Struct NerExtractor 

pub struct NerExtractor { /* private fields */ }

Rule-based and pattern-matching NER extractor.

Example

use scirs2_text::ner::{NerExtractor, NerConfig, EntityType};

let config = NerConfig::default();
let extractor = NerExtractor::new(config);
let entities = extractor.extract("Contact us at support@example.com").unwrap();
assert!(entities.iter().any(|e| e.entity_type == EntityType::Email));

Implementations

impl NerExtractor

pub fn new(config: NerConfig) -> Self

Create a new extractor from the given configuration.

pub fn try_new(config: NerConfig) -> Result<Self>

Attempt to create a new extractor, returning an error if any custom pattern fails to compile.

pub fn add_gazetteer_entry(&mut self, word: &str, entity_type: EntityType)

Add a single entry to the gazetteer.

pub fn extract(&self, text: &str) -> Result<Vec<Entity>>

Extract all entities from text.

Entities are returned sorted by byte start offset. Overlapping entities are resolved by keeping the longer match.