summaryrefslogtreecommitdiffstats
path: root/third_party/rust/idna/src/punycode.rs
blob: b1afc96f7630ba96f0030dfe854c55865b392dc8 (plain)
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
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
// Copyright 2013 The rust-url developers.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.

//! Punycode ([RFC 3492](http://tools.ietf.org/html/rfc3492)) implementation.
//!
//! Since Punycode fundamentally works on unicode code points,
//! `encode` and `decode` take and return slices and vectors of `char`.
//! `encode_str` and `decode_to_string` provide convenience wrappers
//! that convert from and to Rust’s UTF-8 based `str` and `String` types.

use alloc::{string::String, vec::Vec};
use core::char;
use core::u32;

// Bootstring parameters for Punycode
static BASE: u32 = 36;
static T_MIN: u32 = 1;
static T_MAX: u32 = 26;
static SKEW: u32 = 38;
static DAMP: u32 = 700;
static INITIAL_BIAS: u32 = 72;
static INITIAL_N: u32 = 0x80;
static DELIMITER: char = '-';

#[inline]
fn adapt(mut delta: u32, num_points: u32, first_time: bool) -> u32 {
    delta /= if first_time { DAMP } else { 2 };
    delta += delta / num_points;
    let mut k = 0;
    while delta > ((BASE - T_MIN) * T_MAX) / 2 {
        delta /= BASE - T_MIN;
        k += BASE;
    }
    k + (((BASE - T_MIN + 1) * delta) / (delta + SKEW))
}

/// Convert Punycode to an Unicode `String`.
///
/// This is a convenience wrapper around `decode`.
#[inline]
pub fn decode_to_string(input: &str) -> Option<String> {
    decode(input).map(|chars| chars.into_iter().collect())
}

/// Convert Punycode to Unicode.
///
/// Return None on malformed input or overflow.
/// Overflow can only happen on inputs that take more than
/// 63 encoded bytes, the DNS limit on domain name labels.
pub fn decode(input: &str) -> Option<Vec<char>> {
    Some(Decoder::default().decode(input).ok()?.collect())
}

#[derive(Default)]
pub(crate) struct Decoder {
    insertions: Vec<(usize, char)>,
}

impl Decoder {
    /// Split the input iterator and return a Vec with insertions of encoded characters
    pub(crate) fn decode<'a>(&'a mut self, input: &'a str) -> Result<Decode<'a>, ()> {
        self.insertions.clear();
        // Handle "basic" (ASCII) code points.
        // They are encoded as-is before the last delimiter, if any.
        let (base, input) = match input.rfind(DELIMITER) {
            None => ("", input),
            Some(position) => (
                &input[..position],
                if position > 0 {
                    &input[position + 1..]
                } else {
                    input
                },
            ),
        };

        if !base.is_ascii() {
            return Err(());
        }

        let base_len = base.len();
        let mut length = base_len as u32;
        let mut code_point = INITIAL_N;
        let mut bias = INITIAL_BIAS;
        let mut i = 0;
        let mut iter = input.bytes();
        loop {
            let previous_i = i;
            let mut weight = 1;
            let mut k = BASE;
            let mut byte = match iter.next() {
                None => break,
                Some(byte) => byte,
            };

            // Decode a generalized variable-length integer into delta,
            // which gets added to i.
            loop {
                let digit = match byte {
                    byte @ b'0'..=b'9' => byte - b'0' + 26,
                    byte @ b'A'..=b'Z' => byte - b'A',
                    byte @ b'a'..=b'z' => byte - b'a',
                    _ => return Err(()),
                } as u32;
                if digit > (u32::MAX - i) / weight {
                    return Err(()); // Overflow
                }
                i += digit * weight;
                let t = if k <= bias {
                    T_MIN
                } else if k >= bias + T_MAX {
                    T_MAX
                } else {
                    k - bias
                };
                if digit < t {
                    break;
                }
                if weight > u32::MAX / (BASE - t) {
                    return Err(()); // Overflow
                }
                weight *= BASE - t;
                k += BASE;
                byte = match iter.next() {
                    None => return Err(()), // End of input before the end of this delta
                    Some(byte) => byte,
                };
            }

            bias = adapt(i - previous_i, length + 1, previous_i == 0);
            if i / (length + 1) > u32::MAX - code_point {
                return Err(()); // Overflow
            }

            // i was supposed to wrap around from length+1 to 0,
            // incrementing code_point each time.
            code_point += i / (length + 1);
            i %= length + 1;
            let c = match char::from_u32(code_point) {
                Some(c) => c,
                None => return Err(()),
            };

            // Move earlier insertions farther out in the string
            for (idx, _) in &mut self.insertions {
                if *idx >= i as usize {
                    *idx += 1;
                }
            }
            self.insertions.push((i as usize, c));
            length += 1;
            i += 1;
        }

        self.insertions.sort_by_key(|(i, _)| *i);
        Ok(Decode {
            base: base.chars(),
            insertions: &self.insertions,
            inserted: 0,
            position: 0,
            len: base_len + self.insertions.len(),
        })
    }
}

pub(crate) struct Decode<'a> {
    base: core::str::Chars<'a>,
    pub(crate) insertions: &'a [(usize, char)],
    inserted: usize,
    position: usize,
    len: usize,
}

impl<'a> Iterator for Decode<'a> {
    type Item = char;

    fn next(&mut self) -> Option<Self::Item> {
        loop {
            match self.insertions.get(self.inserted) {
                Some((pos, c)) if *pos == self.position => {
                    self.inserted += 1;
                    self.position += 1;
                    return Some(*c);
                }
                _ => {}
            }
            if let Some(c) = self.base.next() {
                self.position += 1;
                return Some(c);
            } else if self.inserted >= self.insertions.len() {
                return None;
            }
        }
    }

    fn size_hint(&self) -> (usize, Option<usize>) {
        let len = self.len - self.position;
        (len, Some(len))
    }
}

impl<'a> ExactSizeIterator for Decode<'a> {
    fn len(&self) -> usize {
        self.len - self.position
    }
}

/// Convert an Unicode `str` to Punycode.
///
/// This is a convenience wrapper around `encode`.
#[inline]
pub fn encode_str(input: &str) -> Option<String> {
    if input.len() > u32::MAX as usize {
        return None;
    }
    let mut buf = String::with_capacity(input.len());
    encode_into(input.chars(), &mut buf).ok().map(|()| buf)
}

/// Convert Unicode to Punycode.
///
/// Return None on overflow, which can only happen on inputs that would take more than
/// 63 encoded bytes, the DNS limit on domain name labels.
pub fn encode(input: &[char]) -> Option<String> {
    if input.len() > u32::MAX as usize {
        return None;
    }
    let mut buf = String::with_capacity(input.len());
    encode_into(input.iter().copied(), &mut buf)
        .ok()
        .map(|()| buf)
}

pub(crate) fn encode_into<I>(input: I, output: &mut String) -> Result<(), ()>
where
    I: Iterator<Item = char> + Clone,
{
    // Handle "basic" (ASCII) code points. They are encoded as-is.
    let (mut input_length, mut basic_length) = (0u32, 0);
    for c in input.clone() {
        input_length = input_length.checked_add(1).ok_or(())?;
        if c.is_ascii() {
            output.push(c);
            basic_length += 1;
        }
    }

    if basic_length > 0 {
        output.push('-')
    }
    let mut code_point = INITIAL_N;
    let mut delta = 0;
    let mut bias = INITIAL_BIAS;
    let mut processed = basic_length;
    while processed < input_length {
        // All code points < code_point have been handled already.
        // Find the next larger one.
        let min_code_point = input
            .clone()
            .map(|c| c as u32)
            .filter(|&c| c >= code_point)
            .min()
            .unwrap();
        if min_code_point - code_point > (u32::MAX - delta) / (processed + 1) {
            return Err(()); // Overflow
        }
        // Increase delta to advance the decoder’s <code_point,i> state to <min_code_point,0>
        delta += (min_code_point - code_point) * (processed + 1);
        code_point = min_code_point;
        for c in input.clone() {
            let c = c as u32;
            if c < code_point {
                delta = delta.checked_add(1).ok_or(())?;
            }
            if c == code_point {
                // Represent delta as a generalized variable-length integer:
                let mut q = delta;
                let mut k = BASE;
                loop {
                    let t = if k <= bias {
                        T_MIN
                    } else if k >= bias + T_MAX {
                        T_MAX
                    } else {
                        k - bias
                    };
                    if q < t {
                        break;
                    }
                    let value = t + ((q - t) % (BASE - t));
                    output.push(value_to_digit(value));
                    q = (q - t) / (BASE - t);
                    k += BASE;
                }
                output.push(value_to_digit(q));
                bias = adapt(delta, processed + 1, processed == basic_length);
                delta = 0;
                processed += 1;
            }
        }
        delta += 1;
        code_point += 1;
    }
    Ok(())
}

#[inline]
fn value_to_digit(value: u32) -> char {
    match value {
        0..=25 => (value as u8 + b'a') as char,       // a..z
        26..=35 => (value as u8 - 26 + b'0') as char, // 0..9
        _ => panic!(),
    }
}

#[test]
#[ignore = "slow"]
#[cfg(target_pointer_width = "64")]
fn huge_encode() {
    let mut buf = String::new();
    assert!(encode_into(std::iter::repeat('ß').take(u32::MAX as usize + 1), &mut buf).is_err());
    assert_eq!(buf.len(), 0);
}