Close #17666: Fix reading gzip files with an extra field.

3 files changed, 11 insertions, 1 deletions

diff --git a/Lib/gzip.py b/Lib/gzip.py
index d5b5743ff6..d2d6dd6531 100644
--- a/Lib/gzip.py
+++ b/Lib/gzip.py
@@ -302,7 +302,8 @@ class GzipFile(io.BufferedIOBase):
 
         if flag & FEXTRA:
             # Read & discard the extra field, if present
-            self._read_exact(struct.unpack("<H", self._read_exact(2)))
+            extra_len, = struct.unpack("<H", self._read_exact(2))
+            self._read_exact(extra_len)
         if flag & FNAME:
             # Read and discard a null-terminated string containing the filename
             while True:
diff --git a/Lib/test/test_gzip.py b/Lib/test/test_gzip.py
index ebd4c43855..37fe8538a7 100644
--- a/Lib/test/test_gzip.py
+++ b/Lib/test/test_gzip.py
@@ -403,6 +403,13 @@ class TestGzip(BaseTest):
             with gzip.GzipFile(fileobj=io.BytesIO(truncated[:i])) as f:
                 self.assertRaises(EOFError, f.read, 1)
 
+    def test_read_with_extra(self):
+        # Gzip data with an extra field
+        gzdata = (b'\x1f\x8b\x08\x04\xb2\x17cQ\x02\xff'
+                  b'\x05\x00Extra'
+                  b'\x0bI-.\x01\x002\xd1Mx\x04\x00\x00\x00')
+        with gzip.GzipFile(fileobj=io.BytesIO(gzdata)) as f:
+            self.assertEqual(f.read(), b'Test')
 
 class TestOpen(BaseTest):
     def test_binary_modes(self):
diff --git a/Misc/NEWS b/Misc/NEWS
index 50bd50ead8..203d5cd988 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -30,6 +30,8 @@ Core and Builtins
 Library
 -------
 
+- Issue #17666: Fix reading gzip files with an extra field.
+
 - Issue #16475: Support object instancing, recursion and interned strings
   in marshal