CallbackHandlerWithInputCapture Class — langchain Architecture

Architecture documentation for the CallbackHandlerWithInputCapture class in test_tools.py from the langchain codebase.

Class python

Entity Profile

Dependency Diagram

graph TD
  32dafc1f_47b3_ccf8_3012_9602da6cebe3["CallbackHandlerWithInputCapture"]
  9659be4f_91d1_3b62_bcfb_38c048777165["FakeCallbackHandler"]
  32dafc1f_47b3_ccf8_3012_9602da6cebe3 -->|extends| 9659be4f_91d1_3b62_bcfb_38c048777165
  8e7836ae_e72c_f670_72a5_4ca6d46e3555["test_tools.py"]
  32dafc1f_47b3_ccf8_3012_9602da6cebe3 -->|defined in| 8e7836ae_e72c_f670_72a5_4ca6d46e3555
  d42476ff_6b17_148a_8f54_3942b7d71a6e["on_tool_start()"]
  32dafc1f_47b3_ccf8_3012_9602da6cebe3 -->|method| d42476ff_6b17_148a_8f54_3942b7d71a6e

Relationship Graph

Source Code

libs/core/tests/unit_tests/test_tools.py lines 2963–2991

class CallbackHandlerWithInputCapture(FakeCallbackHandler):
    """Callback handler that captures inputs passed to on_tool_start."""

    captured_inputs: list[dict | None] = Field(default_factory=list)

    def on_tool_start(
        self,
        serialized: dict[str, Any],
        input_str: str,
        *,
        run_id: Any,
        parent_run_id: Any | None = None,
        tags: list[str] | None = None,
        metadata: dict[str, Any] | None = None,
        inputs: dict[str, Any] | None = None,
        **kwargs: Any,
    ) -> Any:
        """Capture the inputs passed to on_tool_start."""
        self.captured_inputs.append(inputs)
        return super().on_tool_start(
            serialized,
            input_str,
            run_id=run_id,
            parent_run_id=parent_run_id,
            tags=tags,
            metadata=metadata,
            inputs=inputs,
            **kwargs,
        )