API Reference#

This chapter documents the extension points of metadata‑crawler. It describes how to implement custom storage backends, how to add new index systems, and how to provide CLI extensions. The public API consists of abstract base classes and helper functions defined in storage_backend.py, metadata_stores.py, and index.py.

To make any custom implementations available inside metadata-crawler you need to create entry points in your pyproject.toml file:

pyproject.toml

# register a new index system
[project.entry-points."metadata_crawler.index_backends"]
mybackend = "my_package.my_index:MyIndexStore"

# register in your storage backend
[project.entry-points."metadata_crawler.storage_backends"]
foo = "my_package.foo_backend:FooBackend"