From bd916b4746822dd839bcf1f57b2fbed068f84534 Mon Sep 17 00:00:00 2001
From: xtqqczze <45661989+xtqqczze@users.noreply.github.com>
Date: Tue, 17 Mar 2026 02:43:51 +0000
Subject: [PATCH] shuf: use memchr in split_seps

---
 Cargo.lock              | 1 +
 src/uu/shuf/Cargo.toml  | 5 +++--
 src/uu/shuf/src/shuf.rs | 8 ++++----
 3 files changed, 8 insertions(+), 6 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index c5bd7b89b36..c55cc64036f 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -4059,6 +4059,7 @@ dependencies = [
  "codspeed-divan-compat",
  "fluent",
  "itoa",
+ "memchr",
  "rand 0.10.0",
  "rand_chacha 0.10.0",
  "rustc-hash",
diff --git a/src/uu/shuf/Cargo.toml b/src/uu/shuf/Cargo.toml
index d59f022e1af..7eaa497475f 100644
--- a/src/uu/shuf/Cargo.toml
+++ b/src/uu/shuf/Cargo.toml
@@ -20,13 +20,14 @@ path = "src/shuf.rs"
 
 [dependencies]
 clap = { workspace = true }
+fluent = { workspace = true }
 itoa = { workspace = true }
+memchr = { workspace = true }
 rand = { workspace = true }
 rand_chacha = { workspace = true }
+rustc-hash = { workspace = true }
 sha3 = { workspace = true }
 uucore = { workspace = true }
-fluent = { workspace = true }
-rustc-hash = { workspace = true }
 
 [[bin]]
 name = "shuf"
diff --git a/src/uu/shuf/src/shuf.rs b/src/uu/shuf/src/shuf.rs
index de725cc4e20..2b6ed70fbcc 100644
--- a/src/uu/shuf/src/shuf.rs
+++ b/src/uu/shuf/src/shuf.rs
@@ -13,6 +13,7 @@ use std::path::{Path, PathBuf};
 use std::str::FromStr;
 
 use clap::{Arg, ArgAction, Command, builder::ValueParser};
+use memchr::memchr_iter;
 use rand::{
     RngExt as _,
     rngs::ThreadRng,
@@ -267,14 +268,13 @@ fn read_input_file(filename: &Path) -> UResult<Vec<u8>> {
     }
 }
 
-fn split_seps(data: &[u8], sep: u8) -> Vec<&[u8]> {
+pub fn split_seps(data: &[u8], sep: u8) -> Vec<&[u8]> {
     // A single trailing separator is ignored.
     // If data is empty (and does not even contain a single 'sep'
     // to indicate the presence of an empty element), then behave
     // as if the input contained no elements at all.
-    const PREDICTED_LINE_LENGTH: usize = 64;
-    let predicted_capacity = data.len() / PREDICTED_LINE_LENGTH;
-    let mut elements = Vec::with_capacity(predicted_capacity);
+    let sep_count = memchr_iter(sep, data).count();
+    let mut elements = Vec::with_capacity(sep_count + 1);
     elements.extend(data.split(|&b| b == sep));
     let _ = elements.pop_if(|e| e.is_empty());
     elements