Struct ngrammatic::Corpus

source · [−]

pub struct Corpus { /* private fields */ }

Expand description

Holds a corpus of words and their ngrams, allowing fuzzy matches of candidate strings against known strings in the corpus.

Implementations

source

impl Corpus

source

pub fn add_ngram(&mut self, ngram: Ngram)

Add the supplied ngram to the Corpus.

let mut corpus = CorpusBuilder::new().finish();
corpus.add_ngram(NgramBuilder::new("tomato").finish());
let results = corpus.search("tomacco", 0.40);
if let Some(result) = results.first() {
    println!("Closest match to 'tomacco' in the corpus was {}", result.text);
} else {
    println!("The corpus contained no words similar to 'tomacco'.");
}

source

pub fn add_text(&mut self, text: &str)

Generate an Ngram for the supplied text, and add it to the Corpus.

let mut corpus = CorpusBuilder::new().finish();
corpus.add_text("tomato");
let results = corpus.search("tomacco", 0.40);
if let Some(result) = results.first() {
    println!("Closest match to 'tomacco' in the corpus was {}", result.text);
} else {
    println!("The corpus contained no words similar to 'tomacco'.");
}

source

pub fn is_empty(&self) -> bool

If the corpus is empty.

source

pub fn key(&self, text: &str) -> Option<String>

Determines whether an exact match exists for the supplied text in the Corpus index, after processing it with the Corpus’s key_trans function.

source

pub fn search(&self, text: &str, threshold: f32) -> Vec<SearchResult>

Perform a fuzzy search of the Corpus for Ngrams above some threshold of similarity to the supplied text. Returns up to 10 results, sorted by highest similarity to lowest.

let mut corpus = CorpusBuilder::new().finish();
corpus.add_text("tomato");
let results = corpus.search("tomacco", 0.40);
if let Some(result) = results.first() {
    println!("Closest match to 'tomacco' in the corpus was {}", result.text);
} else {
    println!("The corpus contained no words similar to 'tomacco'.");
}

source

pub fn search_with_warp(
 &self,
 text: &str,
 warp: f32,
 threshold: f32
) -> Vec<SearchResult>

Perform a fuzzy search of the Corpus for Ngrams with a custom warp for results above some threshold of similarity to the supplied text. Returns up to 10 results, sorted by highest similarity to lowest.

let mut corpus = CorpusBuilder::new().finish();
corpus.add_text("tomato");
let results = corpus.search_with_warp("tomacco", 2.0, 0.40);
if let Some(result) = results.first() {
    println!("Closest match to 'tomacco' in the corpus was {}", result.text);
} else {
    println!("The corpus contained no words similar to 'tomacco'.");
}