ml_switcheroo.config

Runtime Configuration Store. Updated to support Framework Flavours (Hierarchical Selection).

Attributes

tomllib
T

Classes

RuntimeConfig

Global configuration container for the translation engine.

Functions

get_framework_priority_order(→ List[str])	Returns a list of framework keys sorted by UI Priority.
parse_cli_key_values(→ Dict[str, Any])	Parses a list of 'key=value' strings into a dictionary.

Module Contents

ml_switcheroo.config.tomllib = None

ml_switcheroo.config.T

ml_switcheroo.config.get_framework_priority_order() → List[str]

Returns a list of framework keys sorted by UI Priority.

The sort order is determined by the ui_priority attribute of the registered FrameworkAdapter. Lower numbers appear first.

Returns:

Sorted list of framework identifiers (e.g. [‘torch’, ‘jax’]).

Return type:

List[str]

class ml_switcheroo.config.RuntimeConfig(/, **data: Any)

Bases: pydantic.BaseModel

Global configuration container for the translation engine.

source_framework: str = None

target_framework: str = None

source_flavour: str | None = None

target_flavour: str | None = None

strict_mode: bool = None

plugin_settings: Dict[str, Any] = None

plugin_paths: List[pathlib.Path] = None

validation_report: pathlib.Path | None = None

classmethod validate_framework(v: str) → str

Ensures the framework is registered in the system.

Parameters:

v (str) – The framework key to validate.

Returns:

The normalized (lowercase) framework key.

Return type:

str

Raises:

ValueError – If the framework is not found in the registry.

property effective_source: str

Resolves the specific framework key to use for source logic.

If a flavour (e.g. ‘flax_nnx’) is provided, it overrides the general framework (‘jax’).

Returns:

The active source framework key.

Return type:

str

property effective_target: str

Resolves the specific framework key to use for target logic.

Returns:

The active target framework key.

Return type:

str

parse_plugin_settings(schema: Type[T]) → T

Validates the raw plugin settings dictionary against a specific Pydantic model.

Parameters:

schema (Type[T]) – The Pydantic model class defining expected settings.

Returns:

An instance of the schema model populated with runtime values.

Return type:

T

classmethod load(source: str | None = None, target: str | None = None, source_flavour: str | None = None, target_flavour: str | None = None, strict_mode: bool | None = None, plugin_settings: Dict[str, Any] | None = None, validation_report: pathlib.Path | None = None, search_path: pathlib.Path | None = None) → RuntimeConfig

Loads configuration from pyproject.toml and overrides with CLI arguments.

Parameters:

• source (Optional[str]) – Override for source framework.

• target (Optional[str]) – Override for target framework.

• source_flavour (Optional[str]) – Override for source flavour.

• target_flavour (Optional[str]) – Override for target flavour.

• strict_mode (Optional[bool]) – Override for strict mode setting.

• plugin_settings (Optional[Dict]) – Additional CLI plugin settings.

• validation_report (Optional[Path]) – Override for validation report path.

• search_path (Optional[Path]) – Directory to start searching for TOML config.

Returns:

The fully resolved configuration object.

Return type:

RuntimeConfig

ml_switcheroo.config.parse_cli_key_values(items: List[str] | None) → Dict[str, Any]

Parses a list of ‘key=value’ strings into a dictionary.

Types are inferred (int, float, bool, or string).

Parameters:

items (Optional[List[str]]) – List of raw CLI strings directly from argparse.

Returns:

Parsed dictionary.

Return type:

Dict[str, Any]