- Share regex input preparation and bound capture storage to each result. - Batch replacement output and avoid rebuilding unchanged strings. - Accelerate case conversion and normalization while preserving Unicode and ownership behavior.
