#20288: fix handling of invalid numeric charrefs in HTMLParser.
diff --git a/Lib/HTMLParser.py b/Lib/HTMLParser.py
index 5a55e26..3f97830 100644
--- a/Lib/HTMLParser.py
+++ b/Lib/HTMLParser.py
@@ -195,9 +195,9 @@
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 6a0e461..cde2bd2 100644
--- a/Lib/test/test_htmlparser.py
+++ b/Lib/test/test_htmlparser.py
@@ -394,6 +394,12 @@
("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_unescape_function(self):
parser = HTMLParser.HTMLParser()