Ultra-fast string similarity and scoring metrics. This repo intends to provide the fastest implementations possible for a set of common string metrics. Speed scores are based on scale of 0 to 100, relative to other algorithms

11KB development, < 2KB minified and gzipped

Speed score: 67. Jaro-Winkler similarity (based on Lars Marius Garshol Java implementation).

Speed score: 35. Levenshtein distance between two strings.

Speed score: 20. Length of longest common substring.

Speed score: 100. Optimized version of Joshaven Potter's string score. Fuzziness between 0 (exact) and 1.

Variation of Phonex created by the author. As opposed to Phonex, first letters are encoded, duplicates are preserved, results are not limited by length.

Returns the Phonex encoding of a string.