From 245bf503e2a948ac98170516d11df632e85a948b Mon Sep 17 00:00:00 2001 From: bendn Date: Sun, 25 May 2025 02:01:04 +0700 Subject: [PATCH] increase perf of charsearcher for single ascii characters --- library/core/src/str/iter.rs | 2 +- library/core/src/str/pattern.rs | 32 +++++++++++++++++++++++++++++++- 2 files changed, 32 insertions(+), 2 deletions(-) diff --git a/library/core/src/str/iter.rs b/library/core/src/str/iter.rs index 425c4eaee28ee..49c581f352eb3 100644 --- a/library/core/src/str/iter.rs +++ b/library/core/src/str/iter.rs @@ -656,7 +656,7 @@ impl<'a, P: Pattern> SplitInternal<'a, P> { None } - #[inline] + #[inline(always)] fn next(&mut self) -> Option<&'a str> { if self.finished { return None; diff --git a/library/core/src/str/pattern.rs b/library/core/src/str/pattern.rs index bcbbb11c83b2f..e8189a2187b63 100644 --- a/library/core/src/str/pattern.rs +++ b/library/core/src/str/pattern.rs @@ -429,8 +429,23 @@ unsafe impl<'a> Searcher<'a> for CharSearcher<'a> { SearchStep::Done } } - #[inline] + #[inline(always)] fn next_match(&mut self) -> Option<(usize, usize)> { + if self.utf8_size == 1 { + return match self + .haystack + .as_bytes() + .get(self.finger..self.finger_back)? + .iter() + .position(|x| *x == self.utf8_encoded[0]) + { + Some(x) => { + self.finger += x + 1; + Some((self.finger - 1, self.finger)) + } + None => None, + }; + } loop { // get the haystack after the last character found let bytes = self.haystack.as_bytes().get(self.finger..self.finger_back)?; @@ -498,6 +513,21 @@ unsafe impl<'a> ReverseSearcher<'a> for CharSearcher<'a> { } #[inline] fn next_match_back(&mut self) -> Option<(usize, usize)> { + if self.utf8_size == 1 { + return match self + .haystack + .get(self.finger..self.finger_back)? + .as_bytes() + .iter() + .rposition(|&x| x == self.utf8_encoded[0]) + { + Some(x) => { + self.finger_back = self.finger + x; + Some((self.finger_back, self.finger_back + 1)) + } + None => None, + }; + } let haystack = self.haystack.as_bytes(); loop { // get the haystack up to but not including the last character searched