[3.13] gh-86155: Fix data loss after unclosed script or style tag in HTMLParser ...

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 53383e90e4df7029f792b7aa81aa2e4cff348ed0)

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

diff --git a/Lib/html/parser.py b/Lib/html/parser.py
index 1b8b6ea0e5ab7ab5acada5dc0dd896137b558520..1e30956fe24f83b6e88a4cb2e9114b16760969ed 100644 (file)
--- a/Lib/html/parser.py
+++ b/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:

diff --git a/Lib/test/test_htmlparser.py b/Lib/test/test_htmlparser.py
index 68649e9d6d5e9c86a038bb4450a26e86bcee030e..61fa24fab574f25f07eac855fcfc223d166d0b14 100644 (file)
--- a/Lib/test/test_htmlparser.py
+++ b/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!-->'

diff --git a/Misc/NEWS.d/next/Library/2023-02-13-21-41-34.gh-issue-86155.ppIGSC.rst b/Misc/NEWS.d/next/Library/2023-02-13-21-41-34.gh-issue-86155.ppIGSC.rst
new file mode 100644 (file)
index 0000000..bb85481
--- /dev/null
+++ b/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.