refactor(dynamodb): split evaluate_single_key_condition + find_partiql_where_indices

[vieiralucas]

Summary

• `evaluate_single_key_condition` split into `key_cond_begins_with` / `key_cond_between` / `key_cond_simple_comparison`. The dispatch function is now a 12-line guard ladder. The `key_cond_` prefix avoids collision with the existing `eval_begins_with` that backs the FilterExpression evaluator — the two share the function-call grammar but have different return types and operand-type handling, so they intentionally stay separate.
• `find_partiql_where_indices` split into `split_partiql_and_clauses` (the case-insensitive AND splitter) and `parse_partiql_equality_conditions` (the `col = literal` pair extractor). The outer function is now a 12-line filter loop.

Already-clean targets from the original audit list (judged not worth splitting)

• `parse_path_segments` (40L): single-pass char state machine; the bracket-index sub-block is short and reading it inline matches the surrounding parser style.
• `apply_update_expression` (40L): already a clean dispatch over the `UpdateAction` enum that delegates to per-action helpers (`apply_set_assignment`, `apply_add_assignment`, `apply_delete_assignment`).
• `parse_update_clauses` (35L): focused on keyword-position scan + segment slicing; can't be split without rearranging the algorithm.
• `assign_list_index`: short, one job (replace-or-append at index, validate type).
• `evaluate_size_comparison` (45L): two clear sub-tasks already inline (parse `size(path) op :val`, evaluate against the resolved attribute size).

Test plan

• `cargo fmt`
• `cargo clippy -p fakecloud-dynamodb --all-targets -- -D warnings`
• `cargo test -p fakecloud-dynamodb --lib` — 92 passed
• `cargo test --workspace --exclude fakecloud-e2e --exclude fakecloud-conformance`

---

Summary by cubic

Refactors DynamoDB expression evaluators by splitting two large functions into small, focused helpers with no behavior changes. This simplifies key-condition and PartiQL WHERE processing and makes them easier to test.

• Refactors
  • Split evaluate_single_key_condition into key_cond_begins_with, key_cond_between, and key_cond_simple_comparison; dispatch is now a small guard ladder and preserves S-only begins_with and correct operator matching.
  • Split find_partiql_where_indices into split_partiql_and_clauses (case-insensitive AND) and parse_partiql_equality_conditions (col = literal/param); the outer function is now a short filter loop and keeps prior drop-on-parse-fail behavior.

^{Written for commit 5a55fe1. Summary will update on new commits.}

[cubic-dev-ai]

No issues found across 1 file