Add grapheme iteration benchmarks for various languages.

Cargo.toml

@@ -16,10 +16,15 @@ This crate provides Grapheme Cluster, Word and Sentence boundaries
 according to Unicode Standard Annex #29 rules.
 """
 
-exclude = [ "target/*", "Cargo.lock", "scripts/tmp", "*.txt" ]
+exclude = [ "target/*", "Cargo.lock", "scripts/tmp", "benches/texts/*", "*.txt", ]
 
 [features]
 no_std = [] # This is a no-op, preserved for backward compatibility only.
 
 [dev-dependencies]
 quickcheck = "0.7"
+bencher = "0.1"
+
+[[bench]]
+name = "graphemes"
+harness = false

benches/graphemes.rs

@@ -0,0 +1,109 @@
+#[macro_use]
+extern crate bencher;
+extern crate unicode_segmentation;
+
+use bencher::Bencher;
+use unicode_segmentation::UnicodeSegmentation;
+
+const TEXT_ARABIC: &str = include_str!("texts/arabic.txt");
+const TEXT_ENGLISH: &str = include_str!("texts/english.txt");
+const TEXT_HINDI: &str = include_str!("texts/hindi.txt");
+const TEXT_JAPANESE: &str = include_str!("texts/japanese.txt");
+const TEXT_KOREAN: &str = include_str!("texts/korean.txt");
+const TEXT_MANDARIN: &str = include_str!("texts/mandarin.txt");
+const TEXT_RUSSIAN: &str = include_str!("texts/russian.txt");
+const TEXT_SOURCE_CODE: &str = include_str!("texts/source_code.txt");
+
+fn graphemes_arabic(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_ARABIC, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_ARABIC.len() as u64;
+}
+
+fn graphemes_english(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_ENGLISH, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_ENGLISH.len() as u64;
+}
+
+fn graphemes_hindi(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_HINDI, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_HINDI.len() as u64;
+}
+
+fn graphemes_japanese(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_JAPANESE, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_JAPANESE.len() as u64;
+}
+
+fn graphemes_korean(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_KOREAN, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_KOREAN.len() as u64;
+}
+
+fn graphemes_mandarin(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_MANDARIN, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_MANDARIN.len() as u64;
+}
+
+fn graphemes_russian(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_RUSSIAN, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_RUSSIAN.len() as u64;
+}
+
+fn graphemes_source_code(bench: &mut Bencher) {
+    bench.iter(|| {
+        for g in UnicodeSegmentation::graphemes(TEXT_SOURCE_CODE, true) {
+            bencher::black_box(g);
+        }
+    });
+
+    bench.bytes = TEXT_SOURCE_CODE.len() as u64;
+}
+
+benchmark_group!(
+    benches,
+    graphemes_arabic,
+    graphemes_english,
+    graphemes_hindi,
+    graphemes_japanese,
+    graphemes_korean,
+    graphemes_mandarin,
+    graphemes_russian,
+    graphemes_source_code,
+);
+
+benchmark_main!(benches);