Mercurial > public > mercurial-scm > hg-stable
comparison rust/hg-core/src/utils.rs @ 52316:f4aede0f01af
rust-manifest: use `memchr` crate for all byte-finding needs
While writing a very dumb manifest diffing algorithm for a proof-of-concept
I saw that `Manifest::find_by_path` was much slower than I was expecting.
It turns out that the Rust stdlib uses slow (all is relative) code when
searching for byte positions for reasons ranging from portability, SIMD
API stability, nobody doing the work, etc. `memch` is much faster for these
purposes, so let's use it.
I was measuring ~670ms of profile time in `find_by_path`, after this patch
it went down to ~230ms.
author | Rapha?l Gom?s <rgomes@octobus.net> |
---|---|
date | Tue, 12 Nov 2024 23:20:04 +0100 |
parents | e6a44bc91bc2 |
children | 36d39726c0af |
comparison
equal
deleted
inserted
replaced
52315:fad30cb98579 | 52316:f4aede0f01af |
---|---|
132 None | 132 None |
133 } | 133 } |
134 } | 134 } |
135 | 135 |
136 fn split_2(&self, separator: u8) -> Option<(&[u8], &[u8])> { | 136 fn split_2(&self, separator: u8) -> Option<(&[u8], &[u8])> { |
137 let mut iter = self.splitn(2, |&byte| byte == separator); | 137 let pos = memchr::memchr(separator, self)?; |
138 let a = iter.next()?; | 138 Some((&self[..pos], &self[pos + 1..])) |
139 let b = iter.next()?; | |
140 Some((a, b)) | |
141 } | 139 } |
142 | 140 |
143 fn split_2_by_slice(&self, separator: &[u8]) -> Option<(&[u8], &[u8])> { | 141 fn split_2_by_slice(&self, separator: &[u8]) -> Option<(&[u8], &[u8])> { |
144 find_slice_in_slice(self, separator) | 142 find_slice_in_slice(self, separator) |
145 .map(|pos| (&self[..pos], &self[pos + separator.len()..])) | 143 .map(|pos| (&self[..pos], &self[pos + separator.len()..])) |