Go和Node.js base64编码，解码问题

问题描述：

I am base64 encoding a string in Go

Then decoding in javascript (I have tried 3 different methods)

I cannot get the javascript result to match the original Go string

Go (encode)

a := []byte {138,143,163,224,178,73,161,15,240,121,53,192,198,182,52,245}
fmt.Println("a", string(a), a, len(a))
b := base64.StdEncoding.EncodeToString(a)
fmt.Println("b", b, []byte(b), len([]byte(b)))

js (decode)

 const b = [105,111,43,106,52,76,74,74,111,81,47,119,101,84,88,65,120,114,89,48,57,81,61,61];
 let bString = aesjs.utils.utf8.fromBytes(b);
 console.log("b", bString, b, b.length);

 let a1String = atob(bString);
 let a2String = Base64.decode(bString);
 let a3String = Buffer.from(bString, 'base64').toString('utf8');
 let a1 = aesjs.utils.utf8.toBytes(a1String);
 let a2 = aesjs.utils.utf8.toBytes(a2String);
 let a3 = aesjs.utils.utf8.toBytes(a3String);
 console.log("a", a1, a1.length, a2, a2.length, a3, a3.length);

All 3 methods fail, i.e. a1 != a, a2 != a, a3 != a

I assumed base64 encoding/decoding would be simple

What am I missing? Thanks

EDIT: The js code had a 'typo' - but the problem remains, that ai != a, even a1 != a2 == a3

EDIT: If the original a is simple, then everything works. But when a contains non-printable chars, I cannot decode in js Example where it works: a := []byte {65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65} Here it does not works: a := []byte {138, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65, 65}

我在Go中使用base64编码字符串 p>

然后在javascript中解码（我尝试了3种不同的方法） p>

我无法获得与原始Go字符串匹配的javascript结果 p>

Go（编码） strong> p>

  a：= [] byte {138,143,163,224,178,73,161,15,240,121,53,192,198,182,52,245} 
fmt.Println（“ a”，string（a），a，len  （a））
b：= base64.StdEncoding.EncodeToString（a）
fmt.Println（“ b”，b，[] byte（b），len（[] byte（b）））
  code>   pre> 
 
   js（解码） strong>  p> 
 
 
  const b = [105,111,43,106,52,76,74，  74,111,81,47,119,101,84,88,65,120,114,89,48,57,81,61,61]; 
让bString = aesjs.utils.utf8.fromBytes（b）; 
 console.log（“ b”  ，bString，b，b.length）; 
 
让a1String = atob（bString）; 
让a2String = Base64.decode（bString）; 
让a3String = Buffer.from（bString，'base64'）。  toString（'utf8'）; 
让a1 = aesjs.utils.utf8.toBytes（a1String）; 
让a2 = aesjs.utils.utf8.toBytes（a2String）; 
让a3 = aesj  s.utils.utf8.toBytes（a3String）; 
 console.log（“ a”，a1，a1.length，a2，a2.length，a3，a3.length）; 
  code>  pre>  
 
 所有3种方法均会失败，即a1！= a，a2！= a，a3！= a  p> 
 
 
我认为base64编码/解码将很简单 p  > 
 
 
我缺少什么？ 谢谢 p> 
 
 
编辑：js代码中有'typo'-但问题仍然存在，即ai！= a，甚至a1！= a2 == a3  p> 
 
  
 编辑：如果原始的a很简单，那么一切正常。 strong>但是，当a包含不可打印的字符时，我无法在js 
中解码它的工作方式：
a：= [] byte  {65，65，65，65，65，65，65，65，65，65，65，65，65，65，65，65} 
此处不起作用：
a：= [] byte {138，  65，65，65，65，65，65，65，65，65，65，65，65，65，65，65}  p> 
  div>

答

In Go, strings are UTF8:

in := "Good afternoon, or こんにちは"
buf := []byte(in)
str := base64.StdEncoding.EncodeToString(buf)
fmt.Println(str)

That prints the base64 encoded (ASCII safe) representation: R29vZCBhZnRlcm5vb24sIG9yIOOBk+OCk+OBq+OBoeOBrw==

JS strings are UTF-16. So you need to both decode from base64 and convert from utf8 to utf-16. You could use a library, or a helper function, for example:

function b64DecodeUnicode(str) {
    return decodeURIComponent(atob(str).split('').map(function(c) {
        return '%' + ('00' + c.charCodeAt(0).toString(16)).slice(-2);
    }).join(''));
}

var str = "R29vZCBhZnRlcm5vb24sIG9yIOOBk+OCk+OBq+OBoeOBrw==";
var result = b64DecodeUnicode(str);
"Good afternoon, or こんにちは"

The b64DecodeUnicode() function is copied from this answer, see also the link to MDN for detailed info.

答

From: https://github.com/ricmoo/aes-js/blob/master/README.md

UTF8 should NOT be used to store arbitrary binary data as it is a string encoding format, not a binary encoding format

So don't used utf8 related functions to convert raw (non utf8) binary data.

Go和Node.js base64编码，解码问题

相关推荐