[3.14] gh-86155: Fix data loss after unclosed script or style tag in HTMLParser (GH-22658) (GH-133844)

When calling .close() the HTMLParser should flush all remaining content,
even when that content is in an unclosed script or style tag.
(cherry picked from commit 53383e90e4)

Co-authored-by: Waylan Limberg <waylan.limberg@icloud.com>

Lib/html/parser.py

@@ -260,7 +260,7 @@ class HTMLParser(_markupbase.ParserBase):
             else:
                 assert 0, "interesting.search() lied"
         # end while
-        if end and i < n and not self.cdata_elem:
+        if end and i < n:
             if self.convert_charrefs and not self.cdata_elem:
                 self.handle_data(unescape(rawdata[i:n]))
             else:

Lib/test/test_htmlparser.py

@@ -317,6 +317,16 @@ text
                                 ("endtag", element_lower)],
                             collector=Collector(convert_charrefs=False))
 
+    def test_EOF_in_cdata(self):
+        content = """<!-- not a comment --> &not-an-entity-ref;
+                  <a href="" /> </p><p> <span></span></style>
+                  '</script' + '>'"""
+        s = f'<script>{content}'
+        self._run_check(s, [
+            ("starttag", 'script', []),
+            ("data", content)
+        ])
+
     def test_comments(self):
         html = ("<!-- I'm a valid comment -->"
                 '<!--me too!-->'

Misc/NEWS.d/next/Library/2023-02-13-21-41-34.gh-issue-86155.ppIGSC.rst

@@ -0,0 +1,2 @@
+:meth:`html.parser.HTMLParser.close` no longer loses data when the
+``<script>`` tag is not closed. Patch by Waylan Limberg.