Split a string keeping the separators

Question

Is there a trivial way to split a string keeping the separators? Instead of this:

let texte = \"Ten. Million. Questions. Let\'s celebrate all we\'ve done tog         


        

Answer 1

I was not able to find anything in the standard library, so I wrote my own:

This version uses the unstable pattern API as it's more flexible, but the link above has a fallback that I've hardcoded for my specific stable usecase.

#![feature(pattern)]

use std::str::pattern::{Pattern, Searcher};

#[derive(Copy, Clone, Debug, PartialEq)]
pub enum SplitType<'a> {
    Match(&'a str),
    Delimiter(&'a str),
}

pub struct SplitKeepingDelimiter<'p, P>
where
    P: Pattern<'p>,
{
    searcher: P::Searcher,
    start: usize,
    saved: Option,
}

impl<'p, P> Iterator for SplitKeepingDelimiter<'p, P>
where
    P: Pattern<'p>,
{
    type Item = SplitType<'p>;

    fn next(&mut self) -> Option {
        if self.start == self.searcher.haystack().len() {
            return None;
        }

        if let Some(end_of_match) = self.saved.take() {
            let s = &self.searcher.haystack()[self.start..end_of_match];
            self.start = end_of_match;
            return Some(SplitType::Delimiter(s));
        }

        match self.searcher.next_match() {
            Some((start, end)) => {
                if self.start == start {
                    let s = &self.searcher.haystack()[start..end];
                    self.start = end;
                    Some(SplitType::Delimiter(s))
                } else {
                    let s = &self.searcher.haystack()[self.start..start];
                    self.start = start;
                    self.saved = Some(end);
                    Some(SplitType::Match(s))
                }
            }
            None => {
                let s = &self.searcher.haystack()[self.start..];
                self.start = self.searcher.haystack().len();
                Some(SplitType::Match(s))
            }
        }
    }
}

pub trait SplitKeepingDelimiterExt: ::std::ops::Index<::std::ops::RangeFull, Output = str> {
    fn split_keeping_delimiter
(&self, pattern: P) -> SplitKeepingDelimiter

    where
        P: for<'a> Pattern<'a>,
    {
        SplitKeepingDelimiter {
            searcher: pattern.into_searcher(&self[..]),
            start: 0,
            saved: None,
        }
    }
}

impl SplitKeepingDelimiterExt for str {}

#[cfg(test)]
mod test {
    use super::SplitKeepingDelimiterExt;

    #[test]
    fn split_with_delimiter() {
        use super::SplitType::*;
        let delims = &[',', ';'][..];
        let items: Vec<_> = "alpha,beta;gamma".split_keeping_delimiter(delims).collect();
        assert_eq!(
            &items,
            &[
                Match("alpha"),
                Delimiter(","),
                Match("beta"),
                Delimiter(";"),
                Match("gamma")
            ]
        );
    }

    #[test]
    fn split_with_delimiter_allows_consecutive_delimiters() {
        use super::SplitType::*;
        let delims = &[',', ';'][..];
        let items: Vec<_> = ",;".split_keeping_delimiter(delims).collect();
        assert_eq!(&items, &[Delimiter(","), Delimiter(";")]);
    }
}

You'll note that I needed to track if something was one of the delimiters or not, but that should be easy to adapt if you don't need it.