1 // Copyright 2013 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
9 //! Punycode ([RFC 3492](http://tools.ietf.org/html/rfc3492)) implementation.
10 //!
11 //! Since Punycode fundamentally works on unicode code points,
12 //! `encode` and `decode` take and return slices and vectors of `char`.
13 //! `encode_str` and `decode_to_string` provide convenience wrappers
14 //! that convert from and to Rust’s UTF-8 based `str` and `String` types.
15
16 use std::char;
17 use std::u32;
18
19 // Bootstring parameters for Punycode
20 static BASE: u32 = 36;
21 static T_MIN: u32 = 1;
22 static T_MAX: u32 = 26;
23 static SKEW: u32 = 38;
24 static DAMP: u32 = 700;
25 static INITIAL_BIAS: u32 = 72;
26 static INITIAL_N: u32 = 0x80;
27 static DELIMITER: char = '-';
28
29 #[inline]
adapt(mut delta: u32, num_points: u32, first_time: bool) -> u3230 fn adapt(mut delta: u32, num_points: u32, first_time: bool) -> u32 {
31 delta /= if first_time { DAMP } else { 2 };
32 delta += delta / num_points;
33 let mut k = 0;
34 while delta > ((BASE - T_MIN) * T_MAX) / 2 {
35 delta /= BASE - T_MIN;
36 k += BASE;
37 }
38 k + (((BASE - T_MIN + 1) * delta) / (delta + SKEW))
39 }
40
41 /// Convert Punycode to an Unicode `String`.
42 ///
43 /// This is a convenience wrapper around `decode`.
44 #[inline]
decode_to_string(input: &str) -> Option<String>45 pub fn decode_to_string(input: &str) -> Option<String> {
46 decode(input).map(|chars| chars.into_iter().collect())
47 }
48
49 /// Convert Punycode to Unicode.
50 ///
51 /// Return None on malformed input or overflow.
52 /// Overflow can only happen on inputs that take more than
53 /// 63 encoded bytes, the DNS limit on domain name labels.
decode(input: &str) -> Option<Vec<char>>54 pub fn decode(input: &str) -> Option<Vec<char>> {
55 Some(Decoder::default().decode(input).ok()?.collect())
56 }
57
58 #[derive(Default)]
59 pub(crate) struct Decoder {
60 insertions: Vec<(usize, char)>,
61 }
62
63 impl Decoder {
64 /// Split the input iterator and return a Vec with insertions of encoded characters
decode<'a>(&'a mut self, input: &'a str) -> Result<Decode<'a>, ()>65 pub(crate) fn decode<'a>(&'a mut self, input: &'a str) -> Result<Decode<'a>, ()> {
66 self.insertions.clear();
67 // Handle "basic" (ASCII) code points.
68 // They are encoded as-is before the last delimiter, if any.
69 let (base, input) = match input.rfind(DELIMITER) {
70 None => ("", input),
71 Some(position) => (
72 &input[..position],
73 if position > 0 {
74 &input[position + 1..]
75 } else {
76 input
77 },
78 ),
79 };
80
81 if !base.is_ascii() {
82 return Err(());
83 }
84
85 let base_len = base.len();
86 let mut length = base_len as u32;
87 let mut code_point = INITIAL_N;
88 let mut bias = INITIAL_BIAS;
89 let mut i = 0;
90 let mut iter = input.bytes();
91 loop {
92 let previous_i = i;
93 let mut weight = 1;
94 let mut k = BASE;
95 let mut byte = match iter.next() {
96 None => break,
97 Some(byte) => byte,
98 };
99
100 // Decode a generalized variable-length integer into delta,
101 // which gets added to i.
102 loop {
103 let digit = match byte {
104 byte @ b'0'..=b'9' => byte - b'0' + 26,
105 byte @ b'A'..=b'Z' => byte - b'A',
106 byte @ b'a'..=b'z' => byte - b'a',
107 _ => return Err(()),
108 } as u32;
109 if digit > (u32::MAX - i) / weight {
110 return Err(()); // Overflow
111 }
112 i += digit * weight;
113 let t = if k <= bias {
114 T_MIN
115 } else if k >= bias + T_MAX {
116 T_MAX
117 } else {
118 k - bias
119 };
120 if digit < t {
121 break;
122 }
123 if weight > u32::MAX / (BASE - t) {
124 return Err(()); // Overflow
125 }
126 weight *= BASE - t;
127 k += BASE;
128 byte = match iter.next() {
129 None => return Err(()), // End of input before the end of this delta
130 Some(byte) => byte,
131 };
132 }
133
134 bias = adapt(i - previous_i, length + 1, previous_i == 0);
135 if i / (length + 1) > u32::MAX - code_point {
136 return Err(()); // Overflow
137 }
138
139 // i was supposed to wrap around from length+1 to 0,
140 // incrementing code_point each time.
141 code_point += i / (length + 1);
142 i %= length + 1;
143 let c = match char::from_u32(code_point) {
144 Some(c) => c,
145 None => return Err(()),
146 };
147
148 // Move earlier insertions farther out in the string
149 for (idx, _) in &mut self.insertions {
150 if *idx >= i as usize {
151 *idx += 1;
152 }
153 }
154 self.insertions.push((i as usize, c));
155 length += 1;
156 i += 1;
157 }
158
159 self.insertions.sort_by_key(|(i, _)| *i);
160 Ok(Decode {
161 base: base.chars(),
162 insertions: &self.insertions,
163 inserted: 0,
164 position: 0,
165 len: base_len + self.insertions.len(),
166 })
167 }
168 }
169
170 pub(crate) struct Decode<'a> {
171 base: std::str::Chars<'a>,
172 pub(crate) insertions: &'a [(usize, char)],
173 inserted: usize,
174 position: usize,
175 len: usize,
176 }
177
178 impl<'a> Iterator for Decode<'a> {
179 type Item = char;
180
next(&mut self) -> Option<Self::Item>181 fn next(&mut self) -> Option<Self::Item> {
182 loop {
183 match self.insertions.get(self.inserted) {
184 Some((pos, c)) if *pos == self.position => {
185 self.inserted += 1;
186 self.position += 1;
187 return Some(*c);
188 }
189 _ => {}
190 }
191 if let Some(c) = self.base.next() {
192 self.position += 1;
193 return Some(c);
194 } else if self.inserted >= self.insertions.len() {
195 return None;
196 }
197 }
198 }
199
size_hint(&self) -> (usize, Option<usize>)200 fn size_hint(&self) -> (usize, Option<usize>) {
201 let len = self.len - self.position;
202 (len, Some(len))
203 }
204 }
205
206 impl<'a> ExactSizeIterator for Decode<'a> {
len(&self) -> usize207 fn len(&self) -> usize {
208 self.len - self.position
209 }
210 }
211
212 /// Convert an Unicode `str` to Punycode.
213 ///
214 /// This is a convenience wrapper around `encode`.
215 #[inline]
encode_str(input: &str) -> Option<String>216 pub fn encode_str(input: &str) -> Option<String> {
217 let mut buf = String::with_capacity(input.len());
218 encode_into(input.chars(), &mut buf).ok().map(|()| buf)
219 }
220
221 /// Convert Unicode to Punycode.
222 ///
223 /// Return None on overflow, which can only happen on inputs that would take more than
224 /// 63 encoded bytes, the DNS limit on domain name labels.
encode(input: &[char]) -> Option<String>225 pub fn encode(input: &[char]) -> Option<String> {
226 let mut buf = String::with_capacity(input.len());
227 encode_into(input.iter().copied(), &mut buf)
228 .ok()
229 .map(|()| buf)
230 }
231
encode_into<I>(input: I, output: &mut String) -> Result<(), ()> where I: Iterator<Item = char> + Clone,232 pub(crate) fn encode_into<I>(input: I, output: &mut String) -> Result<(), ()>
233 where
234 I: Iterator<Item = char> + Clone,
235 {
236 // Handle "basic" (ASCII) code points. They are encoded as-is.
237 let (mut input_length, mut basic_length) = (0, 0);
238 for c in input.clone() {
239 input_length += 1;
240 if c.is_ascii() {
241 output.push(c);
242 basic_length += 1;
243 }
244 }
245
246 if basic_length > 0 {
247 output.push('-')
248 }
249 let mut code_point = INITIAL_N;
250 let mut delta = 0;
251 let mut bias = INITIAL_BIAS;
252 let mut processed = basic_length;
253 while processed < input_length {
254 // All code points < code_point have been handled already.
255 // Find the next larger one.
256 let min_code_point = input
257 .clone()
258 .map(|c| c as u32)
259 .filter(|&c| c >= code_point)
260 .min()
261 .unwrap();
262 if min_code_point - code_point > (u32::MAX - delta) / (processed + 1) {
263 return Err(()); // Overflow
264 }
265 // Increase delta to advance the decoder’s <code_point,i> state to <min_code_point,0>
266 delta += (min_code_point - code_point) * (processed + 1);
267 code_point = min_code_point;
268 for c in input.clone() {
269 let c = c as u32;
270 if c < code_point {
271 delta += 1;
272 if delta == 0 {
273 return Err(()); // Overflow
274 }
275 }
276 if c == code_point {
277 // Represent delta as a generalized variable-length integer:
278 let mut q = delta;
279 let mut k = BASE;
280 loop {
281 let t = if k <= bias {
282 T_MIN
283 } else if k >= bias + T_MAX {
284 T_MAX
285 } else {
286 k - bias
287 };
288 if q < t {
289 break;
290 }
291 let value = t + ((q - t) % (BASE - t));
292 output.push(value_to_digit(value));
293 q = (q - t) / (BASE - t);
294 k += BASE;
295 }
296 output.push(value_to_digit(q));
297 bias = adapt(delta, processed + 1, processed == basic_length);
298 delta = 0;
299 processed += 1;
300 }
301 }
302 delta += 1;
303 code_point += 1;
304 }
305 Ok(())
306 }
307
308 #[inline]
value_to_digit(value: u32) -> char309 fn value_to_digit(value: u32) -> char {
310 match value {
311 0..=25 => (value as u8 + b'a') as char, // a..z
312 26..=35 => (value as u8 - 26 + b'0') as char, // 0..9
313 _ => panic!(),
314 }
315 }
316