matchr-rs 0.1.2

Fast fuzzy string matching — Levenshtein, Jaro-Winkler, and trigram similarity. Usable from Rust and Python.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100

use crate::{
    batch_best_match as bbm_fn, best_match as bm_fn, combined_score as cs_fn,
    jaro_winkler as jw_fn, levenshtein as lev_fn, rank_matches as rm_fn,
    token_set_ratio as tset_fn, token_sort_ratio as tsort_fn, trigram_similarity as tri_fn,
};
use pyo3::prelude::*;

#[pyfunction]
fn levenshtein(a: &str, b: &str) -> usize {
    lev_fn(a, b)
}

#[pyfunction]
fn jaro_winkler(a: &str, b: &str) -> f64 {
    jw_fn(a, b)
}

#[pyfunction]
fn trigram_similarity(a: &str, b: &str) -> f64 {
    tri_fn(a, b)
}

#[pyfunction]
fn combined_score(a: &str, b: &str) -> f64 {
    cs_fn(a, b)
}

#[pyfunction]
fn token_sort_ratio(a: &str, b: &str) -> f64 {
    tsort_fn(a, b)
}

#[pyfunction]
fn token_set_ratio(a: &str, b: &str) -> f64 {
    tset_fn(a, b)
}

#[pyfunction]
#[pyo3(signature = (query, candidates, threshold=None))]
fn best_match(
    query: &str,
    candidates: Vec<String>,
    threshold: Option<f64>,
) -> Option<(String, f64)> {
    let refs: Vec<&str> = candidates.iter().map(|s| s.as_str()).collect();
    let min = threshold.unwrap_or(0.0);
    bm_fn(query, &refs)
        .filter(|(_, score)| *score >= min)
        .map(|(s, score)| (s.to_string(), score))
}

#[pyfunction]
#[pyo3(signature = (query, candidates, threshold=None))]
fn rank_matches(
    query: &str,
    candidates: Vec<String>,
    threshold: Option<f64>,
) -> Vec<(String, f64)> {
    let refs: Vec<&str> = candidates.iter().map(|s| s.as_str()).collect();
    let min = threshold.unwrap_or(0.0);
    rm_fn(query, &refs)
        .into_iter()
        .filter(|(_, score)| *score >= min)
        .map(|(s, score)| (s.to_string(), score))
        .collect()
}

#[pyfunction]
#[pyo3(signature = (queries, candidates, threshold=None))]
fn batch_best_match(
    queries: Vec<String>,
    candidates: Vec<String>,
    threshold: Option<f64>,
) -> Vec<Option<(String, f64)>> {
    let q_refs: Vec<&str> = queries.iter().map(|s| s.as_str()).collect();
    let c_refs: Vec<&str> = candidates.iter().map(|s| s.as_str()).collect();
    let min = threshold.unwrap_or(0.0);

    bbm_fn(&q_refs, &c_refs)
        .into_iter()
        .map(|opt| {
            opt.filter(|(_, score)| *score >= min)
                .map(|(s, score)| (s.to_string(), score))
        })
        .collect()
}

#[pymodule]
fn matchr(m: &Bound<'_, PyModule>) -> PyResult<()> {
    m.add_function(wrap_pyfunction!(levenshtein, m)?)?;
    m.add_function(wrap_pyfunction!(jaro_winkler, m)?)?;
    m.add_function(wrap_pyfunction!(trigram_similarity, m)?)?;
    m.add_function(wrap_pyfunction!(combined_score, m)?)?;
    m.add_function(wrap_pyfunction!(token_sort_ratio, m)?)?;
    m.add_function(wrap_pyfunction!(token_set_ratio, m)?)?;
    m.add_function(wrap_pyfunction!(best_match, m)?)?;
    m.add_function(wrap_pyfunction!(rank_matches, m)?)?;
    m.add_function(wrap_pyfunction!(batch_best_match, m)?)?;
    Ok(())
}