Merge branch 'main' into patch-4

Author: oech3

.github/workflows/benchmarks.yml

@@ -37,6 +37,7 @@ jobs:
           - { package: uu_numfmt }
           - { package: uu_rm }
           - { package: uu_seq }
+          - { package: uu_shuf }
           - { package: uu_sort }
           - { package: uu_split }
           - { package: uu_tsort }

Cargo.lock

@@ -349,7 +349,6 @@ notify = { version = "=8.2.0", features = ["macos_kqueue"] }
 num-bigint = "0.4.4"
 num-prime = "0.4.4"
 num-traits = "0.2.19"
-number_prefix = "0.4"
 onig = { version = "~6.5.1", default-features = false }
 parse_datetime = "0.13.0"
 phf = "0.13.1"
@@ -373,6 +372,7 @@ textwrap = { version = "0.16.1", features = ["terminal_size"] }
 thiserror = "2.0.3"
 time = { version = "0.3.36" }
 unicode-width = "0.2.0"
+unit-prefix = "0.5"
 utmp-classic = "0.1.6"
 uutils_term_grid = "0.7"
 walkdir = "2.5"

Cargo.toml

@@ -27,3 +27,12 @@ fluent = { workspace = true }
 [[bin]]
 name = "shuf"
 path = "src/main.rs"
+
+[[bench]]
+name = "shuf_bench"
+harness = false
+
+[dev-dependencies]
+divan = { workspace = true }
+tempfile = { workspace = true }
+uucore = { workspace = true, features = ["benchmark"] }

fuzz/Cargo.lock

@@ -0,0 +1,53 @@
+// This file is part of the uutils coreutils package.
+//
+// For the full copyright and license information, please view the LICENSE
+// file that was distributed with this source code.
+
+use divan::{Bencher, black_box};
+use uu_shuf::uumain;
+use uucore::benchmark::{run_util_function, setup_test_file, text_data};
+
+/// Benchmark shuffling lines from a file
+/// Tests the default mode with a large number of lines
+#[divan::bench(args = [100_000])]
+fn shuf_lines(bencher: Bencher, num_lines: usize) {
+    let data = text_data::generate_by_lines(num_lines, 80);
+    let file_path = setup_test_file(&data);
+    let file_path_str = file_path.to_str().unwrap();
+
+    bencher.bench(|| {
+        black_box(run_util_function(uumain, &[file_path_str]));
+    });
+}
+
+/// Benchmark shuffling a numeric range with -i
+/// Tests the input-range mode which uses a different algorithm
+#[divan::bench(args = [1_000_000])]
+fn shuf_input_range(bencher: Bencher, range_size: usize) {
+    let range_arg = format!("1-{range_size}");
+
+    bencher.bench(|| {
+        black_box(run_util_function(uumain, &["-i", &range_arg]));
+    });
+}
+
+/// Benchmark shuffling with repeat (sampling with replacement)
+/// Tests the -r flag combined with -n to output a specific count
+#[divan::bench(args = [50_000])]
+fn shuf_repeat_sampling(bencher: Bencher, num_lines: usize) {
+    let data = text_data::generate_by_lines(10_000, 80);
+    let file_path = setup_test_file(&data);
+    let file_path_str = file_path.to_str().unwrap();
+    let count = format!("{num_lines}");
+
+    bencher.bench(|| {
+        black_box(run_util_function(
+            uumain,
+            &["-r", "-n", &count, file_path_str],
+        ));
+    });
+}
+
+fn main() {
+    divan::main();
+}

src/uu/shuf/Cargo.toml

@@ -26,7 +26,7 @@ bstr = { workspace = true }
 chrono = { workspace = true, optional = true }
 clap = { workspace = true }
 uucore_procs = { workspace = true }
-number_prefix = { workspace = true }
+unit-prefix = { workspace = true }
 phf = { workspace = true }
 dns-lookup = { workspace = true, optional = true }
 dunce = { version = "1.0.4", optional = true }

src/uu/shuf/benches/shuf_bench.rs

@@ -9,7 +9,7 @@
 //!
 //! Format sizes like gnulibs human_readable() would
 
-use number_prefix::NumberPrefix;
+use unit_prefix::NumberPrefix;
 
 #[derive(Copy, Clone, PartialEq)]
 pub enum SizeFormat {

src/uucore/Cargo.toml

@@ -121,15 +121,17 @@ test -f "${UU_BUILD_DIR}/[" || (cd ${UU_BUILD_DIR} && ln -s "test" "[")
 
 cd "${path_GNU}" && echo "[ pwd:'${PWD}' ]"
 
+# Always update the PATH to test the uutils coreutils instead of the GNU coreutils
+# This ensures the correct path is used even if the repository was moved or rebuilt in a different location
+sed -i "s/^[[:blank:]]*PATH=.*/  PATH='${UU_BUILD_DIR//\//\\/}\$(PATH_SEPARATOR)'\"\$\$PATH\" \\\/" tests/local.mk
+
 if test -f gnu-built; then
     echo "GNU build already found. Skip"
     echo "'rm -f $(pwd)/gnu-built' to force the build"
     echo "Note: the customization of the tests will still happen"
 else
     # Disable useless checks
     sed -i 's|check-texinfo: $(syntax_checks)|check-texinfo:|' doc/local.mk
-    # Change the PATH to test the uutils coreutils instead of the GNU coreutils
-    sed -i "s/^[[:blank:]]*PATH=.*/  PATH='${UU_BUILD_DIR//\//\\/}\$(PATH_SEPARATOR)'\"\$\$PATH\" \\\/" tests/local.mk
     ./bootstrap --skip-po
     ./configure --quiet --disable-gcc-warnings --disable-nls --disable-dependency-tracking --disable-bold-man-page-references \
       "$([ "${SELINUX_ENABLED}" = 1 ] && echo --with-selinux || echo --without-selinux)"