fuzzy

func AnyTypo ¶

func AnyTypo(term string, candidates []string) bool

AnyTypo reports whether term is a typo of any of the candidates (see IsTypo).

func FindClosestByLevenshtein ¶

func FindClosestByLevenshtein(target string, candidates []string) (string, int)

FindClosestByLevenshtein returns the closest match to target from candidates and its distance. If candidates is empty, it returns ("", 0).

func IsTypo ¶

func IsTypo(term, candidate string) bool

IsTypo reports whether candidate is a plausible typo of term — within a small, length-scaled Levenshtein distance. Terms shorter than 4 characters are never typo-matched: at that length too many unrelated words are a single edit apart (e.g. "car"/"cart"/"care") to match safely. Both inputs should already be normalized (e.g. lowercased) by the caller.

Tolerance is 1 edit for 4–5 character terms and 2 for longer ones. The looser bound on longer terms also admits a single adjacent transposition — which Levenshtein scores as 2 edits — so common typos like "updaet" → "update" still match.

func LevenshteinDistance ¶

func LevenshteinDistance(a, b string) int

LevenshteinDistance computes the Levenshtein edit distance between two strings. The implementation is iterative and uses O(min(m,n)) additional memory.