Struct LabelNormalizer 

pub struct LabelNormalizer { /* private fields */ }

Normalizes entity type labels across schemas and languages.

This is a simple alias table, not a complex ontology graph. Add aliases as needed; the lookup is O(1) via HashMap.

Example

use anno_core::core::ontology::{LabelNormalizer, CoreType};

let norm = LabelNormalizer::default();

// Different schemas resolve to same type
assert_eq!(norm.normalize("PER"), Some(CoreType::Person));
assert_eq!(norm.normalize("PERSON"), Some(CoreType::Person));
assert_eq!(norm.normalize("B-PER"), Some(CoreType::Person)); // BIO prefix stripped

// Cross-lingual
assert_eq!(norm.normalize("personne"), Some(CoreType::Person)); // French

Implementations

impl LabelNormalizer

pub fn new() -> LabelNormalizer

Create empty normalizer (no aliases registered).

pub fn register(&self, alias: &str, core_type: CoreType)

Register an alias for a core type.

pub fn register_many(&self, aliases: &[&str], core_type: CoreType)

Register multiple aliases for a core type.

pub fn normalize(&self, label: &str) -> Option<CoreType>

Normalize a label to its core type.

Handles BIO/BIOES prefixes automatically.

pub fn is_known(&self, label: &str) -> bool

Check if a label is known.

pub fn register_biomedical(&self)

Register biomedical domain types.

These are kept separate because they’re domain-specific and shouldn’t pollute the core type namespace.

pub fn register_legal(&self)

Register legal domain types.

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

Get all known aliases (for debugging/documentation).

Trait Implementations

impl Default for LabelNormalizer

fn default() -> LabelNormalizer

Returns the “default value” for a type.