Regarding axum update to matchit v0.8

The new matching syntax breaks path_colon_params in:

Line 23 in 7e2d71e

pub fn path_colon_params(s: &str) -> Cow<str> {

This is my, not so good, idea to fix it:

pub fn path_colon_params(s: &str) -> Cow<str> {
    // if !s.contains(':') {
    //     return s.into();
    // }

    let mut rewritten = String::with_capacity(s.len());

    #[derive(Clone, Copy)]
    enum State {
        None,
        // this state is no longer needed?
        // WasParam,
        WasWildcard,
    }
    let mut state = State::None;
    let chars_pair = s.chars().zip(s.chars().skip(1));
    let mut do_escape_next = false;

    // look at https://github.com/ibraheemdev/matchit/blob/36e5398589129896c5dad8a6b2580ede59831875/src/escape.rs#L18
    for (i, (c, c_next)) in chars_pair.enumerate() {
        if do_escape_next {
            do_escape_next = false;
            continue;
        }
        if (c == '{' && c_next == '{') || (c == '}' && c_next == '}') {
            // This poses a problem when the url have '{' or '}' escaped not sure how and where to handle this
            rewritten.push(c);
            // skip the next iteration
            do_escape_next = true;
            continue;
        }
        match (state, c, c_next) {
            (State::None, '{', '*') => {
                rewritten.push('{');
                state = State::WasWildcard;
            }
            // This is if, a normal Param is allowed after a wildcard
            // (not sure but I think this is a planned feature in the future?)
            (State::WasWildcard, _, '}') => {
                rewritten.push(c);
                rewritten += "+}";
                state = State::None;
            }
            (_, _, _) => {
                rewritten.push(c);
            }
        }
        if i == s.len() - 2 {
            rewritten.push(c_next);
        }
    }

    rewritten.into()
}

I haven't tested this fully, especially wildcards params, but tested it briefly on:

[patch.crates-io]
axum = { git = 'https://github.com/mladedav/axum.git', rev = "db825f94ec849d39f02f8e9cfc469613680c08f4" }

The new syntax is almost the same except we still need to convert the end if it is {*...} To {...+}.

Code should be:
1: get substring after last /
2: if it starts with {* and ends with } copy string until last / and append contents as { content +} else just return original string.