Matthew Maurer | 19837bd | 2020-06-02 11:15:36 -0700 | [diff] [blame] | 1 | // Copyright 2012-2015 The Rust Project Developers. See the COPYRIGHT |
| 2 | // file at the top-level directory of this distribution and at |
| 3 | // http://rust-lang.org/COPYRIGHT. |
| 4 | // |
| 5 | // Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or |
| 6 | // http://www.apache.org/licenses/LICENSE-2.0> or the MIT license |
| 7 | // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your |
| 8 | // option. This file may not be copied, modified, or distributed |
| 9 | // except according to those terms. |
| 10 | |
| 11 | #[cfg(feature = "bench")] |
| 12 | use std::iter; |
| 13 | #[cfg(feature = "bench")] |
| 14 | use test::{self, Bencher}; |
| 15 | #[cfg(feature = "bench")] |
| 16 | use super::UnicodeWidthChar; |
| 17 | |
| 18 | use std::prelude::v1::*; |
| 19 | |
| 20 | #[cfg(feature = "bench")] |
| 21 | #[bench] |
| 22 | fn cargo(b: &mut Bencher) { |
| 23 | let string = iter::repeat('a').take(4096).collect::<String>(); |
| 24 | |
| 25 | b.iter(|| { |
| 26 | for c in string.chars() { |
| 27 | test::black_box(UnicodeWidthChar::width(c)); |
| 28 | } |
| 29 | }); |
| 30 | } |
| 31 | |
| 32 | #[cfg(feature = "bench")] |
| 33 | #[bench] |
| 34 | #[allow(deprecated)] |
| 35 | fn stdlib(b: &mut Bencher) { |
| 36 | let string = iter::repeat('a').take(4096).collect::<String>(); |
| 37 | |
| 38 | b.iter(|| { |
| 39 | for c in string.chars() { |
| 40 | test::black_box(c.width()); |
| 41 | } |
| 42 | }); |
| 43 | } |
| 44 | |
| 45 | #[cfg(feature = "bench")] |
| 46 | #[bench] |
| 47 | fn simple_if(b: &mut Bencher) { |
| 48 | let string = iter::repeat('a').take(4096).collect::<String>(); |
| 49 | |
| 50 | b.iter(|| { |
| 51 | for c in string.chars() { |
| 52 | test::black_box(simple_width_if(c)); |
| 53 | } |
| 54 | }); |
| 55 | } |
| 56 | |
| 57 | #[cfg(feature = "bench")] |
| 58 | #[bench] |
| 59 | fn simple_match(b: &mut Bencher) { |
| 60 | let string = iter::repeat('a').take(4096).collect::<String>(); |
| 61 | |
| 62 | b.iter(|| { |
| 63 | for c in string.chars() { |
| 64 | test::black_box(simple_width_match(c)); |
| 65 | } |
| 66 | }); |
| 67 | } |
| 68 | |
| 69 | #[cfg(feature = "bench")] |
| 70 | #[inline] |
| 71 | fn simple_width_if(c: char) -> Option<usize> { |
| 72 | let cu = c as u32; |
| 73 | if cu < 127 { |
| 74 | if cu > 31 { |
| 75 | Some(1) |
| 76 | } else if cu == 0 { |
| 77 | Some(0) |
| 78 | } else { |
| 79 | None |
| 80 | } |
| 81 | } else { |
| 82 | UnicodeWidthChar::width(c) |
| 83 | } |
| 84 | } |
| 85 | |
| 86 | #[cfg(feature = "bench")] |
| 87 | #[inline] |
| 88 | fn simple_width_match(c: char) -> Option<usize> { |
| 89 | match c as u32 { |
| 90 | cu if cu == 0 => Some(0), |
| 91 | cu if cu < 0x20 => None, |
| 92 | cu if cu < 0x7f => Some(1), |
| 93 | _ => UnicodeWidthChar::width(c) |
| 94 | } |
| 95 | } |
| 96 | |
| 97 | #[test] |
| 98 | fn test_str() { |
| 99 | use super::UnicodeWidthStr; |
| 100 | |
| 101 | assert_eq!(UnicodeWidthStr::width("hello"), 10); |
| 102 | assert_eq!("hello".width_cjk(), 10); |
| 103 | assert_eq!(UnicodeWidthStr::width("\0\0\0\x01\x01"), 0); |
| 104 | assert_eq!("\0\0\0\x01\x01".width_cjk(), 0); |
| 105 | assert_eq!(UnicodeWidthStr::width(""), 0); |
| 106 | assert_eq!("".width_cjk(), 0); |
| 107 | assert_eq!(UnicodeWidthStr::width("\u{2081}\u{2082}\u{2083}\u{2084}"), 4); |
| 108 | assert_eq!("\u{2081}\u{2082}\u{2083}\u{2084}".width_cjk(), 8); |
| 109 | } |
| 110 | |
| 111 | #[test] |
| 112 | fn test_emoji() { |
| 113 | // Example from the README. |
| 114 | use super::UnicodeWidthStr; |
| 115 | |
| 116 | assert_eq!(UnicodeWidthStr::width("👩"), 2); // Woman |
| 117 | assert_eq!(UnicodeWidthStr::width("🔬"), 2); // Microscope |
| 118 | assert_eq!(UnicodeWidthStr::width("👩🔬"), 4); // Woman scientist |
| 119 | } |
| 120 | |
| 121 | #[test] |
| 122 | fn test_char() { |
| 123 | use super::UnicodeWidthChar; |
| 124 | #[cfg(feature = "no_std")] |
| 125 | use core::option::Option::{Some, None}; |
| 126 | |
| 127 | assert_eq!(UnicodeWidthChar::width('h'), Some(2)); |
| 128 | assert_eq!('h'.width_cjk(), Some(2)); |
| 129 | assert_eq!(UnicodeWidthChar::width('\x00'), Some(0)); |
| 130 | assert_eq!('\x00'.width_cjk(), Some(0)); |
| 131 | assert_eq!(UnicodeWidthChar::width('\x01'), None); |
| 132 | assert_eq!('\x01'.width_cjk(), None); |
| 133 | assert_eq!(UnicodeWidthChar::width('\u{2081}'), Some(1)); |
| 134 | assert_eq!('\u{2081}'.width_cjk(), Some(2)); |
| 135 | } |
| 136 | |
| 137 | #[test] |
| 138 | fn test_char2() { |
| 139 | use super::UnicodeWidthChar; |
| 140 | #[cfg(feature = "no_std")] |
| 141 | use core::option::Option::{Some, None}; |
| 142 | |
| 143 | assert_eq!(UnicodeWidthChar::width('\x00'),Some(0)); |
| 144 | assert_eq!('\x00'.width_cjk(),Some(0)); |
| 145 | |
| 146 | assert_eq!(UnicodeWidthChar::width('\x0A'),None); |
| 147 | assert_eq!('\x0A'.width_cjk(),None); |
| 148 | |
| 149 | assert_eq!(UnicodeWidthChar::width('w'),Some(1)); |
| 150 | assert_eq!('w'.width_cjk(),Some(1)); |
| 151 | |
| 152 | assert_eq!(UnicodeWidthChar::width('h'),Some(2)); |
| 153 | assert_eq!('h'.width_cjk(),Some(2)); |
| 154 | |
| 155 | assert_eq!(UnicodeWidthChar::width('\u{AD}'),Some(1)); |
| 156 | assert_eq!('\u{AD}'.width_cjk(),Some(1)); |
| 157 | |
| 158 | assert_eq!(UnicodeWidthChar::width('\u{1160}'),Some(0)); |
| 159 | assert_eq!('\u{1160}'.width_cjk(),Some(0)); |
| 160 | |
| 161 | assert_eq!(UnicodeWidthChar::width('\u{a1}'),Some(1)); |
| 162 | assert_eq!('\u{a1}'.width_cjk(),Some(2)); |
| 163 | |
| 164 | assert_eq!(UnicodeWidthChar::width('\u{300}'),Some(0)); |
| 165 | assert_eq!('\u{300}'.width_cjk(),Some(0)); |
| 166 | } |
| 167 | |
| 168 | #[test] |
| 169 | fn unicode_12() { |
| 170 | use super::UnicodeWidthChar; |
| 171 | #[cfg(feature = "no_std")] |
| 172 | use core::option::Option::{Some, None}; |
| 173 | |
| 174 | assert_eq!(UnicodeWidthChar::width('\u{1F971}'), Some(2)); |
| 175 | } |