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

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

diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index cbb7354039b86c4ac08fb8d4c8bb3d5174e084a0..8c87f6dd3bfde87fdb8d27935298a45f2861db47 100644 (file)
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -8,7 +8,6 @@ import json
 
 from .common import InfoExtractor
 from ..compat import (
-    compat_HTTPError,
     compat_kwargs,
     compat_parse_qs,
     compat_str,
@@ -39,7 +38,7 @@ class TwitchBaseIE(InfoExtractor):
     _USHER_BASE = 'https://usher.ttvnw.net'
     _LOGIN_FORM_URL = 'https://www.twitch.tv/login'
     _LOGIN_POST_URL = 'https://passport.twitch.tv/login'
-    _CLIENT_ID = 'jzkbprff40iqj646a697cyrvl0zt2m6'
+    _CLIENT_ID = 'kimne78kx3ncx6brgo4mv6wki5h1ko'
     _NETRC_MACHINE = 'twitch'
 
     def _handle_error(self, response):
@@ -52,7 +51,9 @@ class TwitchBaseIE(InfoExtractor):
                 expected=True)
 
     def _call_api(self, path, item_id, *args, **kwargs):
-        kwargs.setdefault('headers', {})['Client-ID'] = self._CLIENT_ID
+        headers = kwargs.get('headers', {}).copy()
+        headers['Client-ID'] = self._CLIENT_ID
+        kwargs['headers'] = headers
         response = self._download_json(
             '%s/%s' % (self._API_BASE, path), item_id,
             *args, **compat_kwargs(kwargs))
@@ -84,20 +85,15 @@ class TwitchBaseIE(InfoExtractor):
             headers = {
                 'Referer': page_url,
                 'Origin': page_url,
-                'Content-Type': 'text/plain;charset=UTF-8'
+                'Content-Type': 'text/plain;charset=UTF-8',
             }
 
-            try:
-                response = self._download_json(
-                    post_url, None, note,
-                    data=json.dumps(form).encode(),
-                    headers=headers)
-            except ExtractorError as e:
-                if isinstance(e.cause, compat_HTTPError) and e.cause.code == 400:
-                    response = self._parse_json(
-                        e.cause.read().decode('utf-8'), None)
-                    fail(response.get('error_description') or response.get('error_code'))
-                raise
+            response = self._download_json(
+                post_url, None, note, data=json.dumps(form).encode(),
+                headers=headers, expected_status=400)
+            error = response.get('error_description') or response.get('error_code')
+            if error:
+                fail(error)
 
             if 'Authenticated successfully' in response.get('message', ''):
                 return None, None
@@ -140,7 +136,12 @@ class TwitchBaseIE(InfoExtractor):
             source = next(f for f in formats if f['format_id'] == 'Source')
             source['preference'] = 10
         except StopIteration:
-            pass  # No Source stream present
+            for f in formats:
+                if '/chunked/' in f['url']:
+                    f.update({
+                        'source_preference': 10,
+                        'format_note': 'Source',
+                    })
         self._sort_formats(formats)
 
 
@@ -565,7 +566,8 @@ class TwitchStreamIE(TwitchBaseIE):
                     TwitchAllVideosIE,
                     TwitchUploadsIE,
                     TwitchPastBroadcastsIE,
-                    TwitchHighlightsIE))
+                    TwitchHighlightsIE,
+                    TwitchClipsIE))
                 else super(TwitchStreamIE, cls).suitable(url))
 
     def _real_extract(self, url):
@@ -639,7 +641,7 @@ class TwitchStreamIE(TwitchBaseIE):
 
 class TwitchClipsIE(TwitchBaseIE):
     IE_NAME = 'twitch:clips'
-    _VALID_URL = r'https?://clips\.twitch\.tv/(?:[^/]+/)*(?P<id>[^/?#&]+)'
+    _VALID_URL = r'https?://(?:clips\.twitch\.tv/(?:[^/]+/)*|(?:www\.)?twitch\.tv/[^/]+/clip/)(?P<id>[^/?#&]+)'
 
     _TESTS = [{
         'url': 'https://clips.twitch.tv/FaintLightGullWholeWheat',
@@ -659,6 +661,9 @@ class TwitchClipsIE(TwitchBaseIE):
         # multiple formats
         'url': 'https://clips.twitch.tv/rflegendary/UninterestedBeeDAESuppy',
         'only_matching': True,
+    }, {
+        'url': 'https://www.twitch.tv/sergeynixon/clip/StormyThankfulSproutFutureMan',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):