Struct TranslatorConfig 

pub struct TranslatorConfig {
    pub model_dir: Option<String>,
    pub working_directory: Option<String>,
    pub execute_threshold: f32,
    pub confirm_threshold: f32,
    pub cache_config: Option<CacheConfig>,
    pub default_limit: u32,
    pub languages: Vec<String>,
    pub model_dir_override: Option<PathBuf>,
    pub allow_unverified_model: bool,
    pub allow_model_download: bool,
    pub model_cache_dir: Option<PathBuf>,
    pub classifier_pool_size: Option<usize>,
}

Configuration for the Translator.

Fields

model_dir: Option<String>

Path to model directory (for classifier feature). Directory should contain: intent_classifier.onnx, tokenizer.json, and optionally calibration.json or temperature.json for confidence calibration

working_directory: Option<String>

Context: current working directory for relative paths.

execute_threshold: f32

Custom confidence thresholds.

confirm_threshold: f32

cache_config: Option<CacheConfig>

Cache configuration. Set to None to disable caching.

default_limit: u32

Default result limit (affects cache key generation).

languages: Vec<String>

Languages to restrict searches to (affects cache key generation).

model_dir_override: Option<PathBuf>

Optional override for the resolved model directory (NL02 resolver chain entry).

When set, the resolver short-circuits the rest of the lookup chain and uses this path directly. Distinct from the legacy Self::model_dir field, which remains in place for backward compatibility.

allow_unverified_model: bool

Permit loading a model whose checksums cannot be verified.

Defaults to false. NL04 enforces strict integrity by default; this flag is the operator escape hatch for development workflows.

allow_model_download: bool

Permit fetching the model from the network when not present locally.

Defaults to false. NL03 wires the gated downloader behind this flag.

model_cache_dir: Option<PathBuf>

Optional cache directory for downloaded model artifacts.

Defaults to None, in which case NL03 selects a platform-appropriate cache root.

classifier_pool_size: Option<usize>

Optional override for the classifier worker pool size.

Defaults to None. NL07 resolves this to the SQRY_NL_CLASSIFIER_POOL_SIZE environment variable, falling back to 2 workers.

Trait Implementations

impl Clone for TranslatorConfig

fn clone(&self) -> TranslatorConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for TranslatorConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for TranslatorConfig

fn default() -> Self

Returns the “default value” for a type.