[3.14] gh-135661: Fix abrupt closing of empty comment in HTMLParser (GH-153007) (GH-153025)

An abruptly closed empty comment ("<!-->" or "<!--->") no longer extends
up to a later "-->" in the same feed() call.

test_htmlparser now also feeds each string source as a single chunk, in
addition to one character at a time, to exercise different input buffering.
(cherry picked from commit ed370d3)

Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

Author: 3 people

Lib/html/parser.py

@@ -387,9 +387,11 @@ def parse_html_declaration(self, i):
     def parse_comment(self, i, report=True):
         rawdata = self.rawdata
         assert rawdata.startswith('<!--', i), 'unexpected call to parse_comment()'
-        match = commentclose.search(rawdata, i+4)
+        # An empty comment is abruptly closed by the first ">" or "->",
+        # taking priority over a later "-->" or "--!>" close.
+        match = commentabruptclose.match(rawdata, i+4)
         if not match:
-            match = commentabruptclose.match(rawdata, i+4)
+            match = commentclose.search(rawdata, i+4)
             if not match:
                 return -1
         if report:

Lib/test/test_htmlparser.py

@@ -116,6 +116,11 @@ def _run_check(self, source, expected_events,
                    *, collector=None, convert_charrefs=False):
         if collector is None:
             collector = self.get_collector(convert_charrefs=convert_charrefs)
+            if isinstance(source, str):
+                # Also feed the whole string at once, not just character by
+                # character (below), to exercise different input buffering.
+                self._run_check([source], expected_events,
+                                convert_charrefs=convert_charrefs)
         parser = collector
         for s in source:
             parser.feed(s)
@@ -593,6 +598,9 @@ def test_comments(self):
                 '<!-- <!-- nested --> -->'
                 '<!--<!-->'
                 '<!--<!--!>'
+                # abruptly closed empty comment must not swallow later text
+                '<!-->x-->'
+                '<!--->y-->'
         )
         expected = [('comment', " I'm a valid comment "),
                     ('comment', 'me too!'),
@@ -613,6 +621,8 @@ def test_comments(self):
                     ('comment', ' <!-- nested '), ('data', ' -->'),
                     ('comment', '<!'),
                     ('comment', '<!'),
+                    ('comment', ''), ('data', 'x-->'),
+                    ('comment', ''), ('data', 'y-->'),
         ]
         self._run_check(html, expected)
 

Misc/NEWS.d/next/Library/2026-07-04-13-00-00.gh-issue-135661.CIkADG.rst

@@ -0,0 +1,3 @@
+Fix :class:`html.parser.HTMLParser`: an abruptly closed empty comment
+(``<!-->`` or ``<!--->``) no longer extends up to a later ``-->`` in the same
+:meth:`~html.parser.HTMLParser.feed` call.