gh-117084: Fix ZIP file extraction for directory entry names with backslashes on...

diff --git a/Lib/test/archivetestdata/zipdir_backslash.zip b/Lib/test/archivetestdata/zipdir_backslash.zip
new file mode 100644 (file)
index 0000000..979126e
Binary files /dev/null and b/Lib/test/archivetestdata/zipdir_backslash.zip differ

diff --git a/Lib/test/test_zipfile/test_core.py b/Lib/test/test_zipfile/test_core.py
index 368e60a37b0555706ac6f814458f34b50ae517d6..a605aa1f14fe3f6e551befdb8cc01da148410f2d 100644 (file)
--- a/Lib/test/test_zipfile/test_core.py
+++ b/Lib/test/test_zipfile/test_core.py
@@ -2937,6 +2937,22 @@ class TestWithDirectory(unittest.TestCase):
         os.mkdir(os.path.join(TESTFN2, "a"))
         self.test_extract_dir()
 
+    def test_extract_dir_backslash(self):
+        zfname = findfile("zipdir_backslash.zip", subdir="archivetestdata")
+        with zipfile.ZipFile(zfname) as zipf:
+            zipf.extractall(TESTFN2)
+        if os.name == 'nt':
+            self.assertTrue(os.path.isdir(os.path.join(TESTFN2, "a")))
+            self.assertTrue(os.path.isdir(os.path.join(TESTFN2, "a", "b")))
+            self.assertTrue(os.path.isfile(os.path.join(TESTFN2, "a", "b", "c")))
+            self.assertTrue(os.path.isdir(os.path.join(TESTFN2, "d")))
+            self.assertTrue(os.path.isdir(os.path.join(TESTFN2, "d", "e")))
+        else:
+            self.assertTrue(os.path.isfile(os.path.join(TESTFN2, "a\\b\\c")))
+            self.assertTrue(os.path.isfile(os.path.join(TESTFN2, "d\\e\\")))
+            self.assertFalse(os.path.exists(os.path.join(TESTFN2, "a")))
+            self.assertFalse(os.path.exists(os.path.join(TESTFN2, "d")))
+
     def test_write_dir(self):
         dirpath = os.path.join(TESTFN2, "x")
         os.mkdir(dirpath)

diff --git a/Lib/zipfile/__init__.py b/Lib/zipfile/__init__.py
index cc08f602fe44e0505297e4b182a50da191fc54d0..b330ece4b276ad7ee63224023956cc1c9b8cbca7 100644 (file)
--- a/Lib/zipfile/__init__.py
+++ b/Lib/zipfile/__init__.py
@@ -605,7 +605,15 @@ class ZipInfo:
 
     def is_dir(self):
         """Return True if this archive member is a directory."""
-        return self.filename.endswith('/')
+        if self.filename.endswith('/'):
+            return True
+        # The ZIP format specification requires to use forward slashes
+        # as the directory separator, but in practice some ZIP files
+        # created on Windows can use backward slashes.  For compatibility
+        # with the extraction code which already handles this:
+        if os.path.altsep:
+            return self.filename.endswith((os.path.sep, os.path.altsep))
+        return False
 
 
 # ZIP encryption uses the CRC32 one-byte primitive for scrambling some

diff --git a/Misc/NEWS.d/next/Library/2024-03-21-17-07-38.gh-issue-117084.w1mTpT.rst b/Misc/NEWS.d/next/Library/2024-03-21-17-07-38.gh-issue-117084.w1mTpT.rst
new file mode 100644 (file)
index 0000000..6e7790e
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2024-03-21-17-07-38.gh-issue-117084.w1mTpT.rst
@@ -0,0 +1,2 @@
+Fix :mod:`zipfile` extraction for directory entries with the name containing
+backslashes on Windows.