1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
|
//! Utilities for working with (unicode) characters/codepoints
use std::fmt::{self, Debug, Display};
#[cfg(feature = "unicode-casefold")]
use crate::chars::case_fold::CASE_FOLDING_SIMPLE;
use crate::Config;
//autogenerated by generate-ucd
#[expect(warnings)]
#[rustfmt::skip]
#[cfg(feature = "unicode-casefold")]
mod case_fold;
#[cfg(feature = "unicode-normalization")]
mod normalize;
pub(crate) trait Char: Copy + Eq + Ord + fmt::Display {
const ASCII: bool;
fn char_class(self, config: &Config) -> CharClass;
fn char_class_and_normalize(self, config: &Config) -> (Self, CharClass);
fn normalize(self, config: &Config) -> Self;
}
/// repr tansparent wrapper around u8 with better formatting and `PartialEq<char>` implementation
#[repr(transparent)]
#[derive(PartialEq, Eq, PartialOrd, Ord, Clone, Copy)]
pub(crate) struct AsciiChar(pub u8);
impl AsciiChar {
pub fn cast(bytes: &[u8]) -> &[AsciiChar] {
unsafe { &*(bytes as *const [u8] as *const [AsciiChar]) }
}
}
impl fmt::Display for AsciiChar {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
Display::fmt(&(self.0 as char), f)
}
}
impl PartialEq<AsciiChar> for char {
fn eq(&self, other: &AsciiChar) -> bool {
other.0 as char == *self
}
}
impl Char for AsciiChar {
const ASCII: bool = true;
#[inline]
fn char_class(self, config: &Config) -> CharClass {
let c = self.0;
// using manual if conditions instead optimizes better
if c >= b'a' && c <= b'z' {
CharClass::Lower
} else if c >= b'A' && c <= b'Z' {
CharClass::Upper
} else if c >= b'0' && c <= b'9' {
CharClass::Number
} else if c.is_ascii_whitespace() {
CharClass::Whitespace
} else if config.delimiter_chars.contains(&c) {
CharClass::Delimiter
} else {
CharClass::NonWord
}
}
#[inline(always)]
fn char_class_and_normalize(mut self, config: &Config) -> (Self, CharClass) {
let char_class = self.char_class(config);
if config.ignore_case && char_class == CharClass::Upper {
self.0 += 32
}
(self, char_class)
}
#[inline(always)]
fn normalize(mut self, config: &Config) -> Self {
if config.ignore_case && self.0 >= b'A' && self.0 <= b'Z' {
self.0 += 32
}
self
}
}
fn char_class_non_ascii(c: char) -> CharClass {
if c.is_lowercase() {
CharClass::Lower
} else if is_upper_case(c) {
CharClass::Upper
} else if c.is_numeric() {
CharClass::Number
} else if c.is_alphabetic() {
CharClass::Letter
} else if c.is_whitespace() {
CharClass::Whitespace
} else {
CharClass::NonWord
}
}
impl Char for char {
const ASCII: bool = false;
#[inline(always)]
fn char_class(self, config: &Config) -> CharClass {
if self.is_ascii() {
return AsciiChar(self as u8).char_class(config);
}
char_class_non_ascii(self)
}
#[inline(always)]
fn char_class_and_normalize(mut self, config: &Config) -> (Self, CharClass) {
if self.is_ascii() {
let (c, class) = AsciiChar(self as u8).char_class_and_normalize(config);
return (c.0 as char, class);
}
let char_class = char_class_non_ascii(self);
#[cfg(feature = "unicode-casefold")]
let mut case_fold = char_class == CharClass::Upper;
#[cfg(feature = "unicode-normalization")]
if config.normalize {
self = normalize::normalize(self);
case_fold = true
}
#[cfg(feature = "unicode-casefold")]
if case_fold && config.ignore_case {
self = CASE_FOLDING_SIMPLE
.binary_search_by_key(&self, |(upper, _)| *upper)
.map_or(self, |idx| CASE_FOLDING_SIMPLE[idx].1)
}
(self, char_class)
}
#[inline(always)]
fn normalize(mut self, config: &Config) -> Self {
#[cfg(feature = "unicode-normalization")]
if config.normalize {
self = normalize::normalize(self);
}
#[cfg(feature = "unicode-casefold")]
if config.ignore_case {
self = to_lower_case(self)
}
self
}
}
#[cfg(feature = "unicode-normalization")]
pub use normalize::normalize;
#[cfg(feature = "unicode-segmentation")]
use unicode_segmentation::UnicodeSegmentation;
/// Converts a character to lower case using simple unicode case folding
#[cfg(feature = "unicode-casefold")]
#[inline(always)]
pub fn to_lower_case(c: char) -> char {
CASE_FOLDING_SIMPLE
.binary_search_by_key(&c, |(upper, _)| *upper)
.map_or(c, |idx| CASE_FOLDING_SIMPLE[idx].1)
}
/// Checks if a character is upper case according to simple unicode case folding.
/// if the `unicode-casefold` feature is disable the equivalent std function is used
#[inline(always)]
pub fn is_upper_case(c: char) -> bool {
#[cfg(feature = "unicode-casefold")]
let val = CASE_FOLDING_SIMPLE
.binary_search_by_key(&c, |(upper, _)| *upper)
.is_ok();
#[cfg(not(feature = "unicode-casefold"))]
let val = c.is_uppercase();
val
}
#[derive(Debug, Eq, PartialEq, PartialOrd, Ord, Copy, Clone, Hash)]
pub(crate) enum CharClass {
Whitespace,
NonWord,
Delimiter,
Lower,
Upper,
Letter,
Number,
}
/// Nucleo cannot match graphemes as single units. To work around
/// that we only use the first codepoint of each grapheme. This
/// iterator returns the first character of each unicode grapheme
/// in a string and is used for constructing `Utf32Str(ing)`.
pub fn graphemes(text: &str) -> impl Iterator<Item = char> + '_ {
#[cfg(feature = "unicode-segmentation")]
let res = text.graphemes(true).map(|grapheme| {
// we need to special-case this check since `\r\n` is a single grapheme and is
// therefore the exception to the rule that normalization of a grapheme should
// map to the first character.
if grapheme == "\r\n" {
'\n'
} else {
grapheme
.chars()
.next()
.expect("graphemes must be non-empty")
}
});
#[cfg(not(feature = "unicode-segmentation"))]
let res = text.chars();
res
}
|
