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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228
mod error; use core::str::from_utf8; use std::borrow::Cow; pub use crate::error::DecodingError; pub fn decode(bytes: &[u8]) -> Result<Cow<str>, DecodingError> { if let Ok(str) = from_utf8(bytes) { return Ok(Cow::Borrowed(str)); } let mut decoded = Vec::with_capacity(bytes.len()); let mut iter = bytes.iter(); macro_rules! err { () => { return Err(DecodingError) }; } macro_rules! next { () => { match iter.next() { Some(&byte) => byte, None => return Err(DecodingError), } }; } macro_rules! next_continuation { () => {{ let byte = next!(); if is_continuation_byte(byte) { byte } else { return Err(DecodingError); } }}; } loop { let first = match iter.next() { Some(&byte) => byte, None => break, }; if first <= MAX_ASCII_CODEPOINT { decoded.push(first) } else { let width = match utf8_char_width(first) { Some(v) => v, None => err!(), }; let second = next_continuation!(); match width { 2 => decoded.extend_from_slice(&[first, second]), 3 => { let third = next_continuation!(); match (first, second) { (0xE0, 0xA0..=0xBF) | (0xE1..=0xEC, 0x80..=0xBF) | (0xED, 0x80..=0x9F) | (0xEE..=0xEF, 0x80..=0xBF) => { decoded.extend_from_slice(&[first, second, third]); } (0xED, 0xA0..=0xAF) => { let fourth = next!(); if fourth != 0xED { err!(); } let fifth = next_continuation!(); if fifth < 0xB0 || 0xBF < fifth { err!(); } let sixth = next_continuation!(); decoded.extend_from_slice(&decode_surrogate_pair( second, third, fifth, sixth, )); } _ => err!(), } } _ => err!(), } } } debug_assert!(from_utf8(&decoded).is_ok()); Ok(Cow::Owned(unsafe { String::from_utf8_unchecked(decoded) })) } fn decode_surrogate_pair(second: u8, third: u8, fifth: u8, sixth: u8) -> [u8; 4] { let surrogate1 = decode_surrogate(second, third); let surrogate2 = decode_surrogate(fifth, sixth); let codepoint = 0x10000 + ((surrogate1 - 0xD800) << 10 | (surrogate2 - 0xDC00)); decode_codepoint(codepoint) } fn decode_surrogate(second: u8, third: u8) -> u32 { 0xD000 | ((second & VAL_MASK) as u32) << 6 | (third & VAL_MASK) as u32 } fn decode_codepoint(codepoint: u32) -> [u8; 4] { const STRT_TAG: u8 = 0b11110000; [ STRT_TAG | ((codepoint & 0b1_1100_0000_0000_0000_0000) >> 18) as u8, CONT_TAG | ((codepoint & 0b0_0011_1111_0000_0000_0000) >> 12) as u8, CONT_TAG | ((codepoint & 0b0_0000_0000_1111_1100_0000) >> 6) as u8, CONT_TAG | ((codepoint & 0b0_0000_0000_0000_0011_1111) as u8), ] } pub fn encode(text: &str) -> Cow<[u8]> { if is_valid(text) { return Cow::Borrowed(text.as_bytes()); } let bytes = text.as_bytes(); let capacity = encoded_capacity(text); let mut encoded = Vec::with_capacity(capacity); let mut index = 0; while index < bytes.len() { let byte = bytes[index]; if byte <= MAX_ASCII_CODEPOINT { encoded.push(byte); index += 1; } else { let width = utf8_char_width(byte).unwrap(); assert!(index + width <= bytes.len()); let slice_range = index..index + width; if width <= CESU8_MAX_CHAR_WIDTH { encoded.extend(&bytes[slice_range]) } else { let str = &text[slice_range]; let codepoint = str.chars().next().unwrap() as u32; let surrogate_pair = to_surrogate_pair(codepoint); let encoded_pair = encode_surrogate_pair(surrogate_pair); encoded.extend(&encoded_pair) } } } Cow::Owned(encoded) } fn encode_surrogate_pair(surrogate_pair: [u16; 2]) -> [u8; 6] { let [b1, b2, b3] = encode_surrogate(surrogate_pair[0]); let [b4, b5, b6] = encode_surrogate(surrogate_pair[1]); [b1, b2, b3, b4, b5, b6] } fn encode_surrogate(surrogate: u16) -> [u8; 3] { const STRT_TAG: u8 = 0b11100000; [ STRT_TAG | ((surrogate & 0b11110000_00000000) >> 12) as u8, CONT_TAG | ((surrogate & 0b00001111_11000000) >> 6) as u8, CONT_TAG | ((surrogate & 0b00000000_00111111) as u8), ] } fn to_surrogate_pair(codepoint: u32) -> [u16; 2] { let codepoint = codepoint - 0x10000; let first = ((codepoint >> 10) as u16) | 0xD800; let second = ((codepoint & 0x3FF) as u16) | 0xDC00; [first, second] } fn encoded_capacity(text: &str) -> usize { let bytes = text.as_bytes(); let mut capacity = 0; let mut index = 0; while index < bytes.len() { let byte = bytes[index]; if byte <= MAX_ASCII_CODEPOINT { capacity += 1; index += 1; } else { let width = utf8_char_width(byte).unwrap(); capacity += if width <= CESU8_MAX_CHAR_WIDTH { width } else { 6 }; index += width; } } capacity } const MAX_ASCII_CODEPOINT: u8 = 0x7F; pub fn is_valid(str: &str) -> bool { for byte in str.bytes() { if is_continuation_byte(byte) { continue; } if let Some(width) = utf8_char_width(byte) { if width > CESU8_MAX_CHAR_WIDTH { return false; } } else { return false; } } true } const CESU8_MAX_CHAR_WIDTH: usize = 3; fn is_continuation_byte(byte: u8) -> bool { byte & TAG_MASK == CONT_TAG } const VAL_MASK: u8 = 0b00111111; const TAG_MASK: u8 = 0b11000000; const CONT_TAG: u8 = 0b10000000; fn utf8_char_width(byte: u8) -> Option<usize> { match byte { 0x00..=0x7F => Some(1), 0xC2..=0xDF => Some(2), 0xE0..=0xEF => Some(3), 0xF0..=0xF4 => Some(4), _ => None, } }