add benchmark for join

Author: sylvestre

.github/workflows/benchmarks.yml

@@ -31,6 +31,7 @@ jobs:
           - { package: uu_du }
           - { package: uu_expand }
           - { package: uu_fold }
+          - { package: uu_join }
           - { package: uu_ls }
           - { package: uu_mv }
           - { package: uu_nl }

Cargo.lock

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

src/uu/join/Cargo.toml

@@ -0,0 +1,141 @@
+// 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_join::uumain;
+use uucore::benchmark::{run_util_function, setup_test_file};
+
+/// Benchmark basic join with sorted data
+#[divan::bench]
+fn join_basic_sorted(bencher: Bencher) {
+    bencher.bench(|| {
+        let num_lines = 10000;
+        let mut file1_data = Vec::new();
+        let mut file2_data = Vec::new();
+
+        // Generate sorted test data with 50% overlap
+        for i in 0..num_lines {
+            let key = if i < num_lines / 2 {
+                format!("key_{:06}", i)
+            } else {
+                format!("unique1_{:06}", i - num_lines / 2)
+            };
+
+            file1_data.extend_from_slice(
+                format!("{}\tfile1_field1_{}\tfile1_field2_{}\n", key, i, i).as_bytes(),
+            );
+
+            let key2 = if i < num_lines / 2 {
+                format!("key_{:06}", i)
+            } else {
+                format!("unique2_{:06}", i - num_lines / 2)
+            };
+
+            file2_data.extend_from_slice(
+                format!("{}\tfile2_field1_{}\tfile2_field2_{}\n", key2, i, i).as_bytes(),
+            );
+        }
+
+        // Sort the data
+        let file1_lines: Vec<&str> = std::str::from_utf8(&file1_data).unwrap().lines().collect();
+        let file2_lines: Vec<&str> = std::str::from_utf8(&file2_data).unwrap().lines().collect();
+
+        let mut sorted_file1: Vec<_> = file1_lines.clone();
+        let mut sorted_file2: Vec<_> = file2_lines.clone();
+
+        sorted_file1.sort();
+        sorted_file2.sort();
+
+        let sorted_file1_data = (sorted_file1.join("\n") + "\n").into_bytes();
+        let sorted_file2_data = (sorted_file2.join("\n") + "\n").into_bytes();
+
+        let file1_path = setup_test_file(&sorted_file1_data);
+        let file2_path = setup_test_file(&sorted_file2_data);
+
+        black_box(run_util_function(
+            uumain,
+            &[file1_path.to_str().unwrap(), file2_path.to_str().unwrap()],
+        ));
+    });
+}
+
+/// Benchmark join with custom delimiter
+#[divan::bench]
+fn join_custom_delimiter(bencher: Bencher) {
+    bencher.bench(|| {
+        let num_lines = 10000;
+        let mut file1_data = Vec::new();
+        let mut file2_data = Vec::new();
+
+        // Generate CSV-style data
+        for i in 0..num_lines {
+            let key = format!("key_{:06}", i);
+            file1_data.extend_from_slice(format!("{},value1_{},data1_{}\n", key, i, i).as_bytes());
+            file2_data.extend_from_slice(format!("{},value2_{},data2_{}\n", key, i, i).as_bytes());
+        }
+
+        let file1_path = setup_test_file(&file1_data);
+        let file2_path = setup_test_file(&file2_data);
+
+        black_box(run_util_function(
+            uumain,
+            &[
+                "-t",
+                ",",
+                file1_path.to_str().unwrap(),
+                file2_path.to_str().unwrap(),
+            ],
+        ));
+    });
+}
+
+/// Benchmark join with no order checking (unsorted data)
+#[divan::bench]
+fn join_nocheck_order(bencher: Bencher) {
+    bencher.bench(|| {
+        let num_lines = 10000;
+        let mut file1_data = Vec::new();
+        let mut file2_data = Vec::new();
+
+        // Generate unsorted test data with 50% overlap
+        for i in 0..num_lines {
+            let key = if i < num_lines / 2 {
+                format!("key_{:06}", i)
+            } else {
+                format!("unique1_{:06}", i - num_lines / 2)
+            };
+
+            file1_data.extend_from_slice(
+                format!("{}\tfile1_field1_{}\tfile1_field2_{}\n", key, i, i).as_bytes(),
+            );
+
+            let key2 = if i < num_lines / 2 {
+                format!("key_{:06}", i)
+            } else {
+                format!("unique2_{:06}", i - num_lines / 2)
+            };
+
+            file2_data.extend_from_slice(
+                format!("{}\tfile2_field1_{}\tfile2_field2_{}\n", key2, i, i).as_bytes(),
+            );
+        }
+
+        let file1_path = setup_test_file(&file1_data);
+        let file2_path = setup_test_file(&file2_data);
+
+        black_box(run_util_function(
+            uumain,
+            &[
+                "--nocheck-order",
+                file1_path.to_str().unwrap(),
+                file2_path.to_str().unwrap(),
+            ],
+        ));
+    });
+}
+
+fn main() {
+    divan::main();
+}