fix lite: peek space skip comment

[YydsNone]

I'm now using [email protected] and noticed that pattern (?x)[0-#-contrived comment!\n9]+ and pattern (?x)[0-# contrived comment!\n9]+ behave differently. I think the cause is that Parser's peek_space method doesn't skip comment correctly (not consistent with bump_space method). below is a simple comparison between (?x)[0-#-contrived comment!\n9]+, (?x)[0-#]contrived comment!\n9]+ and (?x)[0-# contrived comment!\n9]+.

#[cfg(test)]
mod tests {
    use regex_lite::Regex;

    #[test]
    fn test_regex_lite() {
        let hyphen_pattern = "(?x)[0-#-contrived comment!\n9]+";
        let right_square_bracket_pattern = "(?x)[0-#]contrived comment!\n9]+";
        let another_pattern = "(?x)[0-# contrived comment!\n9]+";

        let haystack = "0123456789";

        let hyphen_match = Regex::new(hyphen_pattern).unwrap().find(haystack).unwrap();
        let right_square_bracket_match = Regex::new(right_square_bracket_pattern)
            .unwrap()
            .find(haystack)
            .unwrap();
        let another_match = Regex::new(another_pattern).unwrap().find(haystack).unwrap();
        assert_eq!(hyphen_match, right_square_bracket_match);
        assert_eq!(hyphen_match, another_match);
    }
}

test tests::test_regex_lite ... FAILED

failures:

---- tests::test_regex_lite stdout ----
thread 'tests::test_regex_lite' panicked at src/main.rs:24:9:
assertion `left == right` failed
  left: Match { start: 0, end: 1, string: "0" }
 right: Match { start: 0, end: 10, string: "0123456789" }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

[BurntSushi]

Can you compare this with how regex behaves please?

[YydsNone]

Can you compare this with how regex behaves please?

#[cfg(test)]
mod tests {
    use regex::Regex;

    #[test]
    fn test_regex_lite() {
        let hyphen_pattern = "(?x)[0-#-contrived comment!\n9]+";
        let right_square_bracket_pattern = "(?x)[0-#]contrived comment!\n9]+";
        let another_pattern = "(?x)[0-# contrived comment!\n9]+";

        let haystack = "0123456789";

        let hyphen_match = Regex::new(hyphen_pattern).unwrap().find(haystack).unwrap();
        let right_square_bracket_match = Regex::new(right_square_bracket_pattern)
            .unwrap()
            .find(haystack)
            .unwrap();
        let another_match = Regex::new(another_pattern).unwrap().find(haystack).unwrap();
        assert_eq!(hyphen_match, right_square_bracket_match);
        assert_eq!(hyphen_match, another_match);
    }
}

running 1 test
test tests::test_regex_lite ... FAILED

failures:

---- tests::test_regex_lite stdout ----
thread 'tests::test_regex_lite' panicked at src/main.rs:24:9:
assertion `left == right` failed
  left: Match { start: 0, end: 1, string: "0" }
 right: Match { start: 0, end: 10, string: "0123456789" }
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

[YydsNone]

test codes are basically the same, it seems that the behavior of regex and regex-lite is consistent

[YydsNone]

(!in_commit || ch == '#') -> (!in_commit && ch == '#')