Struct EvaluatorRegistry 

pub struct EvaluatorRegistry { /* private fields */ }

Global registry of condition evaluators.

Evaluators are registered at startup (typically from generated code) and looked up at runtime based on the detected message type and format version.

Thread-safe: uses RwLock for concurrent read access with exclusive write access during registration.

Implementations

impl EvaluatorRegistry

pub fn new() -> Self

Create a new empty registry.

pub fn register<E: ConditionEvaluator + 'static>(&self, evaluator: E)

Register a condition evaluator for a message type and format version.

Overwrites any previously registered evaluator for the same key.

pub fn register_as<E: ConditionEvaluator + 'static>( &self, evaluator: E, message_type: &str, format_version: &str, )

Register a condition evaluator under a custom key, overriding the evaluator’s own message_type() and format_version().

Useful for type aliases where the underlying evaluator reports a different format version than the one we want to register under.

pub fn register_arc(&self, evaluator: Arc<dyn ConditionEvaluator>)

Register an already-Arc’d evaluator.

pub fn get( &self, message_type: &str, format_version: &str, ) -> Option<Arc<dyn ConditionEvaluator>>

Look up an evaluator by message type and format version.

Returns None if no evaluator is registered for the given key.

pub fn get_or_err( &self, message_type: &str, format_version: &str, ) -> Result<Arc<dyn ConditionEvaluator>, ValidationError>

Look up an evaluator, returning an error if not found.

pub fn registered_keys(&self) -> Vec<(String, String)>

List all registered (message_type, format_version) keys.

pub fn clear(&self)

Clear all registered evaluators. Primarily for testing.

Trait Implementations

impl Default for EvaluatorRegistry

fn default() -> Self

Returns the “default value” for a type.