S3 Driver requires hash algorithm and value on claim payloads (#1443)

Author: jmaeagle99

temporalio/contrib/aws/s3driver/_driver.py

@@ -188,23 +188,28 @@ async def _download(claim: StorageDriverClaim) -> Payload:
                     f"S3StorageDriver retrieve failed [bucket={bucket}, key={key}]"
                 ) from e
 
-            expected_hash = claim.claim_data.get("hash_value")
             hash_algorithm = claim.claim_data.get("hash_algorithm")
-            if expected_hash and hash_algorithm:
-                if hash_algorithm != "sha256":
-                    raise ValueError(
-                        f"S3StorageDriver unsupported hash algorithm "
-                        f"[bucket={bucket}, key={key}]: "
-                        f"expected sha256, got {hash_algorithm}"
-                    )
-                actual_hash = hashlib.sha256(payload_bytes).hexdigest().lower()
-                if actual_hash != expected_hash:
-                    raise ValueError(
-                        f"S3StorageDriver integrity check failed "
-                        f"[bucket={bucket}, key={key}]: "
-                        f"expected {hash_algorithm}:{expected_hash}, "
-                        f"got {hash_algorithm}:{actual_hash}"
-                    )
+            expected_hash = claim.claim_data.get("hash_value")
+            if not hash_algorithm or not expected_hash:
+                raise ValueError(
+                    f"S3StorageDriver claim is missing required content hash information "
+                    f"[bucket={bucket}, key={key}]: "
+                    f"claim_data must contain 'hash_algorithm' and 'hash_value'"
+                )
+            if hash_algorithm != "sha256":
+                raise ValueError(
+                    f"S3StorageDriver unsupported hash algorithm "
+                    f"[bucket={bucket}, key={key}]: "
+                    f"expected sha256, got {hash_algorithm}"
+                )
+            actual_hash = hashlib.sha256(payload_bytes).hexdigest().lower()
+            if actual_hash != expected_hash:
+                raise ValueError(
+                    f"S3StorageDriver integrity check failed "
+                    f"[bucket={bucket}, key={key}]: "
+                    f"expected {hash_algorithm}:{expected_hash}, "
+                    f"got {hash_algorithm}:{actual_hash}"
+                )
 
             payload = Payload()
             payload.ParseFromString(payload_bytes)

tests/contrib/aws/s3driver/test_s3driver.py

@@ -489,7 +489,7 @@ async def test_retrieve_rejects_unsupported_hash_algorithm(
     async def test_retrieve_without_hash_in_claim(
         self, driver_client: S3StorageDriverClient
     ) -> None:
-        """Claims without hash fields still retrieve successfully (backward compat)."""
+        """Claims missing content hash fields raise ValueError on retrieve."""
         driver = S3StorageDriver(client=driver_client, bucket=BUCKET)
         payload = make_payload("no-hash-claim")
         [claim] = await driver.store(make_store_context(), [payload])
@@ -500,10 +500,11 @@ async def test_retrieve_without_hash_in_claim(
                 "key": claim.claim_data["key"],
             },
         )
-        [retrieved] = await driver.retrieve(
-            StorageDriverRetrieveContext(), [legacy_claim]
-        )
-        assert retrieved == payload
+        with pytest.raises(
+            ValueError,
+            match=r"S3StorageDriver claim is missing required content hash information",
+        ):
+            await driver.retrieve(StorageDriverRetrieveContext(), [legacy_claim])
 
 
 # ---------------------------------------------------------------------------