JavaScript实现Unicode字符串与UTF-8字节数据互转详解

JavaScript字符串编码基础与转换需求

JavaScript中的String类型变量在内部使用Unicode编码来存储字符串。当需要将字符串持久化存储或进行网络传输时，通常需要先将其转换为特定的编码格式，例如UTF-8或UTF-32。常见的场景是将数据以UTF-8编码形式存入数据库，而在读取后如何将其还原为正确的字符串对象便成为了关键问题。本文提供的解决方案能够正确处理包含中文、Emoji表情以及英文字符的混合字符串，实现Unicode码与UTF-8字节数据之间的双向转换。

/**
 * Created by hdwang on 2019/1/28.
 */
var convertUtf8 = (function() {

  /**
   * unicode string to utf-8
   * @param text 字符串
   * @returns {*} utf-8编码
   */
  function toBytes(text) {
    var result = [], i = 0;
    text = encodeURI(text);
    while (i < text.length) {
      var c = text.charCodeAt(i++);

      // if it is a % sign, encode the following 2 bytes as a hex value
      if (c === 37) {
        result.push(parseInt(text.substr(i, 2), 16))
        i += 2;

        // otherwise, just the actual byte
      } else {
        result.push(c)
      }
    }

    return coerceArray(result);
  }


  /**
   * utf8 byte to unicode string
   * @param utf8Bytes
   * @returns {string}
   */
  function utf8ByteToUnicodeStr(utf8Bytes){
    var unicodeStr ="";
    for (var pos = 0; pos < utf8Bytes.length;){
      var flag= utf8Bytes[pos];
      var unicode = 0 ;
      if ((flag >>>7) === 0 ) {
        unicodeStr+= String.fromCharCode(utf8Bytes[pos]);
        pos += 1;

      } else if ((flag &0xFC) === 0xFC ){
        unicode = (utf8Bytes[pos] & 0x3) << 30;
        unicode |= (utf8Bytes[pos+1] & 0x3F) << 24;
        unicode |= (utf8Bytes[pos+2] & 0x3F) << 18;
        unicode |= (utf8Bytes[pos+3] & 0x3F) << 12;
        unicode |= (utf8Bytes[pos+4] & 0x3F) << 6;
        unicode |= (utf8Bytes[pos+5] & 0x3F);
        unicodeStr+= String.fromCodePoint(unicode) ;
        pos += 6;

      }else if ((flag &0xF8) === 0xF8 ){
        unicode = (utf8Bytes[pos] & 0x7) << 24;
        unicode |= (utf8Bytes[pos+1] & 0x3F) << 18;
        unicode |= (utf8Bytes[pos+2] & 0x3F) << 12;
        unicode |= (utf8Bytes[pos+3] & 0x3F) << 6;
        unicode |= (utf8Bytes[pos+4] & 0x3F);
        unicodeStr+= String.fromCodePoint(unicode) ;
        pos += 5;

      } else if ((flag &0xF0) === 0xF0 ){
        unicode = (utf8Bytes[pos] & 0xF) << 18;
        unicode |= (utf8Bytes[pos+1] & 0x3F) << 12;
        unicode |= (utf8Bytes[pos+2] & 0x3F) << 6;
        unicode |= (utf8Bytes[pos+3] & 0x3F);
        unicodeStr+= String.fromCodePoint(unicode) ;
        pos += 4;

      } else if ((flag &0xE0) === 0xE0 ){
        unicode = (utf8Bytes[pos] & 0x1F) << 12;;
        unicode |= (utf8Bytes[pos+1] & 0x3F) << 6;
        unicode |= (utf8Bytes[pos+2] & 0x3F);
        unicodeStr+= String.fromCharCode(unicode) ;
        pos += 3;

      } else if ((flag &0xC0) === 0xC0 ){ //110
        unicode = (utf8Bytes[pos] & 0x3F) << 6;
        unicode |= (utf8Bytes[pos+1] & 0x3F);
        unicodeStr+= String.fromCharCode(unicode) ;
        pos += 2;

      } else{
        unicodeStr+= String.fromCharCode(utf8Bytes[pos]);
        pos += 1;
      }
    }
    return unicodeStr;
  }



  function checkInt(value) {
    return (parseInt(value) === value);
  }

  function checkInts(arrayish) {
    if (!checkInt(arrayish.length)) { return false; }

    for (var i = 0; i < arrayish.length; i++) {
      if (!checkInt(arrayish[i]) || arrayish[i] < 0 || arrayish[i] > 255) {
        return false;
      }
    }

    return true;
  }

  function coerceArray(arg, copy) {

    // ArrayBuffer view
    if (arg.buffer && arg.name === 'Uint8Array') {

      if (copy) {
        if (arg.slice) {
          arg = arg.slice();
        } else {
          arg = Array.prototype.slice.call(arg);
        }
      }

      return arg;
    }

    // It's an array; check it is a valid representation of a byte
    if (Array.isArray(arg)) {
      if (!checkInts(arg)) {
        throw new Error('Array contains invalid value: ' + arg);
      }

      return new Uint8Array(arg);
    }

    // Something else, but behaves like an array (maybe a Buffer? Arguments?)
    if (checkInt(arg.length) && checkInts(arg)) {
      return new Uint8Array(arg);
    }

    throw new Error('unsupported array-like object');
  }

  return {
    toBytes: toBytes,
    fromBytes: utf8ByteToUnicodeStr
  }
})()

Emoji字符的UTF-8处理机制

针对Emoji字符，其在UTF-8编码中通常占用4个字节，对应两个Unicode码位（即两个UTF-16字符）。虽然可以使用String.fromCharCode方法通过两次调用实现转换，但这种方式不如直接使用fromPointCode直观。下文展示了将UTF-8编码的4字节数据还原为Unicode（UTF-16）字符串的具体处理过程。

//高char10位[一个unicode字符] (2+6+2=10)
unicode =  ((utf8Bytes[pos] & 0x3)) << 8 |((utf8Bytes[pos+1] & 0x3f) << 2) |((utf8Bytes[pos+2] >> 4) & 0x03);

//减去‭1F600‬中的1，这里减去6个0即可，低位char已经占据10位
unicode = unicode - parseInt('1000000',2)

//加上utf-16高char的标识符
unicode = 0xD800 + unicode;
console.log(unicode);
unicodeStr += String.fromCharCode(unicode);

//低char10位[一个unicode字符](4+6)
unicode = ((utf8Bytes[pos+2] & 0x0F) << 6) | (utf8Bytes[pos+3] & 0x3F);
//加上utf-16低char的标识符
unicode = 0xDC00 + unicode;
console.log(unicode);
unicodeStr+= String.fromCharCode(unicode);
pos += 4;

上述内容介绍了在JavaScript中实现Unicode字符串与UTF-8字节数据相互转换的完整方案，旨在帮助开发者解决多语言及特殊符号的编码难题。如果读者在实际应用中遇到疑问，欢迎留言交流，小编会及时回复。同时感谢各位对本站的支持。