expand: add a benchmark

Author: sylvestre

Cargo.lock

@@ -24,6 +24,15 @@ uucore = { workspace = true }
 thiserror = { workspace = true }
 fluent = { workspace = true }
 
+[dev-dependencies]
+divan = { workspace = true }
+tempfile = { workspace = true }
+uucore = { workspace = true, features = ["benchmark"] }
+
 [[bin]]
 name = "expand"
 path = "src/main.rs"
+
+[[bench]]
+name = "expand_bench"
+harness = false

src/uu/expand/Cargo.toml

@@ -0,0 +1,79 @@
+// 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_expand::uumain;
+use uucore::benchmark::{create_test_file, run_util_function, text_data};
+
+/// Helper function to run expand benchmark with generated data
+fn bench_expand(bencher: Bencher, data: impl AsRef<[u8]>, args: &[&str]) {
+    let temp_dir = tempfile::tempdir().unwrap();
+    let file_path = create_test_file(data.as_ref(), temp_dir.path());
+    let file_path_str = file_path.to_str().unwrap();
+
+    let mut all_args = vec![];
+    all_args.extend_from_slice(args);
+    all_args.push(file_path_str);
+
+    bencher.bench(|| {
+        black_box(run_util_function(uumain, &all_args));
+    });
+}
+
+/// Benchmark expanding tabs on files with many short lines
+#[divan::bench(args = [100_000])]
+fn expand_many_lines(bencher: Bencher, num_lines: usize) {
+    let data = (0..num_lines)
+        .map(|i| format!("line{}\tvalue{}\tdata{}\n", i, i * 2, i * 3))
+        .collect::<String>();
+    bench_expand(bencher, data, &[]);
+}
+
+/// Benchmark expanding tabs with custom tab stops
+#[divan::bench(args = [50_000])]
+fn expand_custom_tabstops(bencher: Bencher, num_lines: usize) {
+    let data = (0..num_lines)
+        .map(|i| format!("a\tb\tc\td\te{}\n", i))
+        .collect::<String>();
+    bench_expand(bencher, data, &["--tabs=4,8,12"]);
+}
+
+/// Benchmark expanding tabs on large file (by size)
+#[divan::bench(args = [10])]
+fn expand_large_file(bencher: Bencher, size_mb: usize) {
+    let line = "word1\tword2\tword3\tword4\tword5\n";
+    let repetitions = (size_mb * 1024 * 1024) / line.len();
+    let data = line.repeat(repetitions);
+    bench_expand(bencher, data, &[]);
+}
+
+/// Benchmark with initial-only flag (only expand leading tabs)
+#[divan::bench(args = [50_000])]
+fn expand_initial_only(bencher: Bencher, num_lines: usize) {
+    let data = (0..num_lines)
+        .map(|i| format!("\t\tindented{}\tdata\n", i))
+        .collect::<String>();
+    bench_expand(bencher, data, &["--initial"]);
+}
+
+/// Benchmark with no tabs (worst case - all processing, no expansion)
+#[divan::bench(args = [100_000])]
+fn expand_no_tabs(bencher: Bencher, num_lines: usize) {
+    let data = text_data::generate_by_lines(num_lines, 80);
+    bench_expand(bencher, data, &[]);
+}
+
+/// Benchmark with UTF-8 disabled (ASCII mode)
+#[divan::bench(args = [50_000])]
+fn expand_ascii_mode(bencher: Bencher, num_lines: usize) {
+    let data = (0..num_lines)
+        .map(|i| format!("line{}\tvalue{}\n", i, i))
+        .collect::<String>();
+    bench_expand(bencher, data, &["--no-utf8"]);
+}
+
+fn main() {
+    divan::main();
+}