#20288: fix handling of invalid numeric charrefs in HTMLParser.

diff --git a/Lib/html/parser.py b/Lib/html/parser.py
index 2d3bef351b0ad0fe2285166908dec2ba4f9070ca..63fe77425bdc45457243c373984c6666fe8ce57b 100644 (file)
--- a/Lib/html/parser.py
+++ b/Lib/html/parser.py
@@ -228,9 +228,9 @@ class HTMLParser(_markupbase.ParserBase):
                     i = self.updatepos(i, k)
                     continue
                 else:
-                    if ";" in rawdata[i:]: #bail by consuming &#
-                        self.handle_data(rawdata[0:2])
-                        i = self.updatepos(i, 2)
+                    if ";" in rawdata[i:]:  # bail by consuming &#
+                        self.handle_data(rawdata[i:i+2])
+                        i = self.updatepos(i, i+2)
                     break
             elif startswith('&', i):
                 match = entityref.match(rawdata, i)

diff --git a/Lib/test/test_htmlparser.py b/Lib/test/test_htmlparser.py
index c977a9dd4d7111a3e1794bbadf007f5b73aeb390..11d9c9ce8b8c1a4c60250f57226d1d19f38eca33 100644 (file)
--- a/Lib/test/test_htmlparser.py
+++ b/Lib/test/test_htmlparser.py
@@ -151,6 +151,12 @@ text
             ("data", "&#bad;"),
             ("endtag", "p"),
         ])
+        # add the [] as a workaround to avoid buffering (see #20288)
+        self._run_check(["<div>&#bad;</div>"], [
+            ("starttag", "div", []),
+            ("data", "&#bad;"),
+            ("endtag", "div"),
+        ])
 
     def test_unclosed_entityref(self):
         self._run_check("&entityref foo", [

diff --git a/Misc/NEWS b/Misc/NEWS
index 5a84af86cd779ef2ecd8fb8cd7d44182818f1a7c..d2efc232a2d50027b2735247fcc9ff164c80c4d0 100644 (file)
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -45,6 +45,8 @@ Core and Builtins
 Library
 -------
 
+- Issue #20288: fix handling of invalid numeric charrefs in HTMLParser.
+
 - Issue #20424: Python implementation of io.StringIO now supports lone surrogates.
 
 - Issue #19456: ntpath.join() now joins relative paths correctly when a drive