[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/commonmistakes.py b/youtube_dl/extractor/commonmistakes.py
index dbbf27a7486ab78a239df5a21a761b911a1e4a6e..7e12499b1e83ab39026fb0448673dfdacddaf968 100644 (file)
--- a/youtube_dl/extractor/commonmistakes.py
+++ b/youtube_dl/extractor/commonmistakes.py
@@ -1,5 +1,7 @@
 from __future__ import unicode_literals
 
+import sys
+
 from .common import InfoExtractor
 from ..utils import ExtractorError
 
@@ -7,7 +9,7 @@ from ..utils import ExtractorError
 class CommonMistakesIE(InfoExtractor):
     IE_DESC = False  # Do not list
     _VALID_URL = r'''(?x)
-        (?:url|URL)
+        (?:url|URL)$
     '''
 
     _TESTS = [{
@@ -27,3 +29,22 @@ class CommonMistakesIE(InfoExtractor):
         if not self._downloader.params.get('verbose'):
             msg += ' Add -v to the command line to see what arguments and configuration youtube-dl got.'
         raise ExtractorError(msg, expected=True)
+
+
+class UnicodeBOMIE(InfoExtractor):
+    IE_DESC = False
+    _VALID_URL = r'(?P<bom>\ufeff)(?P<id>.*)$'
+
+    # Disable test for python 3.2 since BOM is broken in re in this version
+    # (see https://github.com/ytdl-org/youtube-dl/issues/9751)
+    _TESTS = [] if (3, 0) < sys.version_info <= (3, 3) else [{
+        'url': '\ufeffhttp://www.youtube.com/watch?v=BaW_jenozKc',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        real_url = self._match_id(url)
+        self.report_warning(
+            'Your URL starts with a Byte Order Mark (BOM). '
+            'Removing the BOM and looking for "%s" ...' % real_url)
+        return self.url_result(real_url)