feat: add should_complete to CompletionConfig

[ayushiahjolia]

Issue #, if available: #519

Description of changes:
Adds a should_complete predicate to CompletionConfig for map and parallel operations, giving full control over when a batch completes early.

from aws_durable_execution_sdk_python import complete_batch, continue_batch

config = CompletionConfig(
    should_complete=lambda status: (
        complete_batch() if status.success_count >= 2
        else continue_batch()
    )
)

The predicate receives a CompletionStatus snapshot (counts + per-item statuses) and returns a CompletionDecision - either continue_batch() or complete_batch(outcome). The outcome determines whether CUSTOM_COMPLETION_SUCCEEDED or CUSTOM_COMPLETION_FAILED is reported.

Key design decisions -

• Cannot be combined with threshold fields (raises ValidationError)
• Predicate must be deterministic and side-effect-free
• Replay uses the checkpointed decision, never re-invokes the predicate
• On resumed invocations, deferred until prior checkpoints replay

Testing -

• Unit tests, integration tests and examples

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[yaythomas]

just two minor noin-func nits :)

[github-actions]

Codex AI review

Two actionable correctness findings in the custom completion coordinator. Static review only; tests were not executed per request.

Reviewed commit 1dd230e62c595c09baf233e449b849cf0a59ed44. Workflow run