index.js

/**
 * modify from npm words-count
 */

//Match latin, cyrillic, Malayalam letters and numbers
const common = "(\\d+)|[a-zA-Z\u00C0-\u00FF\u0100-\u017F\u0180-\u024F\u0250-\u02AF\u1E00-\u1EFF\u0400-\u04FF\u0500-\u052F\u0D00-\u0D7F]+|";

//Match Chinese Hànzì, the Japanese Kanji and the Korean Hanja
const cjk = "\u2E80-\u2EFF\u2F00-\u2FDF\u3000-\u303F\u31C0-\u31EF\u3200-\u32FF\u3300-\u33FF\u3400-\u3FFF\u4000-\u4DBF\u4E00-\u4FFF\u5000-\u5FFF\u6000-\u6FFF\u7000-\u7FFF\u8000-\u8FFF\u9000-\u9FFF\uF900-\uFAFF";

//Match Japanese Hiragana, Katakana, Rōmaji
const jp = "\u3040-\u309F\u30A0-\u30FF\u31F0-\u31FF\u3190-\u319F";

//Match Korean Hangul
const kr = "\u1100-\u11FF\u3130-\u318F\uA960-\uA97F\uAC00-\uAFFF\uB000-\uBFFF\uC000-\uCFFF\uD000-\uD7AF\uD7B0-\uD7FF";

//words, no symbols
const reg = new RegExp(common + "[" + cjk + jp + kr + "]", "g");

/**
 * Split the text by words.
 * @param {*} text 
 * @returns {number[]} position indexs of each word
 */
function splitWords(text) {
    if (!text) return [];

    //Split words by white space (For European languages)
    var words = text.split(' ');

    var indexs = [];
    var lastPos = 0;

    for (var i = 0; i < words.length; i++) {
        var word = removeSymbols(words[i]);

        var pos = i === 0 ? 0 : lastPos + words[i - 1].length + 1;

        if (word.trim()) {
            var carry = 0, m;

            while (m = reg.exec(word)) {
                indexs.push(carry === 0 ? pos : pos + m.index);
                carry++;
            }

            if (carry === 0) {//not supported language
                indexs.push(pos);
            }
        }

        lastPos = pos;
    }
    return indexs;
}

function removeSymbols(word) {
    return word.replace(/[,，.。:：;；\[【\]】{｛}｝(（)）<《>》$￥!！?？~～'’"“”*/\\&%@#^、\uFF00-\uFFEF\u2000-\u206F]/g, ' ');
}

function cropText(text, maxCount, splittedWords = splitWords(text)) {
    var v = splittedWords[maxCount];
    if (!v) return text;
    return text.substr(0, v);
}

module.exports = { splitWords, cropText };