Sometimes fuzzy is too loose and you want a literal substring. fzf spells that with a leading quote. Today you add exact terms - a term prefixed with a quote matches only a contiguous substring.
Treat a term beginning with a single quote as an exact-substring match rather than a fuzzy one.
Fuzzy matching is forgiving, but sometimes you know exactly what you want and the forgiveness gets in the way - mn fuzzily matches main.go, but if you meant the literal letters mn you would rather it did not. fzf’s answer is the exact term: prefix a term with a single quote and it matches only where its characters appear contiguously, as a real substring.
The elegant part is scoring. An exact match is just a run of adjacent positions, so you can feed that run straight into the same score function - it earns full consecutive bonuses and whatever boundary bonus its start qualifies for, with no gaps. If the substring appears more than once, score each occurrence and keep the best. Parsing is a one-character change: a leading ' sets the Exact flag on the Term you built last lesson and strips the quote. The subsequence gate and the exact gate now coexist, chosen per term.
// parseQuery: a leading ' sets Term.Exact and strips the quote.// An exact term matches only where its text appears as a CONTIGUOUS// substring. Score that occurrence's run of positions with the same// score() you already have; pick the best-scoring occurrence.if strings.HasPrefix(word, "'") {t := Term{Text: word[1:], Exact: true}// ...}