Tech debt: Fix falsy response handling, inconsistencies, and duplicated code in idempotency utility

[leandrodamascena]

Why is this needed?

Found a few issues during code review of the idempotency utility:

1. Falsy responses are silently discarded (base.py:299, 273-274)

if response else None treats valid falsy return values (0, False, {}, [], "") as None. This means a function returning 0 or False won't have its response stored, and on replay it will re-execute instead of returning the cached value.

# Current
serialized_response: dict = self.output_serializer.to_dict(response) if response else None

# Should be
serialized_response: dict = self.output_serializer.to_dict(response) if response is not None else None

Same issue on line 273-274 when reading cached responses.

2. Inconsistent status constant usage (base.py:170)

Uses string literal "INPROGRESS" instead of STATUS_CONSTANTS["INPROGRESS"].

3. str(None) produces "None" string in Redis persistence (redis.py:335-336)

response_data=str(item.get(self.data_attr)),          # str(None) = "None"
payload_hash=str(item.get(self.validation_key_attr)),  # str(None) = "None"

When the value is missing from Redis, this creates the string "None" instead of actual None.

4. Duplicated null-check for idempotency key (persistence/base.py)

The same 4-line block is copy-pasted in save_success, save_inprogress, delete_record, and get_record:

idempotency_key = self._get_hashed_idempotency_key(data=data)
if idempotency_key is None:
    # If the idempotency key is None, no data will be saved in the Persistence Layer.
    # See: https://github.com/aws-powertools/powertools-lambda-python/issues/2465
    return None

Can be extracted into a single helper method to reduce duplication.

Which area does this relate to?

Idempotency

Suggestion

Fix this.

Acknowledgment

• This request meets Powertools for AWS Lambda (Python) Tenets
• Should this be considered in other Powertools for AWS Lambda languages? i.e. Java, TypeScript, and .NET