过滤非法字符

const replacement = '\ufffd';
// https://gist.github.com/mathiasbynens/bbe7f870208abcfec860
const loneSurrogates = /[\uD800-\uDBFF](?![\uDC00-\uDFFF])|([^\uD800-\uDBFF]|^)[\uDC00-\uDFFF]/g;
 
// string sanitizes the given str by replacing invalid UTF-16 code unit
// sequences with the unicode replacement character. Returns a new string.
//用 \ufffd 取代无效的增补字节的，无效的高位或者低位
exports.string = function(str) {
    return str.replace(loneSurrogates, '$1' + replacement);
}

字符串中包含了一个非标准的 UTF-16 字符