Slices, exercise 5: why does a Chinese character take 3 bytes?

[s-tikhomirov]

I don't understand the solution to exercise 5 about slices:

fn main() {
    let s = "你好，世界";
    // Modify this line to make the code work
    let slice = &s[0..2];

    assert!(slice == "你");

    println!("Success!");
}

The solution is:

fn main() {
    let s = "你好，世界";
    let slice = &s[0..3];

    assert!(slice == "你");
}

Earlier, a comment to exercise 2 said that

Each of the two chars '中' and '国' occupies 4 bytes, 2 * 4 = 8

so I assumed 你 would occupy 4 bytes, and the slice would be &s[0..4]. Yet the correct answer is &s[0..3]. Apparently, in the fifth exercise, each Chinese character takes 3 bytes. Why so?

[Mohammed785]

I think this might help you to understand
https://doc.rust-lang.org/book/ch08-02-strings.html#bytes-and-scalar-values-and-grapheme-clusters-oh-my