bpo-34606: decode extra data only if zip64 end of central directory record is present

Lib/test/test_zipfile.py

@@ -2262,6 +2262,30 @@ def test_from_dir(self):
         self.assertEqual(zi.file_size, 0)
 
 
+class ZipInfoWithExtraTest(unittest.TestCase):
+    path = pathlib.Path(TESTFN2)
+    def setUp(self):
+        os.mkdir(ZipInfoWithExtraTest.path)
+
+    def tearDown(self):
+        rmtree(ZipInfoWithExtraTest.path)
+
+    @requires_zlib
+    def test_extra_field(self):
+
+        f = ZipInfoWithExtraTest.path / "f.zip"
+        with zipfile.ZipFile(f, 'w') as zf:
+            zi = zipfile.ZipInfo("0")
+            zi.extra = b"12345"
+            zf.writestr(zi, b"some string")
+        try:
+            with zipfile.ZipFile(f) as zf:
+                infolist = zf.infolist()
+                self.assertTrue(len(infolist) == 1)
+                self.assertEqual(infolist[0].extra, b"12345")
+        except zipfile.BadZipfile as e:
+            self.fail(f"Unexpected exception: {e}")
+
 class CommandLineTest(unittest.TestCase):
 
     def zipfilecmd(self, *args, **kwargs):

Lib/zipfile.py

@@ -1277,7 +1277,8 @@ def _RealGetContents(self):
 
         # "concat" is zero, unless zip was concatenated to another file
         concat = endrec[_ECD_LOCATION] - size_cd - offset_cd
-        if endrec[_ECD_SIGNATURE] == stringEndArchive64:
+        hasZip64ECD = endrec[_ECD_SIGNATURE] == stringEndArchive64
+        if hasZip64ECD:
             # If Zip64 extension structures are present, account for them
             concat -= (sizeEndCentDir64 + sizeEndCentDir64Locator)
 
@@ -1324,7 +1325,8 @@ def _RealGetContents(self):
             x.date_time = ( (d>>9)+1980, (d>>5)&0xF, d&0x1F,
                             t>>11, (t>>5)&0x3F, (t&0x1F) * 2 )
 
-            x._decodeExtra()
+            if hasZip64ECD:
+                x._decodeExtra()
             x.header_offset = x.header_offset + concat
             self.filelist.append(x)
             self.NameToInfo[x.filename] = x

Misc/NEWS.d/next/Library/2018-09-07-14-25-22.bpo-34606.1-64yh.rst

@@ -0,0 +1 @@
+Decode extra data only if Zip64 End of central directory record is present