bpo-30693: zip+tarfile: sort directory listing (#2263)

tarfile and zipfile now sort directory listing to generate tar and zip archives
in a more reproducible way.

See also https://reproducible-builds.org/docs/stable-inputs/ on that topic.

3 files changed, 28 insertions, 4 deletions

diff --git a/Lib/tarfile.py b/Lib/tarfile.py
index 0b8d31f..a24ee42 100755
--- a/Lib/tarfile.py
+++ b/Lib/tarfile.py
@@ -1943,7 +1943,7 @@ class TarFile(object):
         elif tarinfo.isdir():
             self.addfile(tarinfo)
             if recursive:
-                for f in os.listdir(name):
+                for f in sorted(os.listdir(name)):
                     self.add(os.path.join(name, f), os.path.join(arcname, f),
                             recursive, filter=filter)
 
diff --git a/Lib/test/test_tarfile.py b/Lib/test/test_tarfile.py
index 179cbc6..8ef4294 100644
--- a/Lib/test/test_tarfile.py
+++ b/Lib/test/test_tarfile.py
@@ -1129,6 +1129,30 @@ class WriteTest(WriteTestBase, unittest.TestCase):
         finally:
             support.rmdir(path)
 
+    # mock the following:
+    #  os.listdir: so we know that files are in the wrong order
+    @unittest.mock.patch('os.listdir')
+    def test_ordered_recursion(self, mock_listdir):
+        path = os.path.join(TEMPDIR, "directory")
+        os.mkdir(path)
+        open(os.path.join(path, "1"), "a").close()
+        open(os.path.join(path, "2"), "a").close()
+        mock_listdir.return_value = ["2", "1"]
+        try:
+            tar = tarfile.open(tmpname, self.mode)
+            try:
+                tar.add(path)
+                paths = []
+                for m in tar.getmembers():
+                    paths.append(os.path.split(m.name)[-1])
+                self.assertEqual(paths, ["directory", "1", "2"]);
+            finally:
+                tar.close()
+        finally:
+            support.unlink(os.path.join(path, "1"))
+            support.unlink(os.path.join(path, "2"))
+            support.rmdir(path)
+
     def test_gettarinfo_pathlike_name(self):
         with tarfile.open(tmpname, self.mode) as tar:
             path = pathlib.Path(TEMPDIR) / "file"
diff --git a/Lib/zipfile.py b/Lib/zipfile.py
index 5df7b1b..b90b60f 100644
--- a/Lib/zipfile.py
+++ b/Lib/zipfile.py
@@ -1940,7 +1940,7 @@ class PyZipFile(ZipFile):
                 if self.debug:
                     print("Adding", arcname)
                 self.write(fname, arcname)
-                dirlist = os.listdir(pathname)
+                dirlist = sorted(os.listdir(pathname))
                 dirlist.remove("__init__.py")
                 # Add all *.py files and package subdirectories
                 for filename in dirlist:
@@ -1965,7 +1965,7 @@ class PyZipFile(ZipFile):
                 # This is NOT a package directory, add its files at top level
                 if self.debug:
                     print("Adding files from directory", pathname)
-                for filename in os.listdir(pathname):
+                for filename in sorted(os.listdir(pathname)):
                     path = os.path.join(pathname, filename)
                     root, ext = os.path.splitext(filename)
                     if ext == ".py":
@@ -2116,7 +2116,7 @@ def main(args=None):
             elif os.path.isdir(path):
                 if zippath:
                     zf.write(path, zippath)
-                for nm in os.listdir(path):
+                for nm in sorted(os.listdir(path)):
                     addToZip(zf,
                              os.path.join(path, nm), os.path.join(zippath, nm))
             # else: ignore