percent_encoding/
ascii_set.rs

1// Copyright 2013-2016 The rust-url developers.
2//
3// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
4// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
5// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
6// option. This file may not be copied, modified, or distributed
7// except according to those terms.
8
9use core::{mem, ops};
10
11/// Represents a set of characters or bytes in the ASCII range.
12///
13/// This is used in [`percent_encode`] and [`utf8_percent_encode`].
14/// This is similar to [percent-encode sets](https://url.spec.whatwg.org/#percent-encoded-bytes).
15///
16/// Use the `add` method of an existing set to define a new set. For example:
17///
18/// [`percent_encode`]: crate::percent_encode
19/// [`utf8_percent_encode`]: crate::utf8_percent_encode
20///
21/// ```
22/// use percent_encoding::{AsciiSet, CONTROLS};
23///
24/// /// https://url.spec.whatwg.org/#fragment-percent-encode-set
25/// const FRAGMENT: &AsciiSet = &CONTROLS.add(b' ').add(b'"').add(b'<').add(b'>').add(b'`');
26/// ```
27#[derive(Debug, PartialEq, Eq)]
28pub struct AsciiSet {
29    mask: [Chunk; ASCII_RANGE_LEN / BITS_PER_CHUNK],
30}
31
32type Chunk = u32;
33
34const ASCII_RANGE_LEN: usize = 0x80;
35
36const BITS_PER_CHUNK: usize = 8 * mem::size_of::<Chunk>();
37
38impl AsciiSet {
39    /// An empty set.
40    pub const EMPTY: Self = Self {
41        mask: [0; ASCII_RANGE_LEN / BITS_PER_CHUNK],
42    };
43
44    /// Called with UTF-8 bytes rather than code points.
45    /// Not used for non-ASCII bytes.
46    pub(crate) const fn contains(&self, byte: u8) -> bool {
47        let chunk = self.mask[byte as usize / BITS_PER_CHUNK];
48        let mask = 1 << (byte as usize % BITS_PER_CHUNK);
49        (chunk & mask) != 0
50    }
51
52    pub(crate) fn should_percent_encode(&self, byte: u8) -> bool {
53        !byte.is_ascii() || self.contains(byte)
54    }
55
56    pub const fn add(&self, byte: u8) -> Self {
57        let mut mask = self.mask;
58        mask[byte as usize / BITS_PER_CHUNK] |= 1 << (byte as usize % BITS_PER_CHUNK);
59        Self { mask }
60    }
61
62    pub const fn remove(&self, byte: u8) -> Self {
63        let mut mask = self.mask;
64        mask[byte as usize / BITS_PER_CHUNK] &= !(1 << (byte as usize % BITS_PER_CHUNK));
65        Self { mask }
66    }
67
68    /// Return the union of two sets.
69    pub const fn union(&self, other: Self) -> Self {
70        let mask = [
71            self.mask[0] | other.mask[0],
72            self.mask[1] | other.mask[1],
73            self.mask[2] | other.mask[2],
74            self.mask[3] | other.mask[3],
75        ];
76        Self { mask }
77    }
78
79    /// Return the negation of the set.
80    pub const fn complement(&self) -> Self {
81        let mask = [!self.mask[0], !self.mask[1], !self.mask[2], !self.mask[3]];
82        Self { mask }
83    }
84}
85
86impl ops::Add for AsciiSet {
87    type Output = Self;
88
89    fn add(self, other: Self) -> Self {
90        self.union(other)
91    }
92}
93
94impl ops::Not for AsciiSet {
95    type Output = Self;
96
97    fn not(self) -> Self {
98        self.complement()
99    }
100}
101
102/// The set of 0x00 to 0x1F (C0 controls), and 0x7F (DEL).
103///
104/// Note that this includes the newline and tab characters, but not the space 0x20.
105///
106/// <https://url.spec.whatwg.org/#c0-control-percent-encode-set>
107pub const CONTROLS: &AsciiSet = &AsciiSet {
108    mask: [
109        !0_u32, // C0: 0x00 to 0x1F (32 bits set)
110        0,
111        0,
112        1 << (0x7F_u32 % 32), // DEL: 0x7F (one bit set)
113    ],
114};
115
116macro_rules! static_assert {
117    ($( $bool: expr, )+) => {
118        fn _static_assert() {
119            $(
120                let _ = mem::transmute::<[u8; $bool as usize], u8>;
121            )+
122        }
123    }
124}
125
126static_assert! {
127    CONTROLS.contains(0x00),
128    CONTROLS.contains(0x1F),
129    !CONTROLS.contains(0x20),
130    !CONTROLS.contains(0x7E),
131    CONTROLS.contains(0x7F),
132}
133
134/// Everything that is not an ASCII letter or digit.
135///
136/// This is probably more eager than necessary in any context.
137pub const NON_ALPHANUMERIC: &AsciiSet = &CONTROLS
138    .add(b' ')
139    .add(b'!')
140    .add(b'"')
141    .add(b'#')
142    .add(b'$')
143    .add(b'%')
144    .add(b'&')
145    .add(b'\'')
146    .add(b'(')
147    .add(b')')
148    .add(b'*')
149    .add(b'+')
150    .add(b',')
151    .add(b'-')
152    .add(b'.')
153    .add(b'/')
154    .add(b':')
155    .add(b';')
156    .add(b'<')
157    .add(b'=')
158    .add(b'>')
159    .add(b'?')
160    .add(b'@')
161    .add(b'[')
162    .add(b'\\')
163    .add(b']')
164    .add(b'^')
165    .add(b'_')
166    .add(b'`')
167    .add(b'{')
168    .add(b'|')
169    .add(b'}')
170    .add(b'~');
171
172#[cfg(test)]
173mod tests {
174    use super::*;
175
176    #[test]
177    fn add_op() {
178        let left = AsciiSet::EMPTY.add(b'A');
179        let right = AsciiSet::EMPTY.add(b'B');
180        let expected = AsciiSet::EMPTY.add(b'A').add(b'B');
181        assert_eq!(left + right, expected);
182    }
183
184    #[test]
185    fn not_op() {
186        let set = AsciiSet::EMPTY.add(b'A').add(b'B');
187        let not_set = !set;
188        assert!(!not_set.contains(b'A'));
189        assert!(not_set.contains(b'C'));
190    }
191
192    /// This test ensures that we can get the union of two sets as a constant value, which is
193    /// useful for defining sets in a modular way.
194    #[test]
195    fn union() {
196        const A: AsciiSet = AsciiSet::EMPTY.add(b'A');
197        const B: AsciiSet = AsciiSet::EMPTY.add(b'B');
198        const UNION: AsciiSet = A.union(B);
199        const EXPECTED: AsciiSet = AsciiSet::EMPTY.add(b'A').add(b'B');
200        assert_eq!(UNION, EXPECTED);
201    }
202
203    /// This test ensures that we can get the complement of a set as a constant value, which is
204    /// useful for defining sets in a modular way.
205    #[test]
206    fn complement() {
207        const BOTH: AsciiSet = AsciiSet::EMPTY.add(b'A').add(b'B');
208        const COMPLEMENT: AsciiSet = BOTH.complement();
209        assert!(!COMPLEMENT.contains(b'A'));
210        assert!(!COMPLEMENT.contains(b'B'));
211        assert!(COMPLEMENT.contains(b'C'));
212    }
213}
percent_encoding/ascii_set.rs

percent_encoding/
ascii_set.rs
