Suggest `&string[6..11]` for `str::from_utf8(string.as_bytes()[6..11]).unwrap()`

[Luro02]

This

use std::str;

let string = "Hello World!";
assert_eq!(str::from_utf8(string.as_bytes()[6..11]).unwrap(), "World");

could be written as

let string = "Hello World!";
assert_eq!(&string[6..11], "World");

[l0calh05t]

assert_eq!(str::from_utf8(string.as_bytes()[6..11]).unwrap(), "World");

should be

assert_eq!(str::from_utf8(&string.as_bytes()[6..11]).unwrap(), "World");

[illicitonion]

I believe these are only equivalent if string is only made of single-byte characters. Counter-example:

let string = "🐈ello World!";
assert_eq!(::std::str::from_utf8(&string.as_bytes()[6..11]).unwrap(), "World");

^^ This assert fails

[Luro02]

Well of course your example fails. Neither string[6..11] nor ::std::str::from_utf8(&string.as_bytes()[6..11]).unwrap() are equal to "World".

The indexing works with bytes and not chars, which is why your example is not working as expected, because the emoji ("🐈".len()) is 4 bytes in length.

For example

let string = "🐈ello World!";

dbg!(&string[1..11]);

will fail with

thread 'main' panicked at 'byte index 1 is not a char boundary; it is inside '🐈' (bytes 0..4) of `🐈ello World!`', src/libcore/str/mod.rs:2219:5
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

This is your example working correctly:

use std::str;

fn main() {
    let string = "🐈ello World!";

    assert_eq!(&string[6..11], "lo Wo");

    assert_eq!(
        str::from_utf8(&string.as_bytes()[6..11]).unwrap(),
        &string[6..11]
    );
}

playground

The above code succeeds, which means that str::from_utf8(&string.as_bytes()[6..11]).unwrap() and &string[6..11] are equal, so the lint is also valid for utf8 strings.

[illicitonion]

Well of course your example fails. Neither string[6..11] nor ::std::str::from_utf8(&string.as_bytes()[6..11]).unwrap() are equal to "World".

The indexing works with bytes and not chars, which is why your example is not working as expected, because the emoji ("🐈".len()) is 4 bytes in length.

Apologies, of course you're right - for some reason I was thinking slicing strings worked on chars and as you've shown I put together a really poor example to try to show that! Thanks for taking the time to make a better one which shows the truth :)