10 lines
407 B
JavaScript
10 lines
407 B
JavaScript
// Character bigrams of an already normalized text, used by the Dice similarity.
|
|
// "abc" -> ["ab", "bc"]. A text shorter than two characters has no bigram.
|
|
export function buildBigrams(text) {
|
|
if (typeof text !== 'string' || text.length < 2) return []
|
|
const bigrams = []
|
|
for (let index = 0; index < text.length - 1; index += 1) {
|
|
bigrams.push(text.slice(index, index + 2))
|
|
}
|
|
return bigrams
|
|
}
|