Tracking Issue for `const_is_char_boundary`

[zachs18]

Feature gate: #![feature(const_is_char_boundary)]

This is a tracking issue for using str::is_char_boundary in const, which allows checking that index-th byte is the first byte in a UTF-8 code point sequence or the end of the string during const-eval.

Public API

// core::str
impl str {
    pub const fn is_char_boundary(&self, index: usize) -> bool;
}

Steps / History

• Implementation: Mark str::is_char_boundary and str::split_at* unstably const. #131520
• Final comment period (FCP)¹
• Stabilization PR

Unresolved Questions

• None yet.

Footnotes

1. https://std-dev-guide.rust-lang.org/feature-lifecycle/stabilization.html ↩

[okaneco]

I was looking at this as well, it opens the door for making str::split_at* and the round_char_boundary unstable feature functions const (the latter needs the iterator code to be rewritten).

I don't believe the check for index == 0 is necessary so one possible implementation is as follows.

pub const fn is_char_boundary(&self, index: usize) -> bool {
    if index < self.as_bytes().len() {
        self.as_bytes()[index].is_utf8_char_boundary()
    } else {
        // The start and end of a string are considered boundaries
        index == self.len()
    }
}

[zachs18]

Yeah, this was motivated by making str::split_at* const (cc #131518, which #131520 also implements). I looked at {floor,ceil}_char_boundary but did not want to mess with them at the moment since the rest were relatively simple changes.