Presence and counting

Choose a boolean, a total, or a count for each dictionary entry without retaining a result array.

match(text, options?)

Returns a boolean and stops after the first hit without creating a match object. Use it when you only need presence, not every result.

count(text, options?)

Returns the total number of occurrences, including overlaps and duplicate entries, equal to search(text).length. It uses aggregate state counts without creating match objects or walking output links. Empty text or an empty dictionary returns 0. Invalid text throws TypeError; a count exceeding Number.MAX_SAFE_INTEGER throws RangeError. No strategy option is accepted.

import AhoCorasick from 'modern-ahocorasick'

new AhoCorasick(['a', 'aa', 'a']).count('aaa') // 8

countByPattern(text, options?)

Returns a fresh number[] in original input order. Counts include overlaps; missing patterns have zero counts, and duplicate entries retain separate slots. For example, new AhoCorasick(['a', 'aa', 'a', 'b']).countByPattern('aaa') returns [3, 2, 3, 0]. Empty dictionaries return []; invalid text throws TypeError.

The scanner aggregates state visits through failure links without enumerating occurrences. Excluding segmentation, time is O(g log(d + 1) + s + p) and temporary memory is O(s + p), for g text graphemes, s states, p input patterns and maximum transition degree d. Use count() when only a total is needed: it avoids the state-sized temporary array.

Filtered queries

All three methods accept QueryOptions: wholeWord, locale, start, end, and anchored; none accepts a selection strategy. Counts include overlaps and duplicate entries among the accepted matches. Aggregate scan costs above describe default exact matching. Word or constructor boundary filters and transformed matchers may enumerate occurrences and allocate additional state.

See Query options.