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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333
#![warn(missing_docs)] //#![doc(html_playground_url = "https://playground.example.com/")] // TODO add playground if possible //! This crate implements Fuzzy Searching with trigrams //! //! //! Fuzzy searching allows to compare strings by similarity rather than by equality:\ //! Similar strings will get a high score (close to `1.0f32`) while dissimilar strings will get a lower score (closer to `0.0f32`). //! //! Fuzzy searching tolerates changes in word order:\ //! ex. `"John Dep"` and `"Dep John"` will get a high score. //! //! //! The crate exposes 5 main functions: //! - [fuzzy_compare] will take 2 strings and return a score representing how similar those strings are. //! - [fuzzy_search] applies [fuzzy_compare] to a list of strings and returns a list of tuples: (word, score). //! - [fuzzy_search_sorted] is similar to [fuzzy_search] but orders the output in descending order. //! - [fuzzy_search_threshold] will take an additional `f32` as input and returns only tuples with score greater than the threshold. //! - [fuzzy_search_best_n] will take an additional `usize` arguments and returns the first `n` tuples. //! //! The Algorithm used is taken from : <https://dev.to/kaleman15/fuzzy-searching-with-postgresql-97o> //! //! Basic idea: //! //! 1. From both strings extracts all groups of 3 adjacent letters.\ //! (`"House"` becomes `[' H', ' Ho', 'Hou', 'ous', 'use', 'se ']`).\ //! Note the 2 spaces added to the head of the string and the one on the tail, used to make the algorithm work on zero length words. //! //! 1. Then counts the number of trigrams of the first words that are also present on the second word and divide by the number of trigrams of the first word.\ //! //! //! Example: Comparing 2 strings //! ```rust //! fn test () { //! use rust_fuzzy_search::fuzzy_compare; //! let score : f32 = fuzzy_compare("kolbasobulko", "kolbasobulko"); //! println!("score = {:?}", score); //! } //! ``` //! //! Example: Comparing a string with a list of strings and retrieving only the best matches //! ```rust //! fn test() { //! use rust_fuzzy_search::fuzzy_search_best_n; //! let s = "bulko"; //! let list : Vec<&str> = vec![ //! "kolbasobulko", //! "sandviĉo", //! "ŝatas", //! "domo", //! "emuo", //! "fabo", //! "fazano" //! ]; //! let n : usize = 3; //! let res : Vec<(&str, f32)> = fuzzy_search_best_n(s,&list, n); //! for (_word, score) in res { //! println!("{:?}",score) //! } //! } //! ``` //! Example: if you have a `Vec` of `String`s you need to convert it to a list of `&str` //! ```rust //! fn works_with_strings() { //! use rust_fuzzy_search::fuzzy_search; //! let s = String::from("varma"); //! let list: Vec<String> = vec![String::from("varma vetero"), String::from("varma ĉokolado")]; //! fuzzy_search(&s, &list.iter().map(String::as_ref).collect::<Vec<&str>>()); //! } //! ``` //! use std::iter; fn trigrams(s: &str) -> Vec<(char, char, char)> { let it_1 = iter::once(' ').chain(iter::once(' ')).chain(s.chars()); let it_2 = iter::once(' ').chain(s.chars()); let it_3 = s.chars().chain(iter::once(' ')); let res: Vec<(char, char, char)> = it_1 .zip(it_2) .zip(it_3) .map(|((a, b), c): ((char, char), char)| (a, b, c)) .collect(); res } /// Use this function to compare 2 strings. /// /// The output is a score (between `0.0f32 and 1.0f32`) representing how similar the 2 strings are. /// /// Arguments: /// * `a` : the first string to compare. /// * `b` : the second string to compare. /// /// /// example: /// ```rust /// fn test () { /// use rust_fuzzy_search::fuzzy_compare; /// let score : f32 = fuzzy_compare("kolbasobulko", "kolbasobulko"); /// println!("score = {:?}", score); /// } /// ``` pub fn fuzzy_compare(a: &str, b: &str) -> f32 { // gets length of first input string plus 1 (because of the 3 added spaces (' ')) let string_len = a.chars().count() + 1; // gets the trigrams for both strings let trigrams_a = trigrams(a); let trigrams_b = trigrams(b); // accumulator let mut acc: f32 = 0.0f32; // counts the number of trigrams of the first string that are also present in the second one for t_a in &trigrams_a { for t_b in &trigrams_b { if t_a == t_b { acc += 1.0f32; break; } } } let res = acc / (string_len as f32); // crops between zero and one if (0.0f32..=1.0f32).contains(&res) { res } else { 0.0f32 } } /// Use this function to compare a string (`&str`) with all elements of a list. /// /// /// The result is a list whose elements are tuples of the form `(string, score)`, the first element being the word of the list and the second element the score. /// /// Arguments: /// * `s` : the string to compare. /// * `list` : the list of strings to compare with `s`. /// /// example: /// ```rust /// fn test() { /// use rust_fuzzy_search::fuzzy_search; /// let s = "bulko"; /// let list : Vec<&str> = vec!["kolbasobulko", "sandviĉo"]; /// let res : Vec<(&str, f32)> = fuzzy_search(s,&list); /// for (_word, score) in res { /// println!("{:?}",score) /// } /// } /// ``` /// /// pub fn fuzzy_search<'a>(s: &'a str, list: &'a [&str]) -> Vec<(&'a str, f32)> { list.iter() .map(|&value| { let res = fuzzy_compare(s, value); (value, res) }) .collect() } /// This function is similar to [fuzzy_search] but sorts the result in descending order (the best matches are placed at the beginning). /// /// Arguments: /// * `s` : the string to compare. /// * `list` : the list of strings to compare with `s`. /// /// example: /// ```rust /// fn test() { /// use rust_fuzzy_search::fuzzy_search_sorted; /// let s = "bulko"; /// let list : Vec<&str> = vec!["kolbasobulko", "sandviĉo"]; /// let res : Vec<(&str, f32)> = fuzzy_search_sorted(s,&list); /// for (_word, score) in res { /// println!("{:?}",score) /// } /// } /// ``` /// pub fn fuzzy_search_sorted<'a>(s: &'a str, list: &'a [&str]) -> Vec<(&'a str, f32)> { let mut res = fuzzy_search(s, list); res.sort_by(|(_, d1), (_, d2)| d2.partial_cmp(d1).unwrap()); // TODO to fix the unwrap call res } /// This function is similar to [fuzzy_search] but filters out element with a score lower than the specified one. /// /// Arguments: /// * `s` : the string to compare. /// * `list` : the list of strings to compare with `s`. /// * `threshold` : the minimum allowed score for the elements in the result: elements with lower score will be removed. /// /// ```rust /// fn test() { /// use rust_fuzzy_search::fuzzy_search_threshold; /// let s = "bulko"; /// let list : Vec<&str> = vec!["kolbasobulko", "sandviĉo"]; /// let threshold : f32 = 0.4f32; /// let res : Vec<(&str, f32)> = fuzzy_search_threshold(s,&list, threshold); /// for (_word, score) in res { /// println!("{:?}",score) /// } /// } /// ``` pub fn fuzzy_search_threshold<'a>( s: &'a str, list: &'a [&str], threshold: f32, ) -> Vec<(&'a str, f32)> { fuzzy_search(s, list) .into_iter() .filter(|&(_, score)| score >= threshold) .collect() } /// This function is similar to [fuzzy_search_sorted] but keeps only the `n` best items, those with a better match. /// /// Arguments : /// * `s` : the string to compare. /// * `list` : the list of strings to compare with `s`. /// * `n` : the number of element to retrieve. /// /// example: /// ``` /// fn test() { /// use rust_fuzzy_search::fuzzy_search_best_n; /// let s = "bulko"; /// let list : Vec<&str> = vec!["kolbasobulko", "sandviĉo"]; /// let n : usize = 1; /// let res : Vec<(&str, f32)> = fuzzy_search_best_n(s,&list, n); /// for (_word, score) in res { /// println!("{:?}",score) /// } /// } /// ``` /// pub fn fuzzy_search_best_n<'a>(s: &'a str, list: &'a [&str], n: usize) -> Vec<(&'a str, f32)> { fuzzy_search_sorted(s, list).into_iter().take(n).collect() } #[cfg(test)] mod tests { use crate::{ fuzzy_compare, fuzzy_search, fuzzy_search_best_n, fuzzy_search_sorted, fuzzy_search_threshold, }; #[test] fn perfect_match_1() { assert_eq!(fuzzy_compare("kolbasobulko", "kolbasobulko"), 1.0f32) } #[test] fn perfect_match_2() { assert_eq!(fuzzy_compare("sandviĉo", "sandviĉo"), 1.0f32) } #[test] fn perfect_match_3() { assert_eq!(fuzzy_compare("domo", "domo"), 1.0f32) } #[test] fn perfect_match_4() { assert_eq!(fuzzy_compare("ŝatas", "ŝatas"), 1.0f32) } #[test] fn perfect_match_5() { assert_eq!(fuzzy_compare("mirinda estonto", "mirinda estonto"), 1.0f32) } #[test] fn no_match() { assert_eq!(fuzzy_compare("abc", "def"), 0.0f32) } #[test] fn empty_word() { assert_eq!(fuzzy_compare("", ""), 1.0f32) } #[test] fn one_letter() { assert_eq!(fuzzy_compare("a", "a"), 1.0f32) } #[test] fn utf8_one_letter_1() { assert_eq!(fuzzy_compare("ĉ", "ĉ"), 1.0f32) } #[test] fn utf8_one_letter_2() { assert_eq!(fuzzy_compare("ł", "ł"), 1.0f32) } #[test] fn utf8_no_match() { assert_eq!(fuzzy_compare("cgs", "ĉĝŝ"), 0.0f32) } #[test] fn test_fuzzy_search_1() { let s: &str = "bulko"; let list: Vec<&str> = vec!["kolbasobulko", "sandviĉo", "kolbasobulkejo"]; let res: Vec<(&str, f32)> = fuzzy_search(s, &list); assert_eq!(res.into_iter().count(), 3); } #[test] fn test_fuzzy_search_sorted() { let s: &str = "bulko"; let list: Vec<&str> = vec!["kolbasobulko", "sandviĉo", "kolbasobulkejo"]; let res: Vec<(&str, f32)> = fuzzy_search_sorted(s, &list); assert_eq!(res.into_iter().count(), 3); } #[test] fn no_lowers() { let threshold = 0.5f32; let s: &str = "bulko"; let list: Vec<&str> = vec!["kolbasobulko", "sandviĉo", "kolbasobulkejo"]; for (_word, score) in fuzzy_search_threshold(s, &list, threshold) { assert!(score > threshold) } } #[test] fn test_fuzzy_search_best_n() { let s: &str = "bulko"; let list: Vec<&str> = vec!["kolbasobulko", "sandviĉo", "kolbasobulkejo"]; let res: Vec<(&str, f32)> = fuzzy_search_best_n(s, &list, 2); assert_eq!(res.into_iter().count(), 2); } #[test] fn works_with_strings() { let s = String::from("varma"); let list: Vec<String> = vec![String::from("varma vetero"), String::from("varma ĉokolado")]; fuzzy_search(&s, &list.iter().map(String::as_ref).collect::<Vec<&str>>()); } }