fix(idempotency): pass by value on idem key to guard inadvert mutations (#1090)

Co-authored-by: heitorlessa <[email protected]>

Author: ojongerius

aws_lambda_powertools/utilities/idempotency/base.py

@@ -1,5 +1,6 @@
 import logging
 from typing import Any, Callable, Dict, Optional, Tuple
+from copy import deepcopy
 
 from aws_lambda_powertools.utilities.idempotency.config import IdempotencyConfig
 from aws_lambda_powertools.utilities.idempotency.exceptions import (
@@ -69,7 +70,7 @@ def __init__(
             Function keyword arguments
         """
         self.function = function
-        self.data = _prepare_data(function_payload)
+        self.data = deepcopy(_prepare_data(function_payload))
         self.fn_args = function_args
         self.fn_kwargs = function_kwargs
 

tests/functional/idempotency/conftest.py

@@ -14,7 +14,8 @@
 from aws_lambda_powertools.utilities.idempotency.idempotency import IdempotencyConfig
 from aws_lambda_powertools.utilities.jmespath_utils import extract_data_from_envelope
 from aws_lambda_powertools.utilities.validation import envelopes
-from tests.functional.utils import hash_idempotency_key, json_serialize, load_event
+from tests.functional.idempotency.utils import hash_idempotency_key
+from tests.functional.utils import json_serialize, load_event
 
 TABLE_NAME = "TEST_TABLE"
 

tests/functional/idempotency/test_idempotency.py

@@ -22,7 +22,12 @@
 from aws_lambda_powertools.utilities.idempotency.idempotency import idempotent, idempotent_function
 from aws_lambda_powertools.utilities.idempotency.persistence.base import BasePersistenceLayer, DataRecord
 from aws_lambda_powertools.utilities.validation import envelopes, validator
-from tests.functional.utils import hash_idempotency_key, json_serialize, load_event
+from tests.functional.idempotency.utils import (
+    build_idempotency_put_item_stub,
+    build_idempotency_update_item_stub,
+    hash_idempotency_key,
+)
+from tests.functional.utils import json_serialize, load_event
 
 TABLE_NAME = "TEST_TABLE"
 
@@ -275,6 +280,40 @@ def lambda_handler(event, context):
     stubber.deactivate()
 
 
+@pytest.mark.parametrize("idempotency_config", [{"use_local_cache": True, "event_key_jmespath": "body"}], indirect=True)
+def test_idempotent_lambda_first_execution_event_mutation(
+    idempotency_config: IdempotencyConfig,
+    persistence_store: DynamoDBPersistenceLayer,
+    lambda_apigw_event,
+    lambda_response,
+    lambda_context,
+):
+    """
+    Test idempotent decorator where lambda_handler mutates the event.
+    Ensures we're passing data by value, not reference.
+    """
+    event = copy.deepcopy(lambda_apigw_event)
+    stubber = stub.Stubber(persistence_store.table.meta.client)
+    ddb_response = {}
+    stubber.add_response("put_item", ddb_response, build_idempotency_put_item_stub(data=event["body"]))
+    stubber.add_response(
+        "update_item",
+        ddb_response,
+        build_idempotency_update_item_stub(data=event["body"], handler_response=lambda_response),
+    )
+    stubber.activate()
+
+    @idempotent(config=idempotency_config, persistence_store=persistence_store)
+    def lambda_handler(event, context):
+        event.pop("body")  # remove exact key we're using for idempotency
+        return lambda_response
+
+    lambda_handler(event, lambda_context)
+
+    stubber.assert_no_pending_responses()
+    stubber.deactivate()
+
+
 @pytest.mark.parametrize("idempotency_config", [{"use_local_cache": False}, {"use_local_cache": True}], indirect=True)
 def test_idempotent_lambda_expired(
     idempotency_config: IdempotencyConfig,

tests/functional/idempotency/utils.py

@@ -0,0 +1,45 @@
+import hashlib
+from typing import Any, Dict
+
+from botocore import stub
+
+from tests.functional.utils import json_serialize
+
+
+def hash_idempotency_key(data: Any):
+    """Serialize data to JSON, encode, and hash it for idempotency key"""
+    return hashlib.md5(json_serialize(data).encode()).hexdigest()
+
+
+def build_idempotency_put_item_stub(
+    data: Dict, function_name: str = "test-func", handler_name: str = "lambda_handler"
+) -> Dict:
+    idempotency_key_hash = f"{function_name}.{handler_name}#{hash_idempotency_key(data)}"
+    return {
+        "ConditionExpression": "attribute_not_exists(#id) OR #now < :now",
+        "ExpressionAttributeNames": {"#id": "id", "#now": "expiration"},
+        "ExpressionAttributeValues": {":now": stub.ANY},
+        "Item": {"expiration": stub.ANY, "id": idempotency_key_hash, "status": "INPROGRESS"},
+        "TableName": "TEST_TABLE",
+    }
+
+
+def build_idempotency_update_item_stub(
+    data: Dict,
+    handler_response: Dict,
+    function_name: str = "test-func",
+    handler_name: str = "lambda_handler",
+) -> Dict:
+    idempotency_key_hash = f"{function_name}.{handler_name}#{hash_idempotency_key(data)}"
+    serialized_lambda_response = json_serialize(handler_response)
+    return {
+        "ExpressionAttributeNames": {"#expiry": "expiration", "#response_data": "data", "#status": "status"},
+        "ExpressionAttributeValues": {
+            ":expiry": stub.ANY,
+            ":response_data": serialized_lambda_response,
+            ":status": "COMPLETED",
+        },
+        "Key": {"id": idempotency_key_hash},
+        "TableName": "TEST_TABLE",
+        "UpdateExpression": "SET #response_data = :response_data, " "#expiry = :expiry, #status = :status",
+    }

tests/functional/utils.py

@@ -1,5 +1,4 @@
 import base64
-import hashlib
 import json
 from pathlib import Path
 from typing import Any
@@ -22,8 +21,3 @@ def b64_to_str(data: str) -> str:
 
 def json_serialize(data):
     return json.dumps(data, sort_keys=True, cls=Encoder)
-
-
-def hash_idempotency_key(data: Any):
-    """Serialize data to JSON, encode, and hash it for idempotency key"""
-    return hashlib.md5(json_serialize(data).encode()).hexdigest()