[utils] Fix unescapeHTML for misformed string like "&a"" (#13935)

[youtube-dl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 39860e9d156baf2ceb205460c5f78fa2e9026f52..2554a2abd7e86eeac1c92dddf6e90a307f8b9c9f 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -365,9 +365,9 @@ def get_elements_by_attribute(attribute, value, html, escape_value=True):
     retlist = []
     for m in re.finditer(r'''(?xs)
         <([a-zA-Z0-9:._-]+)
-         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*?
+         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
          \s+%s=['"]?%s['"]?
-         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*?
+         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
         \s*>
         (?P<content>.*?)
         </\1>
@@ -596,7 +596,7 @@ def unescapeHTML(s):
     assert type(s) == compat_str
 
     return re.sub(
-        r'&([^;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
+        r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
 
 
 def get_subprocess_encoding():
@@ -2733,6 +2733,8 @@ def cli_option(params, command_option, param):
 
 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
     param = params.get(param)
+    if param is None:
+        return []
     assert isinstance(param, bool)
     if separator:
         return [command_option + separator + (true_value if param else false_value)]