gh-153625: rewrite with original filename in append mode for zipfile

Lib/test/test_zipfile/test_core.py

@@ -4827,6 +4827,29 @@ def test_overlap_with_archive_comment(self):
             with self.assertRaisesRegex(zipfile.BadZipFile, 'Overlapped entries'):
                 zipf.read('a')
 
+    def test_append_keep_filename(self):
+        """Files loaded from an archive should keep original filename when
+        rewritten to central directory in append mode."""
+        with mock.patch('os.sep', '/'), mock.patch('os.altsep', None):
+            with zipfile.ZipFile(TESTFN, mode="w") as zipfp:
+                zinfo = zipfile.ZipInfo('MyFolder/My\\File.txt')
+                zipfp.writestr(zinfo, 'foo')
+
+        with mock.patch('os.sep', '\\'), mock.patch('os.altsep', '/'):
+            with zipfile.ZipFile(TESTFN, "a") as zipfp:
+                zi = zipfp.infolist()[0]
+                self.assertEqual(zi.orig_filename, 'MyFolder/My\\File.txt')
+                self.assertEqual(zi.filename, 'MyFolder/My/File.txt')
+                self.assertIsNone(zipfp.testzip())
+                # trigger archive rewriting
+                zipfp.comment = b''
+
+            with zipfile.ZipFile(TESTFN, "r") as zipfp:
+                zi = zipfp.infolist()[0]
+                self.assertEqual(zi.orig_filename, 'MyFolder/My\\File.txt')
+                self.assertEqual(zi.filename, 'MyFolder/My/File.txt')
+                self.assertIsNone(zipfp.testzip())
+
     def tearDown(self):
         unlink(TESTFN)
         unlink(TESTFN2)

Lib/zipfile/__init__.py

@@ -432,7 +432,7 @@ class ZipInfo:
     """Class with attributes describing each file in the ZIP archive."""
 
     __slots__ = (
-        'orig_filename',
+        '_orig_filename',
         'filename',
         'date_time',
         'compress_type',
@@ -456,8 +456,6 @@ class ZipInfo:
     )
 
     def __init__(self, filename="NoName", date_time=(1980,1,1,0,0,0)):
-        self.orig_filename = filename   # Original file name in archive
-
         # Terminate the file name at the first null byte and
         # ensure paths always use forward slashes as the directory separator.
         filename = _sanitize_filename(filename)
@@ -487,11 +485,25 @@ def __init__(self, filename="NoName", date_time=(1980,1,1,0,0,0)):
         self.external_attr = 0          # External file attributes
         self.compress_size = 0          # Size of the compressed file
         self.file_size = 0              # Size of the uncompressed file
+
+        # Special internal attributes set by class ZipFile when read from an archive:
+        self._orig_filename = None      # Original file name in archive
         self._end_offset = None         # Start of the next local header or central directory
+
         # Other attributes are set by class ZipFile:
         # header_offset         Byte offset to the file header
         # CRC                   CRC-32 of the uncompressed file
 
+    @property
+    def orig_filename(self):
+        return (self._orig_filename if self._orig_filename is not None
+                else self.filename)
+
+    # Maintain backward compatibility in case someone sets it.
+    @orig_filename.setter
+    def orig_filename(self, value):
+        self._orig_filename = value
+
     # Maintain backward compatibility with the old protected attribute name.
     @property
     def _compresslevel(self):
@@ -579,9 +591,9 @@ def _encodeFilenameFlags(self):
         else:
             encoding = 'cp437'
         try:
-            return self.filename.encode(encoding), self.flag_bits & ~_MASK_UTF_FILENAME
+            return self.orig_filename.encode(encoding), self.flag_bits & ~_MASK_UTF_FILENAME
         except UnicodeEncodeError:
-            return self.filename.encode('utf-8'), self.flag_bits | _MASK_UTF_FILENAME
+            return self.orig_filename.encode('utf-8'), self.flag_bits | _MASK_UTF_FILENAME
 
     def _decodeExtra(self, filename_crc):
         # Try to decode the extra field.
@@ -2056,6 +2068,7 @@ def _RealGetContents(self):
                 filename = filename.decode(self.metadata_encoding or 'cp437')
             # Create ZipInfo instance to store file information
             x = ZipInfo(filename)
+            x._orig_filename = filename
             x.extra = fp.read(centdir[_CD_EXTRA_FIELD_LENGTH])
             x.comment = fp.read(centdir[_CD_COMMENT_LENGTH])
             x.header_offset = centdir[_CD_LOCAL_HEADER_OFFSET]

Misc/NEWS.d/next/Library/2026-07-12-16-11-18.gh-issue-153625.-wpBTM.rst

@@ -0,0 +1 @@
+Fix an issue that an auto-sanitized filename corrupts a ZIP archive when rewritten to the central directory in append mode.