haive.core.config.runnable

Configuration management for Haive runnables.

This module provides utilities for creating, managing, and manipulating runtime configurations for Haive engines and runnables. It handles parameter management, metadata tracking, and configuration merging.

The main class RunnableConfigManager provides a comprehensive set of static methods for working with RunnableConfig objects, which are used to configure the behavior of engines at runtime.

Classes:

RunnableConfigManager: Static utility class for managing runnable configurations

Example:

Examples

>>> # Create a basic config with thread tracking
>>> config = RunnableConfigManager.create(
>>> thread_id="123",
>>> user_id="user_456"
>>> )
>>>
>>> # Add engine-specific configuration
>>> config = RunnableConfigManager.add_engine_config(
>>> config,
>>> "my_llm",
>>> temperature=0.7,
>>> max_tokens=100
>>> )

Classes

RunnableConfigManager

Enhanced manager for creating and manipulating RunnableConfig objects.

Module Contents

class haive.core.config.runnable.RunnableConfigManager

Enhanced manager for creating and manipulating RunnableConfig objects.

Provides methods for creating standardized configs, extracting values, and managing engine-specific configurations.

static add_engine(config, engine)

Add an engine’s parameters to the RunnableConfig.

Parameters:

• config (langchain_core.runnables.RunnableConfig) – RunnableConfig to add to

• engine (Any) – Engine to add

Returns:

Updated RunnableConfig

Return type:

langchain_core.runnables.RunnableConfig

static add_engine_config(config, engine_name, **params)

Add engine-specific configuration.

Parameters:

• config (langchain_core.runnables.RunnableConfig) – RunnableConfig to add to

• engine_name (str) – Name of the engine to add config for

• **params – Configuration parameters for the engine

Returns:

Updated RunnableConfig

Return type:

langchain_core.runnables.RunnableConfig

static create(thread_id=None, user_id=None, **kwargs)

Create a standardized RunnableConfig with common parameters.

Parameters:

• thread_id (str | None) – Optional thread ID for persistence (generated if not provided)

• user_id (str | None) – Optional user ID for attribution and permissions

• **kwargs – Additional parameters to include in configurable section

Returns:

A properly structured RunnableConfig

Return type:

langchain_core.runnables.RunnableConfig

static create_with_engine(engine, thread_id=None, user_id=None, **kwargs)

Create a RunnableConfig with engine parameters auto-populated.

Parameters:

• engine (Any) – Engine object to extract params from

• thread_id (str | None) – Optional thread ID

• user_id (str | None) – Optional user ID

• **kwargs – Additional configurable parameters

Returns:

RunnableConfig with engine parameters

Return type:

langchain_core.runnables.RunnableConfig

static create_with_metadata(metadata, thread_id=None, user_id=None, **kwargs)

Create a RunnableConfig with metadata.

Useful for tracing, logging, and debugging.

Parameters:

• metadata (dict[str, Any]) – Dictionary of metadata values

• thread_id (str | None) – Optional thread ID

• user_id (str | None) – Optional user ID

• **kwargs – Additional configurable parameters

Returns:

RunnableConfig with metadata section

Return type:

langchain_core.runnables.RunnableConfig

static extract_engine_config(config, engine_name)

Extract engine-specific configuration.

Parameters:

• config (langchain_core.runnables.RunnableConfig) – RunnableConfig to extract from

• engine_name (str) – Name of the engine to extract config for

Returns:

Engine-specific configuration dictionary

Return type:

dict[str, Any]

static extract_engine_type_config(config, engine_type)

Extract configuration for a specific engine type.

Parameters:

• config (langchain_core.runnables.RunnableConfig) – RunnableConfig to extract from

• engine_type (str) – The engine type (e.g., “llm”, “retriever”)

Returns:

Configuration for the engine type

Return type:

dict[str, Any]

static extract_value(config, key, default=None)

Extract a value from RunnableConfig’s configurable section.

Parameters:

• config (langchain_core.runnables.RunnableConfig) – RunnableConfig to extract from

• key (str) – Key to extract

• default (Any) – Default value if key not found

Returns:

Extracted value or default

Return type:

Any

static from_dict(input_dict)

Create a RunnableConfig from a dictionary.

Parameters:

input_dict (dict[str, Any]) – Dictionary to convert

Returns:

Properly structured RunnableConfig

Return type:

langchain_core.runnables.RunnableConfig

static from_model(model)

Create a RunnableConfig from a Pydantic model.

Parameters:

model (pydantic.BaseModel) – Pydantic model to convert

Returns:

Properly structured RunnableConfig

Return type:

langchain_core.runnables.RunnableConfig

static get_thread_id(config)

Extract thread_id from a RunnableConfig.

Parameters:

config (langchain_core.runnables.RunnableConfig) – RunnableConfig to extract from

Returns:

thread_id if present, otherwise None

Return type:

str | None

static get_user_id(config)

Extract user_id from a RunnableConfig.

Parameters:

config (langchain_core.runnables.RunnableConfig) – RunnableConfig to extract from

Returns:

user_id if present, otherwise None

Return type:

str | None

static merge(base, override)

Merge two RunnableConfigs, with override taking precedence.

Parameters:

• base (langchain_core.runnables.RunnableConfig) – Base configuration

• override (langchain_core.runnables.RunnableConfig) – Configuration that takes precedence

Returns:

Merged configuration

Return type:

langchain_core.runnables.RunnableConfig

static to_model(config, model_cls)

Convert a RunnableConfig to a Pydantic model.

Parameters:

• config (langchain_core.runnables.RunnableConfig) – RunnableConfig to convert

• model_cls (type[pydantic.BaseModel]) – Pydantic model class to convert to

Returns:

Instantiated model

Return type:

pydantic.BaseModel