://www.chaturbate.com/siswet19/',
         'info_dict': {
@@ -21,6 +25,9 @@ class ChaturbateIE(InfoExtractor):
             'skip_download': True,
         },
         'skip': 'Room is offline',
+    }, {
+        'url': 'https://chaturbate.com/fullvideo/?b=caylin',
+        'only_matching': True,
     }, {
         'url': 'https://en.chaturbate.com/siswet19/',
         'only_matching': True,
@@ -31,10 +38,38 @@ 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(
+            'https://chaturbate.com/%s/' % video_id, video_id,
+            headers=self.geo_verification_headers())
+
+        found_m3u8_urls = []
+
+        data = self._parse_json(
+            self._search_regex(
+                r'initialRoomDossier\s*=\s*(["\'])(?P<value>(?:(?!\1).)+)\1',
+                webpage, 'data', default='{}', group='value'),
+            video_id, transform_source=lowercase_escape, fatal=False)
+        if data:
+            m3u8_url = url_or_none(data.get('hls_source'))
+            if m3u8_url:
+                found_m3u8_urls.append(m3u8_url)
 
-        m3u8_urls = re.findall(
-            r'var hlsSource.+? = (["\'])(?P<url>http.+?\.m3u8)', webpage)
+        if not found_m3u8_urls:
+            for m in re.finditer(
+                    r'(\\u002[27])(?P<url>http.+?\.m3u8.*?)\1', webpage):
+                found_m3u8_urls.append(lowercase_escape(m.group('url')))
+
+        if not found_m3u8_urls:
+            for m in re.finditer(
+                    r'(["\'])(?P<url>http.+?\.m3u8.*?)\1', webpage):
+                found_m3u8_urls.append(m.group('url'))
+
+        m3u8_urls = []
+        for found_m3u8_url in found_m3u8_urls:
+            m3u8_fast_url, m3u8_no_fast_url = found_m3u8_url, found_m3u8_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 +86,17 @@ class ChaturbateIE(InfoExtractor):
 
         formats = []
         for m3u8_url in m3u8_urls:
-            formats.append(self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4')[0])
-
+            for known_id in ('fast', 'slow'):
+                if '_%s' % known_id in m3u8_url:
+                    m3u8_id = known_id
+                    break
+            else:
+                m3u8_id = None
+            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 {