Merge pull request #97 from timClicks/add-word-benches

Manishearth · web-flow · commit 58d73acf0e92 · 2021-06-02T08:43:37.000-07:00
Benchmark other methods mentioned in README
diff --git a/Cargo.toml b/Cargo.toml
@@ -28,4 +28,12 @@ criterion = "0.3"
 
 [[bench]]
 name = "graphemes"
+harness = false
+
+[[bench]]
+name = "unicode_words"
+harness = false
+
+[[bench]]
+name = "word_bounds"
 harness = false
diff --git a/benches/unicode_words.rs b/benches/unicode_words.rs
@@ -0,0 +1,64 @@
+#[macro_use]
+extern crate bencher;
+extern crate unicode_segmentation;
+
+use bencher::Bencher;
+use unicode_segmentation::UnicodeSegmentation;
+use std::fs;
+
+fn unicode_words(bench: &mut Bencher, path: &str) {
+    let text = fs::read_to_string(path).unwrap();
+    bench.iter(|| {
+        for w in text.unicode_words() {
+            bencher::black_box(w);
+        }
+    });
+
+    bench.bytes = text.len() as u64;
+}
+
+fn unicode_words_arabic(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/arabic.txt");
+}
+
+fn unicode_words_english(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/english.txt");
+}
+
+fn unicode_words_hindi(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/hindi.txt");
+}
+
+fn unicode_words_japanese(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/japanese.txt");
+}
+
+fn unicode_words_korean(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/korean.txt");
+}
+
+fn unicode_words_mandarin(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/mandarin.txt");
+}
+
+fn unicode_words_russian(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/russian.txt");
+}
+
+fn unicode_words_source_code(bench: &mut Bencher) {
+    unicode_words(bench, "benches/texts/source_code.txt");
+}
+
+benchmark_group!(
+    benches,
+    unicode_words_arabic,
+    unicode_words_english,
+    unicode_words_hindi,
+    unicode_words_japanese,
+    unicode_words_korean,
+    unicode_words_mandarin,
+    unicode_words_russian,
+    unicode_words_source_code,
+);
+
+benchmark_main!(benches);
diff --git a/benches/word_bounds.rs b/benches/word_bounds.rs
@@ -0,0 +1,64 @@
+#[macro_use]
+extern crate bencher;
+extern crate unicode_segmentation;
+
+use bencher::Bencher;
+use unicode_segmentation::UnicodeSegmentation;
+use std::fs;
+
+fn word_bounds(bench: &mut Bencher, path: &str) {
+    let text = fs::read_to_string(path).unwrap();
+    bench.iter(|| {
+        for w in text.split_word_bounds() {
+            bencher::black_box(w);
+        }
+    });
+
+    bench.bytes = text.len() as u64;
+}
+
+fn word_bounds_arabic(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/arabic.txt");
+}
+
+fn word_bounds_english(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/english.txt");
+}
+
+fn word_bounds_hindi(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/hindi.txt");
+}
+
+fn word_bounds_japanese(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/japanese.txt");
+}
+
+fn word_bounds_korean(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/korean.txt");
+}
+
+fn word_bounds_mandarin(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/mandarin.txt");
+}
+
+fn word_bounds_russian(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/russian.txt");
+}
+
+fn word_bounds_source_code(bench: &mut Bencher) {
+    word_bounds(bench, "benches/texts/source_code.txt");
+}
+
+benchmark_group!(
+    benches,
+    word_bounds_arabic,
+    word_bounds_english,
+    word_bounds_hindi,
+    word_bounds_japanese,
+    word_bounds_korean,
+    word_bounds_mandarin,
+    word_bounds_russian,
+    word_bounds_source_code,
+);
+
+benchmark_main!(benches);
