2 files changed, 195 insertions, 0 deletions
diff --git a/library/core/benches/str/char_count.rs b/library/core/benches/str/char_count.rs
new file mode 100644
index 000000000..25d9b2e29
--- /dev/null
+++ b/library/core/benches/str/char_count.rs
@@ -0,0 +1,107 @@
+use super::corpora::*;
+use test::{black_box, Bencher};
+
+macro_rules! define_benches {
+    ($( fn $name: ident($arg: ident: &str) $body: block )+) => {
+        define_benches!(mod en_tiny, en::TINY, $($name $arg $body)+);
+        define_benches!(mod en_small, en::SMALL, $($name $arg $body)+);
+        define_benches!(mod en_medium, en::MEDIUM, $($name $arg $body)+);
+        define_benches!(mod en_large, en::LARGE, $($name $arg $body)+);
+        define_benches!(mod en_huge, en::HUGE, $($name $arg $body)+);
+
+        define_benches!(mod zh_tiny, zh::TINY, $($name $arg $body)+);
+        define_benches!(mod zh_small, zh::SMALL, $($name $arg $body)+);
+        define_benches!(mod zh_medium, zh::MEDIUM, $($name $arg $body)+);
+        define_benches!(mod zh_large, zh::LARGE, $($name $arg $body)+);
+        define_benches!(mod zh_huge, zh::HUGE, $($name $arg $body)+);
+
+        define_benches!(mod ru_tiny, ru::TINY, $($name $arg $body)+);
+        define_benches!(mod ru_small, ru::SMALL, $($name $arg $body)+);
+        define_benches!(mod ru_medium, ru::MEDIUM, $($name $arg $body)+);
+        define_benches!(mod ru_large, ru::LARGE, $($name $arg $body)+);
+        define_benches!(mod ru_huge, ru::HUGE, $($name $arg $body)+);
+
+        define_benches!(mod emoji_tiny, emoji::TINY, $($name $arg $body)+);
+        define_benches!(mod emoji_small, emoji::SMALL, $($name $arg $body)+);
+        define_benches!(mod emoji_medium, emoji::MEDIUM, $($name $arg $body)+);
+        define_benches!(mod emoji_large, emoji::LARGE, $($name $arg $body)+);
+        define_benches!(mod emoji_huge, emoji::HUGE, $($name $arg $body)+);
+    };
+    (mod $mod_name: ident, $input: expr, $($name: ident $arg: ident $body: block)+) => {
+        mod $mod_name {
+            use super::*;
+            $(
+                #[bench]
+                fn $name(bencher: &mut Bencher) {
+                    let input = $input;
+                    bencher.bytes = input.len() as u64;
+                    let mut input_s = input.to_string();
+                    bencher.iter(|| {
+                        let $arg: &str = &black_box(&mut input_s);
+                        black_box($body)
+                    })
+                }
+            )+
+        }
+    };
+}
+
+define_benches! {
+    fn case00_libcore(s: &str) {
+        libcore(s)
+    }
+
+    fn case01_filter_count_cont_bytes(s: &str) {
+        filter_count_cont_bytes(s)
+    }
+
+    fn case02_iter_increment(s: &str) {
+        iterator_increment(s)
+    }
+
+    fn case03_manual_char_len(s: &str) {
+        manual_char_len(s)
+    }
+}
+
+fn libcore(s: &str) -> usize {
+    s.chars().count()
+}
+
+#[inline]
+fn utf8_is_cont_byte(byte: u8) -> bool {
+    (byte as i8) < -64
+}
+
+fn filter_count_cont_bytes(s: &str) -> usize {
+    s.as_bytes().iter().filter(|&&byte| !utf8_is_cont_byte(byte)).count()
+}
+
+fn iterator_increment(s: &str) -> usize {
+    let mut c = 0;
+    for _ in s.chars() {
+        c += 1;
+    }
+    c
+}
+
+fn manual_char_len(s: &str) -> usize {
+    let s = s.as_bytes();
+    let mut c = 0;
+    let mut i = 0;
+    let l = s.len();
+    while i < l {
+        let b = s[i];
+        if b < 0x80 {
+            i += 1;
+        } else if b < 0xe0 {
+            i += 2;
+        } else if b < 0xf0 {
+            i += 3;
+        } else {
+            i += 4;
+        }
+        c += 1;
+    }
+    c
+}
diff --git a/library/core/benches/str/corpora.rs b/library/core/benches/str/corpora.rs
new file mode 100644
index 000000000..b4ac62506
--- /dev/null
+++ b/library/core/benches/str/corpora.rs
@@ -0,0 +1,88 @@
+//! Exposes a number of modules with different kinds of strings.
+//!
+//! Each module contains `&str` constants named `TINY`, `SMALL`, `MEDIUM`,
+//! `LARGE`, and `HUGE`.
+//!
+//! - The `TINY` string is generally around 8 bytes.
+//! - The `SMALL` string is generally around 30-40 bytes.
+//! - The `MEDIUM` string is generally around 600-700 bytes.
+//! - The `LARGE` string is the `MEDIUM` string repeated 8x, and is around 5kb.
+//! - The `HUGE` string is the `LARGE` string repeated 8x (or the `MEDIUM`
+//!   string repeated 64x), and is around 40kb.
+//!
+//! Except for `mod emoji` (which is just a bunch of emoji), the strings were
+//! pulled from (localizations of) rust-lang.org.
+
+macro_rules! repeat8 {
+    ($s:expr) => {
+        concat!($s, $s, $s, $s, $s, $s, $s, $s)
+    };
+}
+
+macro_rules! define_consts {
+    ($s:literal) => {
+        pub const MEDIUM: &str = $s;
+        pub const LARGE: &str = repeat8!($s);
+        pub const HUGE: &str = repeat8!(repeat8!(repeat8!($s)));
+    };
+}
+
+pub mod en {
+    pub const TINY: &str = "Mary had";
+    pub const SMALL: &str = "Mary had a little lamb, Little lamb";
+    define_consts! {
+        "Rust is blazingly fast and memory-efficient: with no runtime or garbage
+         collector, it can power performance-critical services, run on embedded
+         devices, and easily integrate with other languages.  Rust’s rich type system
+         and ownership model guarantee memory-safety and thread-safety — enabling you
+         to eliminate many classes of bugs at compile-time.  Rust has great
+         documentation, a friendly compiler with useful error messages, and top-notch
+         tooling — an integrated package manager and build tool, smart multi-editor
+         support with auto-completion and type inspections, an auto-formatter, and
+         more."
+    }
+}
+
+pub mod zh {
+    pub const TINY: &str = "速度惊";
+    pub const SMALL: &str = "速度惊人且内存利用率极高";
+    define_consts! {
+        "Rust   速度惊人且内存利用率极高。由于\
+         没有运行时和垃圾回收，它能够胜任对性能要\
+         求特别高的服务，可以在嵌入式设备上运行，\
+         还能轻松和其他语言集成。Rust 丰富的类型\
+         系统和所有权模型保证了内存安全和线程安全，\
+         让您在编译期就能够消除各种各样的错误。\
+         Rust 拥有出色的文档、友好的编译器和清晰\
+         的错误提示信息， 还集成了一流的工具——\
+         包管理器和构建工具， 智能地自动补全和类\
+         型检验的多编辑器支持， 以及自动格式化代\
+         码等等。"
+    }
+}
+
+pub mod ru {
+    pub const TINY: &str = "Сотни";
+    pub const SMALL: &str = "Сотни компаний по";
+    define_consts! {
+        "Сотни компаний по всему миру используют Rust в реальных\
+         проектах для быстрых кросс-платформенных решений с\
+         ограниченными ресурсами. Такие проекты, как Firefox,\
+         Dropbox и Cloudflare, используют Rust. Rust отлично\
+         подходит как для стартапов, так и для больших компаний,\
+         как для встраиваемых устройств, так и для масштабируемых\
+         web-сервисов. Мой самый большой комплимент Rust."
+    }
+}
+
+pub mod emoji {
+    pub const TINY: &str = "😀😃";
+    pub const SMALL: &str = "😀😃😄😁😆😅🤣😂🙂🙃😉😊😇🥰😍🤩😘";
+    define_consts! {
+        "😀😃😄😁😆😅🤣😂🙂🙃😉😊😇🥰😍🤩😘😗☺😚😙🥲😋😛😜🤪😝🤑🤗🤭🤫🤔🤐🤨😐😑😶😶‍🌫️😏😒\
+         🙄😬😮‍💨🤥😌😔😪🤤😴😷🤒🤕🤢🤮🤧🥵🥶🥴😵😵‍💫🤯��🥳🥸😎🤓🧐😕😟🙁☹😮😯😲😳🥺😦😧😨\
+         😰😥😢😭😱😖😣😞😓😩😫🥱😤😡😠🤬😈👿💀☠💩🤡👹👺👻👽👾🤖😺😸😹😻😼😽🙀😿😾🙈🙉🙊\
+         💋💌💘💝💖💗💓��💕💟❣💔❤️‍🔥❤️‍🩹❤🧡💛💚💙💜🤎🖤🤍💯💢💥💫💦💨🕳💬👁️‍🗨️🗨🗯💭💤👋\
+         🤚🖐✋🖖👌🤌🤏✌"
+    }
+}