From eae6de23b20baaa164eec915b355ef7271b3150b Mon Sep 17 00:00:00 2001
From: tonghuaroot <tonghuaroot@gmail.com>
Date: Thu, 9 Jul 2026 16:42:28 +0800
Subject: [PATCH] gh-153426: Raise BadZipFile for an invalid UTF-8 filename in
 zipfile

When a central directory or local file header entry sets the UTF-8 filename
flag (general purpose bit 11) but stores bytes that are not valid UTF-8,
opening the archive raised a raw UnicodeDecodeError instead of the
BadZipFile that signals a corrupt archive. Wrap the bit-11 filename decode
in both _RealGetContents and ZipFile.open to raise BadZipFile, mirroring
the handling of the unicode path extra field, while leaving the
metadata_encoding decode path unchanged.
---
 Lib/test/test_zipfile/test_core.py            | 32 +++++++++++++++++++
 Lib/zipfile/__init__.py                       | 14 ++++++--
 ...-07-09-16-41-12.gh-issue-153426.Qh3uzJ.rst |  3 ++
 3 files changed, 47 insertions(+), 2 deletions(-)
 create mode 100644 Misc/NEWS.d/next/Library/2026-07-09-16-41-12.gh-issue-153426.Qh3uzJ.rst

diff --git a/Lib/test/test_zipfile/test_core.py b/Lib/test/test_zipfile/test_core.py
index 4f20209927e7b3d..fff0c39e7135564 100644
--- a/Lib/test/test_zipfile/test_core.py
+++ b/Lib/test/test_zipfile/test_core.py
@@ -5754,6 +5754,38 @@ def test_read_with_unsuitable_metadata_encoding(self):
         with self.assertRaises(UnicodeDecodeError):
             zipfile.ZipFile(TESTFN, "r", metadata_encoding='utf-8')
 
+    def test_read_invalid_utf8_filename(self):
+        # A file name marked as UTF-8 (general purpose bit 11) but holding
+        # bytes that are not valid UTF-8 is a corrupt archive, not a decoding
+        # error for the caller to handle.
+        def build(name, flag):
+            lfh = struct.pack(zipfile.structFileHeader,
+                              zipfile.stringFileHeader,
+                              20, 0, flag, 0, 0, 0, 0, 0, 0,
+                              len(name), 0) + name
+            cd = struct.pack(zipfile.structCentralDir,
+                             zipfile.stringCentralDir,
+                             20, 0, 20, 0, flag, 0, 0, 0, 0, 0, 0,
+                             len(name), 0, 0, 0, 0, 0, 0) + name
+            eocd = struct.pack(zipfile.structEndArchive,
+                               zipfile.stringEndArchive,
+                               0, 0, 1, 1, len(cd), len(lfh), 0)
+            return lfh + cd + eocd
+
+        # Central directory path (ZipFile()).
+        data = build(b'\xff\xfe', zipfile._MASK_UTF_FILENAME)
+        with self.assertRaisesRegex(zipfile.BadZipFile, 'is marked as UTF-8'):
+            zipfile.ZipFile(io.BytesIO(data))
+
+        # Local file header path (open()); the central directory name is
+        # valid UTF-8 so it is only the header decode that fails.  Keep the
+        # replacement the same length so header offsets stay valid.
+        good = build(b'good', zipfile._MASK_UTF_FILENAME)
+        corrupt = good.replace(b'good', b'\xff\xfe\xff\xfe', 1)
+        with zipfile.ZipFile(io.BytesIO(corrupt)) as zipfp:
+            with self.assertRaisesRegex(zipfile.BadZipFile, 'is marked as UTF-8'):
+                zipfp.read('good')
+
     def test_read_after_append(self):
         newname = '\u56db'  # Han 'four'
         newname2 = 'fünf'  # representable in cp437, but still stored as UTF-8
diff --git a/Lib/zipfile/__init__.py b/Lib/zipfile/__init__.py
index 418933a2e8d9e87..4641700e8fa351d 100644
--- a/Lib/zipfile/__init__.py
+++ b/Lib/zipfile/__init__.py
@@ -2050,7 +2050,12 @@ def _RealGetContents(self):
             flags = centdir[_CD_FLAG_BITS]
             if flags & _MASK_UTF_FILENAME:
                 # UTF-8 file names extension
-                filename = filename.decode('utf-8')
+                try:
+                    filename = filename.decode('utf-8')
+                except UnicodeDecodeError as exc:
+                    raise BadZipFile(
+                        f"File name {filename!r} is marked as UTF-8 "
+                        f"but is not valid UTF-8") from exc
             else:
                 # Historical ZIP filename encoding
                 filename = filename.decode(self.metadata_encoding or 'cp437')
@@ -2231,7 +2236,12 @@ def open(self, name, mode="r", pwd=None, *, force_zip64=False):
 
             if fheader[_FH_GENERAL_PURPOSE_FLAG_BITS] & _MASK_UTF_FILENAME:
                 # UTF-8 filename
-                fname_str = fname.decode("utf-8")
+                try:
+                    fname_str = fname.decode("utf-8")
+                except UnicodeDecodeError as exc:
+                    raise BadZipFile(
+                        f"File name {fname!r} is marked as UTF-8 "
+                        f"but is not valid UTF-8") from exc
             else:
                 fname_str = fname.decode(self.metadata_encoding or "cp437")
 
diff --git a/Misc/NEWS.d/next/Library/2026-07-09-16-41-12.gh-issue-153426.Qh3uzJ.rst b/Misc/NEWS.d/next/Library/2026-07-09-16-41-12.gh-issue-153426.Qh3uzJ.rst
new file mode 100644
index 000000000000000..7556a312a62efeb
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2026-07-09-16-41-12.gh-issue-153426.Qh3uzJ.rst
@@ -0,0 +1,3 @@
+Raise :exc:`zipfile.BadZipFile` instead of :exc:`UnicodeDecodeError` when
+opening a :class:`zipfile.ZipFile` whose member name is flagged as UTF-8 but
+contains bytes that are not valid UTF-8.