-
Notifications
You must be signed in to change notification settings - Fork 13.5k
Optimize core::str::Chars::count
#90414
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Merged
+379
−29
Merged
Changes from all commits
Commits
Show all changes
5 commits
Select commit
Hold shift + click to select a range
628b217
Optimize `core::str::Chars::count`
thomcc ed01324
Fix zh::SMALL string in core::str benchmarks
thomcc 002aaf2
Ensure non-power-of-two sizes are tested in the Chars::count test
thomcc ebbccaf
Respond to review feedback, and improve implementation somewhat
thomcc 41f8214
Fix comment grammar for `do_count_chars`
thomcc File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,33 +1,10 @@ | ||
| use std::str; | ||
| use test::{black_box, Bencher}; | ||
|
|
||
| const LOREM_SHORT: &str = "Lorem ipsum"; | ||
|
|
||
| const LOREM: &str = "Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. | ||
| Duis autem vel eum iriure dolor in hendrerit in vulputate velit esse molestie consequat, vel illum dolore eu feugiat nulla facilisis at vero eros et accumsan et iusto odio dignissim qui blandit praesent luptatum zzril delenit augue duis dolore te feugait nulla facilisi. Lorem ipsum dolor sit amet, consectetuer adipiscing elit, sed diam nonummy nibh euismod tincidunt ut laoreet dolore magna aliquam erat volutpat. | ||
| Ut wisi enim ad minim veniam, quis nostrud exerci tation ullamcorper suscipit lobortis nisl ut aliquip ex ea commodo consequat. Duis autem vel eum iriure dolor in hendrerit in vulputate velit esse molestie consequat, vel illum dolore eu feugiat nulla facilisis at vero eros et accumsan et iusto odio dignissim qui blandit praesent luptatum zzril delenit augue duis dolore te feugait nulla facilisi. | ||
| Nam liber tempor cum soluta nobis eleifend option congue nihil imperdiet doming id quod mazim placerat facer possim assum. Lorem ipsum dolor sit amet, consectetuer adipiscing elit, sed diam nonummy nibh euismod tincidunt ut laoreet dolore magna aliquam erat volutpat. Ut wisi enim ad minim veniam, quis nostrud exerci tation ullamcorper suscipit lobortis nisl ut aliquip ex ea commodo consequat. | ||
| Duis autem vel eum iriure dolor in hendrerit in vulputate velit esse molestie consequat, vel illum dolore eu feugiat nulla facilisis. | ||
| At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, At accusam aliquyam diam diam dolore dolores duo eirmod eos erat, et nonumy sed tempor et et invidunt justo labore Stet clita ea et gubergren, kasd magna no rebum. sanctus sea sed takimata ut vero voluptua. est Lorem ipsum dolor sit amet. Lorem ipsum dolor sit amet, consetetur"; | ||
|
|
||
| const EMOJI: &str = "😀😃😄😁😆😅🤣😂🙂🙃😉😊😇🥰😍🤩😘😗☺😚😙🥲😋😛😜🤪😝🤑🤗🤭🤫🤔🤐🤨😐😑😶😶🌫️😏😒🙄😬😮💨🤥😌😔😪🤤😴😷🤒🤕🤢🤮🤧🥵🥶🥴😵😵💫🤯🤠🥳🥸😎🤓🧐😕😟🙁☹😮😯😲😳🥺😦😧😨😰😥😢😭😱😖😣😞😓😩😫🥱😤😡😠🤬😈👿💀☠💩🤡👹👺👻👽👾🤖😺😸😹😻😼😽🙀😿😾🙈🙉🙊💋💌💘💝💖💗💓💞💕💟❣💔❤️🔥❤️🩹❤🧡💛💚💙💜🤎🖤🤍💯💢💥💫💦💨🕳💣💬👁️🗨️🗨🗯💭💤👋🤚🖐✋🖖👌🤌🤏✌🤞🤟🤘🤙👈👉👆🖕👇☝👍👎✊👊🤛🤜👏🙌👐🤲🤝🙏✍💅🤳💪🦾🦿🦵🦶👂🦻👃🧠🫀🫁🦷🦴👀👁👅👄👶🧒👦👧🧑👱👨🧔🧔♂️🧔♀️👨🦰👨🦱👨🦳👨🦲👩👩🦰🧑🦰👩🦱🧑🦱👩🦳🧑🦳👩🦲🧑🦲👱♀️👱♂️🧓👴👵🙍🙍♂️🙍♀️🙎🙎♂️🙎♀️🙅🙅♂️🙅♀️🙆🙆♂️🙆♀️💁💁♂️💁♀️🙋🙋♂️🙋♀️🧏🧏♂️🧏♀️🙇🙇♂️🙇♀️🤦🤦♂️🤦♀️🤷🤷♂️🤷♀️🧑⚕️👨⚕️👩⚕️🧑🎓👨🎓👩🎓🧑🏫👨🏫👩🏫🧑⚖️👨⚖️👩⚖️🧑🌾👨🌾👩🌾🧑🍳👨🍳👩🍳🧑🔧👨🔧👩🔧🧑🏭👨🏭👩🏭🧑💼👨💼👩💼🧑🔬👨🔬👩🔬🧑💻👨💻👩💻🧑🎤👨🎤👩🎤🧑🎨👨🎨👩🎨🧑✈️👨✈️👩✈️🧑🚀👨🚀👩🚀🧑🚒👨🚒👩🚒👮👮♂️👮♀️🕵🕵️♂️🕵️♀️💂💂♂️💂♀️🥷👷👷♂️👷♀️🤴👸👳👳♂️👳♀️👲🧕🤵🤵♂️🤵♀️👰👰♂️👰♀️🤰🤱👩🍼👨🍼🧑🍼👼🎅🤶🧑🎄🦸🦸♂️🦸♀️🦹🦹♂️🦹♀️🧙🧙♂️🧙♀️🧚🧚♂️🧚♀️🧛🧛♂️🧛♀️🧜🧜♂️🧜♀️🧝🧝♂️🧝♀️🧞🧞♂️🧞♀️🧟🧟♂️🧟♀️💆💆♂️💆♀️💇💇♂️💇♀️🚶🚶♂️🚶♀️🧍🧍♂️🧍♀️🧎🧎♂️🧎♀️🧑🦯👨🦯👩🦯🧑🦼👨🦼👩🦼🧑🦽👨🦽👩🦽🏃🏃♂️🏃♀️💃🕺🕴👯👯♂️👯♀️🧖🧖♂️🧖♀️🧗🧗♂️🧗♀️🤺🏇⛷🏂🏌🏌️♂️🏌️♀️🏄🏄♂️🏄♀️🚣🚣♂️🚣♀️🏊🏊♂️🏊♀️⛹⛹️♂️⛹️♀️🏋🏋️♂️🏋️♀️🚴🚴♂️🚴♀️🚵🚵♂️🚵♀️🤸🤸♂️🤸♀️🤼🤼♂️🤼♀️🤽🤽♂️🤽♀️🤾🤾♂️🤾♀️🤹🤹♂️🤹♀️🧘🧘♂️🧘♀️🛀🛌🧑🤝🧑👭👫👬💏👩❤️💋👨👨❤️💋👨👩❤️💋👩💑👩❤️👨👨❤️👨👩❤️👩👪👨👩👦👨👩👧👨👩👧👦👨👩👦👦👨👩👧👧👨👨👦👨👨👧👨👨👧👦👨👨👦👦👨👨👧👧👩👩👦👩👩👧👩👩👧👦👩👩👦👦👩👩👧👧👨👦👨👦👦👨👧👨👧👦👨👧👧👩👦👩👦👦👩👧👩👧👦👩👧👧🗣👤👥🫂👣🦰🦱🦳🦲🐵🐒🦍🦧🐶🐕🦮🐕🦺🐩🐺🦊🦝🐱🐈🐈⬛🦁🐯🐅🐆🐴🐎🦄🦓🦌🦬🐮🐂🐃🐄🐷🐖🐗🐽🐏🐑🐐🐪🐫🦙🦒🐘🦣🦏🦛🐭🐁🐀🐹🐰🐇🐿🦫🦔🦇🐻🐻❄️🐨🐼🦥🦦🦨🦘🦡🐾🦃🐔🐓🐣🐤🐥🐦🐧🕊🦅🦆🦢🦉🦤🪶🦩🦚🦜🐸🐊🐢🦎🐍🐲🐉🦕🦖🐳🐋🐬🦭🐟🐠🐡🦈🐙🐚🐌🦋🐛🐜🐝🪲🐞🦗🪳🕷🕸🦂🦟🪰🪱🦠💐🌸💮🏵🌹🥀🌺🌻🌼🌷🌱🪴🌲🌳🌴🌵🌾🌿☘🍀🍁🍂🍃🍇🍈🍉🍊🍋🍌🍍🥭🍎🍏🍐🍑🍒🍓🫐🥝🍅🫒🥥🥑🍆🥔🥕🌽🌶🫑🥒🥬🥦🧄🧅🍄🥜🌰🍞🥐🥖🫓🥨🥯🥞🧇🧀🍖🍗🥩🥓🍔🍟🍕🌭🥪🌮🌯🫔🥙🧆🥚🍳🥘🍲🫕🥣🥗🍿🧈🧂🥫🍱🍘🍙🍚🍛🍜🍝🍠🍢🍣🍤🍥🥮🍡🥟🥠🥡🦀🦞🦐🦑🦪🍦🍧🍨🍩🍪🎂🍰🧁🥧🍫🍬🍭🍮🍯🍼🥛☕🫖🍵🍶🍾🍷🍸🍹🍺🍻🥂🥃🥤🧋🧃🧉🧊🥢🍽🍴🥄🔪🏺🌍🌎🌏🌐🗺🗾🧭🏔⛰🌋🗻🏕🏖🏜🏝🏞🏟🏛🏗🧱🪨🪵🛖🏘🏚🏠🏡🏢🏣🏤🏥🏦🏨🏩🏪🏫🏬🏭🏯🏰💒🗼🗽⛪🕌🛕🕍⛩🕋⛲⛺🌁🌃🏙🌄🌅🌆🌇🌉♨🎠🎡🎢💈🎪🚂🚃🚄🚅🚆🚇🚈🚉🚊🚝🚞🚋🚌🚍🚎🚐🚑🚒🚓🚔🚕🚖🚗🚘🚙🛻🚚🚛🚜🏎🏍🛵🦽🦼🛺🚲🛴🛹🛼🚏🛣🛤🛢⛽🚨🚥🚦🛑🚧⚓⛵🛶🚤🛳⛴🛥🚢✈🛩🛫🛬🪂💺🚁🚟🚠🚡🛰🚀🛸🛎🧳⌛⏳⌚⏰⏱⏲🕰🕛🕧🕐🕜🕑🕝🕒🕞🕓🕟🕔🕠🕕🕡🕖🕢🕗🕣🕘🕤🕙🕥🕚🕦🌑🌒🌓🌔🌕🌖🌗🌘🌙🌚🌛🌜🌡☀🌝🌞🪐⭐🌟🌠🌌☁⛅⛈🌤🌥🌦🌧🌨🌩🌪🌫🌬🌀🌈🌂☂☔⛱⚡❄☃⛄☄🔥💧🌊🎃🎄🎆🎇🧨✨🎈🎉🎊🎋🎍🎎🎏🎐🎑🧧🎀🎁🎗🎟🎫🎖🏆🏅🥇🥈🥉⚽⚾🥎🏀🏐🏈🏉🎾🥏🎳🏏🏑🏒🥍🏓🏸🥊🥋🥅⛳⛸🎣🤿🎽🎿🛷🥌🎯🪀🪁🎱🔮🪄🧿🎮🕹🎰🎲🧩🧸🪅🪆♠♥♦♣♟🃏🀄🎴🎭🖼🎨🧵🪡🧶🪢👓🕶🥽🥼🦺👔👕👖🧣🧤🧥🧦👗👘🥻🩱🩲🩳👙👚👛👜👝🛍🎒🩴👞👟🥾🥿👠👡🩰👢👑👒🎩🎓🧢🪖⛑📿💄💍💎🔇🔈🔉🔊📢📣📯🔔🔕🎼🎵🎶🎙🎚🎛🎤🎧📻🎷🪗🎸🎹🎺🎻🪕🥁"; | ||
|
|
||
| #[bench] | ||
| fn str_char_count_lorem(b: &mut Bencher) { | ||
| b.iter(|| black_box(LOREM).chars().count()); | ||
| } | ||
|
|
||
| #[bench] | ||
| fn str_char_count_lorem_short(b: &mut Bencher) { | ||
| b.iter(|| black_box(LOREM_SHORT).chars().count()); | ||
| } | ||
|
|
||
| #[bench] | ||
| fn str_char_count_emoji(b: &mut Bencher) { | ||
| b.iter(|| black_box(EMOJI).chars().count()); | ||
| } | ||
| mod char_count; | ||
| mod corpora; | ||
|
|
||
| #[bench] | ||
| fn str_validate_emoji(b: &mut Bencher) { | ||
| b.iter(|| str::from_utf8(black_box(EMOJI.as_bytes()))); | ||
| b.iter(|| str::from_utf8(black_box(corpora::emoji::LARGE.as_bytes()))); | ||
| } |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -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 | ||
| } |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -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"; | ||
thomcc marked this conversation as resolved.
Show resolved
Hide resolved
|
||
| 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! { | ||
| "😀😃😄😁😆😅🤣😂🙂🙃😉😊😇🥰😍🤩😘😗☺😚😙🥲😋😛😜🤪😝🤑🤗🤭🤫🤔🤐🤨😐😑😶😶🌫️😏😒\ | ||
| 🙄😬😮💨🤥😌😔😪🤤😴😷🤒🤕🤢🤮🤧🥵🥶🥴😵😵💫🤯��🥳🥸😎🤓🧐😕😟🙁☹😮😯😲😳🥺😦😧😨\ | ||
| 😰😥😢😭😱😖😣😞😓😩😫🥱😤😡😠🤬😈👿💀☠💩🤡👹👺👻👽👾🤖😺😸😹😻😼😽🙀😿😾🙈🙉🙊\ | ||
| 💋💌💘💝💖💗💓��💕💟❣💔❤️🔥❤️🩹❤🧡💛💚💙💜🤎🖤🤍💯💢💥💫💦💨🕳💬👁️🗨️🗨🗯💭💤👋\ | ||
| 🤚🖐✋🖖👌🤌🤏✌" | ||
| } | ||
| } | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,136 @@ | ||
| //! Code for efficiently counting the number of `char`s in a UTF-8 encoded | ||
| //! string. | ||
| //! | ||
| //! Broadly, UTF-8 encodes `char`s as a "leading" byte which begins the `char`, | ||
| //! followed by some number (possibly 0) of continuation bytes. | ||
| //! | ||
| //! The leading byte can have a number of bit-patterns (with the specific | ||
| //! pattern indicating how many continuation bytes follow), but the continuation | ||
| //! bytes are always in the format `0b10XX_XXXX` (where the `X`s can take any | ||
| //! value). That is, the most significant bit is set, and the second most | ||
| //! significant bit is unset. | ||
| //! | ||
| //! To count the number of characters, we can just count the number of bytes in | ||
| //! the string which are not continuation bytes, which can be done many bytes at | ||
| //! a time fairly easily. | ||
| //! | ||
| //! Note: Because the term "leading byte" can sometimes be ambiguous (for | ||
| //! example, it could also refer to the first byte of a slice), we'll often use | ||
| //! the term "non-continuation byte" to refer to these bytes in the code. | ||
| use core::intrinsics::unlikely; | ||
|
|
||
| const USIZE_SIZE: usize = core::mem::size_of::<usize>(); | ||
| const UNROLL_INNER: usize = 4; | ||
|
|
||
| #[inline] | ||
| pub(super) fn count_chars(s: &str) -> usize { | ||
| if s.len() < USIZE_SIZE * UNROLL_INNER { | ||
| // Avoid entering the optimized implementation for strings where the | ||
| // difference is not likely to matter, or where it might even be slower. | ||
| // That said, a ton of thought was not spent on the particular threshold | ||
| // here, beyond "this value seems to make sense". | ||
| char_count_general_case(s.as_bytes()) | ||
| } else { | ||
| do_count_chars(s) | ||
| } | ||
| } | ||
|
|
||
| fn do_count_chars(s: &str) -> usize { | ||
| // For correctness, `CHUNK_SIZE` must be: | ||
| // | ||
| // - Less than or equal to 255, otherwise we'll overflow bytes in `counts`. | ||
| // - A multiple of `UNROLL_INNER`, otherwise our `break` inside the | ||
| // `body.chunks(CHUNK_SIZE)` loop is incorrect. | ||
| // | ||
| // For performance, `CHUNK_SIZE` should be: | ||
| // - Relatively cheap to `/` against (so some simple sum of powers of two). | ||
| // - Large enough to avoid paying for the cost of the `sum_bytes_in_usize` | ||
| // too often. | ||
| const CHUNK_SIZE: usize = 192; | ||
|
|
||
| // Check the properties of `CHUNK_SIZE` and `UNROLL_INNER` that are required | ||
| // for correctness. | ||
| const _: () = assert!(CHUNK_SIZE < 256); | ||
| const _: () = assert!(CHUNK_SIZE % UNROLL_INNER == 0); | ||
|
|
||
| // SAFETY: transmuting `[u8]` to `[usize]` is safe except for size | ||
| // differences which are handled by `align_to`. | ||
| let (head, body, tail) = unsafe { s.as_bytes().align_to::<usize>() }; | ||
|
|
||
| // This should be quite rare, and basically exists to handle the degenerate | ||
| // cases where align_to fails (as well as miri under symbolic alignment | ||
| // mode). | ||
| // | ||
| // The `unlikely` helps discourage LLVM from inlining the body, which is | ||
| // nice, as we would rather not mark the `char_count_general_case` function | ||
| // as cold. | ||
| if unlikely(body.is_empty() || head.len() > USIZE_SIZE || tail.len() > USIZE_SIZE) { | ||
| return char_count_general_case(s.as_bytes()); | ||
| } | ||
|
|
||
| let mut total = char_count_general_case(head) + char_count_general_case(tail); | ||
| // Split `body` into `CHUNK_SIZE` chunks to reduce the frequency with which | ||
| // we call `sum_bytes_in_usize`. | ||
| for chunk in body.chunks(CHUNK_SIZE) { | ||
| // We accumulate intermediate sums in `counts`, where each byte contains | ||
| // a subset of the sum of this chunk, like a `[u8; size_of::<usize>()]`. | ||
| let mut counts = 0; | ||
|
|
||
| let (unrolled_chunks, remainder) = chunk.as_chunks::<UNROLL_INNER>(); | ||
| for unrolled in unrolled_chunks { | ||
| for &word in unrolled { | ||
| // Because `CHUNK_SIZE` is < 256, this addition can't cause the | ||
| // count in any of the bytes to overflow into a subsequent byte. | ||
| counts += contains_non_continuation_byte(word); | ||
| } | ||
| } | ||
|
|
||
| // Sum the values in `counts` (which, again, is conceptually a `[u8; | ||
| // size_of::<usize>()]`), and accumulate the result into `total`. | ||
| total += sum_bytes_in_usize(counts); | ||
|
|
||
| // If there's any data in `remainder`, then handle it. This will only | ||
| // happen for the last `chunk` in `body.chunks()` (because `CHUNK_SIZE` | ||
| // is divisible by `UNROLL_INNER`), so we explicitly break at the end | ||
| // (which seems to help LLVM out). | ||
| if !remainder.is_empty() { | ||
| // Accumulate all the data in the remainder. | ||
| let mut counts = 0; | ||
| for &word in remainder { | ||
| counts += contains_non_continuation_byte(word); | ||
| } | ||
| total += sum_bytes_in_usize(counts); | ||
| break; | ||
| } | ||
| } | ||
| total | ||
| } | ||
|
|
||
| // Checks each byte of `w` to see if it contains the first byte in a UTF-8 | ||
| // sequence. Bytes in `w` which are continuation bytes are left as `0x00` (e.g. | ||
| // false), and bytes which are non-continuation bytes are left as `0x01` (e.g. | ||
| // true) | ||
| #[inline] | ||
| fn contains_non_continuation_byte(w: usize) -> usize { | ||
| const LSB: usize = 0x0101_0101_0101_0101u64 as usize; | ||
| ((!w >> 7) | (w >> 6)) & LSB | ||
| } | ||
|
|
||
| // Morally equivalent to `values.to_ne_bytes().into_iter().sum::<usize>()`, but | ||
| // more efficient. | ||
| #[inline] | ||
| fn sum_bytes_in_usize(values: usize) -> usize { | ||
| const LSB_SHORTS: usize = 0x0001_0001_0001_0001_u64 as usize; | ||
| const SKIP_BYTES: usize = 0x00ff_00ff_00ff_00ff_u64 as usize; | ||
|
|
||
| let pair_sum: usize = (values & SKIP_BYTES) + ((values >> 8) & SKIP_BYTES); | ||
| pair_sum.wrapping_mul(LSB_SHORTS) >> ((USIZE_SIZE - 2) * 8) | ||
| } | ||
|
|
||
| // This is the most direct implementation of the concept of "count the number of | ||
| // bytes in the string which are not continuation bytes", and is used for the | ||
| // head and tail of the input string (the first and last item in the tuple | ||
| // returned by `slice::align_to`). | ||
| fn char_count_general_case(s: &[u8]) -> usize { | ||
| s.iter().filter(|&&byte| !super::validations::utf8_is_cont_byte(byte)).count() | ||
| } |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Uh oh!
There was an error while loading. Please reload this page.