GH-128131: Completely support random read access of uncompressed unencrypted files in ZipFile

Lib/test/test_zipfile/test_core.py

@@ -3447,6 +3447,53 @@ def test_too_short(self):
         self.assertEqual(
             b"zzz", zipfile._Extra.strip(b"zzz", (self.ZIP64_EXTRA,)))
 
+class StoredZipExtFileRandomAccessTest(unittest.TestCase):
+    def test_random_access(self):
+        from _pyio import BytesIO
+        class StatIO(BytesIO):
+            def __init__(self):
+                super().__init__()
+                self.bytes_read = 0
+
+            def read(self, size=-1):
+                bs = super().read(size)
+                self.bytes_read += len(bs)
+                return bs
+
+            def get_bytes_read(self):
+                return self.bytes_read
+
+        sio = StatIO()
+        # 100000 bytes
+        txt = b'0123456789'*10000
+
+        # Check seek on a file
+        with zipfile.ZipFile(sio, "w", compression=zipfile.ZIP_STORED) as zipf:
+            zipf.writestr("foo.txt", txt)
+
+        with zipfile.ZipFile(sio, "r") as zipf:
+            with zipf.open("foo.txt", "r") as fp:
+                br = sio.get_bytes_read()
+                fp.seek(50000, os.SEEK_CUR)
+                self.assertEqual(sio.get_bytes_read() - br, 0, 'seek produces redundant read!')
+
+                b = fp.read(100)
+                self.assertEqual(b, txt[:100])
+
+                # seek length must be greater than ZipExtFile.MIN_READ_SIZE (4096)
+                # backward seek
+                br = sio.get_bytes_read()
+                fp.seek(5000, os.SEEK_CUR)
+                b = fp.read(100)
+                self.assertEqual(b, txt[50000:50100])
+                self.assertLessEqual(sio.get_bytes_read() - br, 4096, 'read redundant bytes during backward seek!')
+
+                # forward seek
+                br = sio.get_bytes_read()
+                fp.seek(-40000, os.SEEK_CUR)
+                b = fp.read(100)
+                self.assertEqual(b, txt[10100:10200])
+                self.assertLessEqual(sio.get_bytes_read() - br, 4096, 'read redundant bytes during forward seek!')
 
 if __name__ == "__main__":
     unittest.main()

Lib/zipfile/__init__.py

@@ -1162,13 +1162,16 @@ def seek(self, offset, whence=os.SEEK_SET):
             self._offset = buff_offset
             read_offset = 0
         # Fast seek uncompressed unencrypted file
-        elif self._compress_type == ZIP_STORED and self._decrypter is None and read_offset > 0:
+        elif self._compress_type == ZIP_STORED and self._decrypter is None:
             # disable CRC checking after first seeking - it would be invalid
             self._expected_crc = None
             # seek actual file taking already buffered data into account
             read_offset -= len(self._readbuffer) - self._offset
             self._fileobj.seek(read_offset, os.SEEK_CUR)
             self._left -= read_offset
+            self._compress_left -= read_offset
+            if self._eof and read_offset < 0:
+                self._eof = False
             read_offset = 0
             # flush read buffer
             self._readbuffer = b''

Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst

@@ -0,0 +1 @@
+Support fast forward seek in uncompressed unencrypted :class:`!zipfile.ZipExtFile`.