我有一些UTF-8编码的数据生活在Javascript Uint8Array元素的范围内。是否有一种有效的方法来解码这些到一个常规的javascript字符串(我相信javascript使用16位Unicode)?我不想一次添加一个字符,因为字符串连接会变得CPU密集。
当前回答
class UTF8{
static encode(str:string){return new UTF8().encode(str)}
static decode(data:Uint8Array){return new UTF8().decode(data)}
private EOF_byte:number = -1;
private EOF_code_point:number = -1;
private encoderError(code_point) {
console.error("UTF8 encoderError",code_point)
}
private decoderError(fatal, opt_code_point?):number {
if (fatal) console.error("UTF8 decoderError",opt_code_point)
return opt_code_point || 0xFFFD;
}
private inRange(a:number, min:number, max:number) {
return min <= a && a <= max;
}
private div(n:number, d:number) {
return Math.floor(n / d);
}
private stringToCodePoints(string:string) {
/** @type {Array.<number>} */
let cps = [];
// Based on http://www.w3.org/TR/WebIDL/#idl-DOMString
let i = 0, n = string.length;
while (i < string.length) {
let c = string.charCodeAt(i);
if (!this.inRange(c, 0xD800, 0xDFFF)) {
cps.push(c);
} else if (this.inRange(c, 0xDC00, 0xDFFF)) {
cps.push(0xFFFD);
} else { // (inRange(c, 0xD800, 0xDBFF))
if (i == n - 1) {
cps.push(0xFFFD);
} else {
let d = string.charCodeAt(i + 1);
if (this.inRange(d, 0xDC00, 0xDFFF)) {
let a = c & 0x3FF;
let b = d & 0x3FF;
i += 1;
cps.push(0x10000 + (a << 10) + b);
} else {
cps.push(0xFFFD);
}
}
}
i += 1;
}
return cps;
}
private encode(str:string):Uint8Array {
let pos:number = 0;
let codePoints = this.stringToCodePoints(str);
let outputBytes = [];
while (codePoints.length > pos) {
let code_point:number = codePoints[pos++];
if (this.inRange(code_point, 0xD800, 0xDFFF)) {
this.encoderError(code_point);
}
else if (this.inRange(code_point, 0x0000, 0x007f)) {
outputBytes.push(code_point);
} else {
let count = 0, offset = 0;
if (this.inRange(code_point, 0x0080, 0x07FF)) {
count = 1;
offset = 0xC0;
} else if (this.inRange(code_point, 0x0800, 0xFFFF)) {
count = 2;
offset = 0xE0;
} else if (this.inRange(code_point, 0x10000, 0x10FFFF)) {
count = 3;
offset = 0xF0;
}
outputBytes.push(this.div(code_point, Math.pow(64, count)) + offset);
while (count > 0) {
let temp = this.div(code_point, Math.pow(64, count - 1));
outputBytes.push(0x80 + (temp % 64));
count -= 1;
}
}
}
return new Uint8Array(outputBytes);
}
private decode(data:Uint8Array):string {
let fatal:boolean = false;
let pos:number = 0;
let result:string = "";
let code_point:number;
let utf8_code_point = 0;
let utf8_bytes_needed = 0;
let utf8_bytes_seen = 0;
let utf8_lower_boundary = 0;
while (data.length > pos) {
let _byte = data[pos++];
if (_byte == this.EOF_byte) {
if (utf8_bytes_needed != 0) {
code_point = this.decoderError(fatal);
} else {
code_point = this.EOF_code_point;
}
} else {
if (utf8_bytes_needed == 0) {
if (this.inRange(_byte, 0x00, 0x7F)) {
code_point = _byte;
} else {
if (this.inRange(_byte, 0xC2, 0xDF)) {
utf8_bytes_needed = 1;
utf8_lower_boundary = 0x80;
utf8_code_point = _byte - 0xC0;
} else if (this.inRange(_byte, 0xE0, 0xEF)) {
utf8_bytes_needed = 2;
utf8_lower_boundary = 0x800;
utf8_code_point = _byte - 0xE0;
} else if (this.inRange(_byte, 0xF0, 0xF4)) {
utf8_bytes_needed = 3;
utf8_lower_boundary = 0x10000;
utf8_code_point = _byte - 0xF0;
} else {
this.decoderError(fatal);
}
utf8_code_point = utf8_code_point * Math.pow(64, utf8_bytes_needed);
code_point = null;
}
} else if (!this.inRange(_byte, 0x80, 0xBF)) {
utf8_code_point = 0;
utf8_bytes_needed = 0;
utf8_bytes_seen = 0;
utf8_lower_boundary = 0;
pos--;
code_point = this.decoderError(fatal, _byte);
} else {
utf8_bytes_seen += 1;
utf8_code_point = utf8_code_point + (_byte - 0x80) * Math.pow(64, utf8_bytes_needed - utf8_bytes_seen);
if (utf8_bytes_seen !== utf8_bytes_needed) {
code_point = null;
} else {
let cp = utf8_code_point;
let lower_boundary = utf8_lower_boundary;
utf8_code_point = 0;
utf8_bytes_needed = 0;
utf8_bytes_seen = 0;
utf8_lower_boundary = 0;
if (this.inRange(cp, lower_boundary, 0x10FFFF) && !this.inRange(cp, 0xD800, 0xDFFF)) {
code_point = cp;
} else {
code_point = this.decoderError(fatal, _byte);
}
}
}
}
//Decode string
if (code_point !== null && code_point !== this.EOF_code_point) {
if (code_point <= 0xFFFF) {
if (code_point > 0)result += String.fromCharCode(code_point);
} else {
code_point -= 0x10000;
result += String.fromCharCode(0xD800 + ((code_point >> 10) & 0x3ff));
result += String.fromCharCode(0xDC00 + (code_point & 0x3ff));
}
}
}
return result;
}
`
其他回答
在节点“缓冲区实例也是Uint8Array实例”,所以buf.toString()在这种情况下工作。
使用base64作为编码格式效果很好。这就是它是如何实现通过url在Firefox发送传递秘密。您将需要base64-js包。下面是Send源代码中的函数:
const b64 = require("base64-js")
function arrayToB64(array) {
return b64.fromByteArray(array).replace(/\+/g, "-").replace(/\//g, "_").replace(/=/g, "")
}
function b64ToArray(str) {
return b64.toByteArray(str + "===".slice((str.length + 3) % 4))
}
在一个Chrome示例应用程序中可以找到,尽管这意味着对于更大的数据块,您可以接受异步转换。
/**
* Converts an array buffer to a string
*
* @private
* @param {ArrayBuffer} buf The buffer to convert
* @param {Function} callback The function to call when conversion is complete
*/
function _arrayBufferToString(buf, callback) {
var bb = new Blob([new Uint8Array(buf)]);
var f = new FileReader();
f.onload = function(e) {
callback(e.target.result);
};
f.readAsText(bb);
}
到目前为止,对我来说最简单的方法是:
//1. Create or fetch the Uint8Array to use in the example
const bufferArray = new Uint8Array([10, 10, 10])
//2. Turn the Uint8Array into a regular array
const array = Array.from(bufferArray);
//3. Stringify it (option A)
JSON.stringify(array);
//3. Stringify it (option B: uses @serdarsenay code snippet to decode each item in array)
let binArrayToString = function(binArray) {
let str = "";
for (let i = 0; i < binArray.length; i++) {
str += String.fromCharCode(parseInt(binArray[i]));
}
return str;
}
binArrayToString(array);
以下是我使用的方法:
var str = String.fromCharCode.apply(null, uint8Arr);