Fix hash caching in ObjectStoragePath (#37769)

apache · Feb 28, 2024 · 382a961 · 382a961
1 parent bee0a7b
commit 382a961
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 3 deletions.
diff --git a/airflow/io/path.py b/airflow/io/path.py
@@ -17,7 +17,6 @@
 from __future__ import annotations
 
 import contextlib
-import functools
 import os
 import shutil
 import typing
@@ -152,9 +151,10 @@ def __new__(
 
         return cls._from_parts(args_list, url=parsed_url, conn_id=conn_id, **kwargs)  # type: ignore
 
-    @functools.lru_cache
     def __hash__(self) -> int:
-        return hash(str(self))
+        if not (_hash := getattr(self, "_hash", None)):
+            self._hash = _hash = hash(str(self))
+        return _hash
 
     def __eq__(self, other: typing.Any) -> bool:
         return self.samestore(other) and str(self) == str(other)

diff --git a/tests/io/test_path.py b/tests/io/test_path.py
@@ -318,3 +318,12 @@ def test_dataset(self):
         o = ObjectStoragePath(i)
         assert o.protocol == p
         assert o.path == f
+
+    def test_hash(self):
+        file_uri_1 = f"file:///tmp/{str(uuid.uuid4())}"
+        file_uri_2 = f"file:///tmp/{str(uuid.uuid4())}"
+        s = set()
+        for _ in range(10):
+            s.add(ObjectStoragePath(file_uri_1))
+            s.add(ObjectStoragePath(file_uri_2))
+        assert len(s) == 2