JavaScript and php encoding output are inconsistent

[huanshiwushuang]

Use JavaScript: Base62x.encode('哈')
output: 'vPE8'

But,

Use PHP: Base62x::encode('哈')->get();
output: 'vU5'

Hope to get your help, thanks

[huanshiwushuang]

maybe not support chinese ?

[wadelau]

I had troubleshooted into the issue and found that the point was mb_str_split funtion which is a built-in function after PHP-7.4.

The Chinese character "哈" with UTF-8 format will be split into a Character-Array as ['哈'] by mb_str_split.
However, the same Chinese character will be split into a Byte-Array as [229, 147, 136] by str_split, which is expected in src/Encoding/Base62x::encode.

We have not figured out why the author of this repo changed the function from str_split to mb_str_split.

My advice is to use str_split for byte-array, mb_str_split for character-array.
Byte is equivalent to character in ASCII, so American-European may have no sense of such issues.
In fact, one character may contains one byte, two bytes, three bytes or even more with UTF-8 charset.

The suggestion is to replace mb_str_split with str_split in src/Encoding/Base62x for test.