hashmap on Buffer objects appears to be wrong, why

There are two different Buffer objects (different values) in a JavaScript, but in JavaScript object, they are treated as being equal. Any idea why? How to force them to be different keys for a JavaScript? Thanks

my node.js version is 10.15.1 on Ubuntu 18.04.

var buf1 = Buffer.from([
    10,
    33,
    45,
    77,
    10,
    33,
    46,
    166,
    140,
    190,
    5,
    241
  ]);
var buf2 = new Buffer.from([
    10,
    33,
    45,
    77,
    10,
    33,
    46,
    168,
    215,
    216,
    5,
    241
]);
let a = {};
a[buf2] = "jon";
console.log(a[buf1]);  //output  jon

>Solution :

a[buf2] is equivalent to a[buf2.toString()], which is equivalent to a[buf2.toString('utf-8')], which is the cause of your problem: your buffers don’t hold valid UTF-8 data, so when converting to an UTF-8 string, invalid code points are replaced by the Unicode replacement character, U+FFFD (encoded as 0xEF 0xBF 0xBD in UTF-8).

To show this, you can run the following code:

console.log('buf1', Buffer.from(buf1.toString()).toString('hex'));
console.log('buf2', Buffer.from(buf2.toString()).toString('hex'));

The result is the hexadecimal representation of each buffer as it’s used as a property "name" (replacement character highlighted):

buf1 0a212d4d0a212eefbfbdefbfbdefbfbd05efbfbd
                   ^^^^^^^^^^^^^^^^^^  ^^^^^^
buf2 0a212d4d0a212eefbfbdefbfbdefbfbd05efbfbd

As you can see, they are equivalent.