parse() — langchain Function Reference

Architecture documentation for the parse() function in output_parser.py from the langchain codebase.

Function python AgentOrchestration ToolInterface

Entity Profile

AgentOrchestration→ ToolInterface→ parse() — langchain Function Reference

Dependency Diagram

graph TD
  ea846c6f_225c_b26e_1924_b844031769a0["parse()"]
  c64d0531_c96b_8590_370b_7ce9e34ea9b9["ReActOutputParser"]
  ea846c6f_225c_b26e_1924_b844031769a0 -->|defined in| c64d0531_c96b_8590_370b_7ce9e34ea9b9
  style ea846c6f_225c_b26e_1924_b844031769a0 fill:#6366f1,stroke:#818cf8,color:#fff

Relationship Graph

Source Code

libs/langchain/langchain_classic/agents/react/output_parser.py lines 14–30

    def parse(self, text: str) -> AgentAction | AgentFinish:
        action_prefix = "Action: "
        if not text.strip().split("\n")[-1].startswith(action_prefix):
            msg = f"Could not parse LLM Output: {text}"
            raise OutputParserException(msg)
        action_block = text.strip().split("\n")[-1]

        action_str = action_block[len(action_prefix) :]
        # Parse out the action and the directive.
        re_matches = re.search(r"(.*?)\[(.*?)\]", action_str)
        if re_matches is None:
            msg = f"Could not parse action directive: {action_str}"
            raise OutputParserException(msg)
        action, action_input = re_matches.group(1), re_matches.group(2)
        if action == "Finish":
            return AgentFinish({"output": action_input}, text)
        return AgentAction(action, action_input, text)