Text this: Evaluation of Alignment Between Large Language Models and Expert Clinicians in Suicide Risk Assessment.