Skip to content

oa-configurator

A shared configuration layer for the OMOP-oriented Python stack.

oa-configurator gives OMOP tools a single typed configuration file instead of a tangle of environment variables and package-local .env files.

Key Concepts

  • Connection: A concrete database endpoint (host, dialect, credentials)
  • Resource: A logical role bundle, e.g. primary OMOP CDM DB, embedding DB, artficat paths
  • Profile: A named environment (e.g. local, prod) that patches resources and tools
  • Tool: Per-tool defaults, e.g. backend, default resource, storage roots
  • Logging: One call configures consistent log output for the entire OMOP Python stack

Info

Configuration lives in one TOML file (default ~/.config/omop/config.toml, overridable via OA_CONFIG_PATH) and is loaded once. The Resolver turns logical names into typed, credential-resolved handles ready for use.

Quick Example

from oa_configurator import load_stack_config, Resolver

config = load_stack_config()                        # reads CONFIG_PATH (default ~/.config/omop/config.toml)
resolver = Resolver(config)

resource = resolver.resolve_resource("default")
engine   = resource.create_engine()                # SQLAlchemy Engine, schema_translate_map applied

from oa_configurator import StackConfig, DatabaseConfig, ResourceConfig, Resolver

config = StackConfig.for_session(
    databases={"local": DatabaseConfig(dialect="postgresql", host="localhost",
                                       database_name="omop", password="omop")},
    resources={"default": ResourceConfig(database="local", cdm_schema="cdm")},
)
engine = Resolver(config).resolve_resource("default").create_engine()

from oa_configurator import load_stack_config, DatabaseConfig, ResourceConfig, Resolver

# Load shared team config, redirect one resource to a local SQLite database
engine = (
    Resolver(load_stack_config())
    .with_overrides(
        databases={"local": DatabaseConfig(dialect="sqlite", database_name="/data/local.db")},
        resources={"default": ResourceConfig(database="local", cdm_schema="omop")},
    )
    .resolve_resource("default")
    .create_engine()
)

Next Steps