Using Langchain Output Parsers To Get What You Want Out Of Llms