haive.core.engine.document.loaders.registry¶

Document loader registry system.

This module provides a registry for document loaders, allowing them to be registered, looked up, and managed throughout the application.

Classes¶

`DocumentLoaderRegistry`	Registry for document loaders.
`LoaderMetadata`	Metadata for a document loader.

Functions¶

`create_loader`(loader_name, **kwargs)	Create a loader instance by name.
`get_default_registry`()	Get the default document loader registry.
`get_loader`(loader_name)	Get a loader by name from the default registry.
`register_loader`(source_type[, name, description, ...])	Decorator to register a document loader.

Module Contents¶

class haive.core.engine.document.loaders.registry.DocumentLoaderRegistry[source]¶

Bases: haive.core.registry.base.AbstractRegistry[type[langchain_core.document_loaders.base.BaseLoader]]

Registry for document loaders.

This registry keeps track of document loader classes and their metadata, allowing for discovery and instantiation of loaders based on source types.

Initialize the registry with empty storage.

clear()[source]¶

Clear all registrations.

Return type:: None

find_by_id(id)[source]¶

Find a loader by name (used for compatibility with AbstractRegistry).

Parameters:: id (str) – Loader name
Returns:: Loader class if found, None otherwise
Return type:: type[langchain_core.document_loaders.base.BaseLoader] | None

find_by_name(name)[source]¶

Find a loader by name.

Parameters:: name (str) – Loader name
Returns:: Loader class if found, None otherwise
Return type:: type[langchain_core.document_loaders.base.BaseLoader] | None

find_loader_for_file(file_path)[source]¶

Find loaders that can handle a specific file extension.

Parameters:: file_path (str) – Path to the file
Returns:: List of loader classes that can handle this file
Return type:: list[type[langchain_core.document_loaders.base.BaseLoader]]

find_loader_for_url(url)[source]¶

Find loaders that can handle a specific URL pattern.

Parameters:: url (str) – URL to handle
Returns:: List of loader classes that can handle this URL
Return type:: list[type[langchain_core.document_loaders.base.BaseLoader]]

get(item_type, name)[source]¶

Get a loader by source type and name.

Parameters:

item_type (haive.core.engine.document.loaders.sources.source_types.SourceCategory) – Source type
name (str) – Loader name

Returns:

Loader class if found, None otherwise

Return type:

type[langchain_core.document_loaders.base.BaseLoader] | None

get_all(item_type)[source]¶

Get all loaders for a specific source type.

Parameters:: item_type (haive.core.engine.document.loaders.sources.source_types.SourceCategory) – Source type
Returns:: Dictionary mapping loader names to loader classes
Return type:: dict[str, type[langchain_core.document_loaders.base.BaseLoader]]

get_all_metadata()[source]¶

Get metadata for all registered loaders.

Returns:: Dictionary mapping loader names to metadata
Return type:: dict[str, LoaderMetadata]

classmethod get_instance()[source]¶

Get the singleton instance of the registry.

Return type:: DocumentLoaderRegistry

get_metadata(name)[source]¶

Get metadata for a specific loader.

Parameters:: name (str) – Loader name
Returns:: Loader metadata if found, None otherwise
Return type:: LoaderMetadata | None

list(item_type)[source]¶

List all loader names for a specific source type.

Parameters:: item_type (haive.core.engine.document.loaders.sources.source_types.SourceCategory) – Source type
Returns:: List of loader names
Return type:: list[str]

register(loader_class, metadata)[source]¶

Parameters:

loader_class (type[langchain_core.document_loaders.base.BaseLoader]) – Loader class to register
metadata (LoaderMetadata) – Metadata for the loader

Returns:

The registered loader class

Return type:

type[langchain_core.document_loaders.base.BaseLoader]

class haive.core.engine.document.loaders.registry.LoaderMetadata(/, **data)[source]¶

Bases: pydantic.BaseModel

Metadata for a document loader.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:: data (Any)

model_config¶: Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].

haive.core.engine.document.loaders.registry.create_loader(loader_name, **kwargs)[source]¶

Create a loader instance by name.

Parameters:: loader_name (str)
Return type:: langchain_core.document_loaders.base.BaseLoader | None

haive.core.engine.document.loaders.registry.get_default_registry()[source]¶

Get the default document loader registry.

Return type:: DocumentLoaderRegistry

haive.core.engine.document.loaders.registry.get_loader(loader_name)[source]¶

Get a loader by name from the default registry.

Parameters:: loader_name (str)
Return type:: type[langchain_core.document_loaders.base.BaseLoader] | None

haive.core.engine.document.loaders.registry.register_loader(source_type, name=None, description=None, requires_async=False, file_extensions=None, url_patterns=None, config_schema=None)[source]¶

Decorator to register a document loader.

Parameters:

source_type (haive.core.engine.document.loaders.sources.source_types.SourceCategory) – Type of source this loader handles
name (str | None) – Optional custom name for the loader
description (str | None) – Optional description of the loader
requires_async (bool) – Whether this loader requires async operations
file_extensions (list[str] | None) – List of file extensions this loader can handle
url_patterns (list[str] | None) – List of URL patterns this loader can handle
config_schema (type[pydantic.BaseModel] | None) – Optional Pydantic model for configuration

Returns:

Decorator function

Return type:

collections.abc.Callable[[type[langchain_core.document_loaders.base.BaseLoader]], type[langchain_core.document_loaders.base.BaseLoader]]