fix(dynamodb): align size() type semantics with AWS

[vieiralucas]

Summary

• attribute_size returned string-length for N-type attributes, so size(count) = :ten with count = N("10") and :ten = N("10") evaluated to false ("10".len() == 2, not 10). That isn't AWS behavior: size() is defined only for S, B, SS, NS, BS, L, and M; on N, BOOL, or NULL it's a type mismatch that silently filters the row out in FilterExpression context.
• Drop the N and BOOL/NULL branches of attribute_size so size() returns None on those types, which flows to false in the filter evaluator.
• Unignore filter_functions_with_sdk_space_before_paren: switch count from N("10") to S("1234567890") so size() still returns 10 and exercises both the size(X) and size (X) spacings.
• Add filter_size_on_numeric_is_invalid locking the new semantics (size on N, BOOL, and a missing attribute all evaluate to false).

Test plan

• cargo test -p fakecloud-dynamodb --lib
• cargo test -p fakecloud-e2e --test dynamodb
• cargo clippy -p fakecloud-dynamodb --all-targets -- -D warnings
• cargo fmt --all

---

Summary by cubic

Fixes DynamoDB size() evaluation to match AWS: it’s only valid for S, B, SS, NS, BS, L, and M. On N, BOOL, or NULL it now evaluates to false in filters, avoiding incorrect matches.

• Bug Fixes
  • Removed N/BOOL/NULL handling in attribute_size so size() returns None for those types.
  • Unignored filter_functions_with_sdk_space_before_paren; switched count to S("1234567890") to keep size() = 10 for both size(X) and size (X).
  • Added filter_size_on_numeric_is_invalid to assert false for size() on N, BOOL, and missing attributes.

^{Written for commit 3fef3f6. Summary will update on new commits.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[cubic-dev-ai]

No issues found across 2 files