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

Lib/test/test_zipfile/test_core.py

@@ -3447,6 +3447,62 @@ def test_too_short(self):
         self.assertEqual(
             b"zzz", zipfile._Extra.strip(b"zzz", (self.ZIP64_EXTRA,)))
 
+class StoredZipExtFileRandomReadTest(unittest.TestCase):
+    def test_random_read(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
+
+        sio = StatIO()
+        # 20000 bytes
+        txt = b'0123456789' * 2000
+
+        # seek length must be greater than ZipExtFile.MIN_READ_SIZE (4096)
+        min_size = zipfile.ZipExtFile.MIN_READ_SIZE
+        self.assertGreaterEqual(min_size, 100)
+
+        with zipfile.ZipFile(sio, "w", compression=zipfile.ZIP_STORED) as zipf:
+            zipf.writestr("foo.txt", txt)
+
+        # check random seek and read on a file
+        with zipfile.ZipFile(sio, "r") as zipf:
+            with zipf.open("foo.txt", "r") as fp:
+                # forward seek
+                old_count = sio.bytes_read
+                fp.seek(10002, os.SEEK_CUR)
+                self.assertEqual(fp.tell(), 10002)
+                self.assertEqual(fp._left, fp._compress_left)
+                arr = fp.read(100)
+                self.assertEqual(fp.tell(), 10102)
+                self.assertEqual(arr, txt[10002:10102])
+                self.assertEqual(fp._left, fp._compress_left)
+                d = sio.bytes_read - old_count
+                self.assertLessEqual(d, min_size)
+
+                # backward seek
+                old_count = sio.bytes_read
+                fp.seek(-5003, os.SEEK_CUR)
+                self.assertEqual(fp.tell(), 5099)
+                self.assertEqual(fp._left, fp._compress_left)
+                arr = fp.read(100)
+                self.assertEqual(fp.tell(), 5199)
+                self.assertEqual(arr, txt[5099:5199])
+                self.assertEqual(fp._left, fp._compress_left)
+                d = sio.bytes_read - old_count
+                self.assertLessEqual(d, min_size)
+
+                # eof flags test
+                fp.seek(0, os.SEEK_END)
+                self.assertTrue(fp._eof)
+                fp.seek(12345, os.SEEK_SET)
+                self.assertFalse(fp._eof)
 
 if __name__ == "__main__":
     unittest.main()

Lib/zipfile/__init__.py

@@ -1162,13 +1162,15 @@ 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
+            self._eof = self._left <= 0
             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 +1,2 @@
+Completely support random access of uncompressed unencrypted read-only
+zip files obtained by :meth:`ZipFile.open <zipfile.ZipFile.open>`.