StructuredChatOutputParser Class — langchain Architecture

Architecture documentation for the StructuredChatOutputParser class in output_parser.py from the langchain codebase.

Class python

Entity Profile

Dependency Diagram

graph TD
  2a95d029_9ce2_559a_374e_c9256ca9f87e["StructuredChatOutputParser"]
  d9685a4b_26b0_bca9_f857_03bb2ffc9dd4["AgentOutputParser"]
  2a95d029_9ce2_559a_374e_c9256ca9f87e -->|extends| d9685a4b_26b0_bca9_f857_03bb2ffc9dd4
  2617733a_ade1_11ca_e0e5_eb0cecb70df6["output_parser.py"]
  2a95d029_9ce2_559a_374e_c9256ca9f87e -->|defined in| 2617733a_ade1_11ca_e0e5_eb0cecb70df6
  8ae1b605_5f96_9c45_9d6e_e0bd57a87dd3["get_format_instructions()"]
  2a95d029_9ce2_559a_374e_c9256ca9f87e -->|method| 8ae1b605_5f96_9c45_9d6e_e0bd57a87dd3
  c8b3b8b6_c924_2b06_cfd4_9d645b98239b["parse()"]
  2a95d029_9ce2_559a_374e_c9256ca9f87e -->|method| c8b3b8b6_c924_2b06_cfd4_9d645b98239b
  c5bb2aea_f732_f6d1_d7de_99e5d9bf8826["_type()"]
  2a95d029_9ce2_559a_374e_c9256ca9f87e -->|method| c5bb2aea_f732_f6d1_d7de_99e5d9bf8826

Relationship Graph

Source Code

libs/langchain/langchain_classic/agents/structured_chat/output_parser.py lines 21–59

class StructuredChatOutputParser(AgentOutputParser):
    """Output parser for the structured chat agent."""

    format_instructions: str = FORMAT_INSTRUCTIONS
    """Default formatting instructions"""

    pattern: Pattern = re.compile(r"```(?:json\s+)?(\W.*?)```", re.DOTALL)
    """Regex pattern to parse the output."""

    @override
    def get_format_instructions(self) -> str:
        """Returns formatting instructions for the given output parser."""
        return self.format_instructions

    @override
    def parse(self, text: str) -> AgentAction | AgentFinish:
        try:
            action_match = self.pattern.search(text)
            if action_match is not None:
                response = json.loads(action_match.group(1).strip(), strict=False)
                if isinstance(response, list):
                    # gpt turbo frequently ignores the directive to emit a single action
                    logger.warning("Got multiple action responses: %s", response)
                    response = response[0]
                if response["action"] == "Final Answer":
                    return AgentFinish({"output": response["action_input"]}, text)
                return AgentAction(
                    response["action"],
                    response.get("action_input", {}),
                    text,
                )
            return AgentFinish({"output": text}, text)
        except Exception as e:
            msg = f"Could not parse LLM output: {text}"
            raise OutputParserException(msg) from e

    @property
    def _type(self) -> str:
        return "structured_chat"