1 /* punycode.c - punycode encoding/decoding
2 Copyright (C) 2011-2021 Simon Josefsson
3
4 Libidn2 is free software: you can redistribute it and/or modify it
5 under the terms of either:
6
7 * the GNU Lesser General Public License as published by the Free
8 Software Foundation; either version 3 of the License, or (at
9 your option) any later version.
10
11 or
12
13 * the GNU General Public License as published by the Free
14 Software Foundation; either version 2 of the License, or (at
15 your option) any later version.
16
17 or both in parallel, as here.
18
19 This program is distributed in the hope that it will be useful,
20 but WITHOUT ANY WARRANTY; without even the implied warranty of
21 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
22 GNU General Public License for more details.
23
24 You should have received copies of the GNU General Public License and
25 the GNU Lesser General Public License along with this program. If
26 not, see <http://www.gnu.org/licenses/>.
27 */
28
29 /*
30 Code copied from http://www.nicemice.net/idn/punycode-spec.gz on
31 2011-01-04 with SHA-1 a966a8017f6be579d74a50a226accc7607c40133
32 labeled punycode-spec 1.0.3 (2006-Mar-23-Thu). It is modified for
33 Libidn2 by Simon Josefsson. License on the original code:
34
35 punycode-spec 1.0.3 (2006-Mar-23-Thu)
36 http://www.nicemice.net/idn/
37 Adam M. Costello
38 http://www.nicemice.net/amc/
39
40 B. Disclaimer and license
41
42 Regarding this entire document or any portion of it (including
43 the pseudocode and C code), the author makes no guarantees and
44 is not responsible for any damage resulting from its use. The
45 author grants irrevocable permission to anyone to use, modify,
46 and distribute it in any way that does not diminish the rights
47 of anyone else to use, modify, and distribute it, provided that
48 redistributed derivative works do not contain misleading author or
49 version information. Derivative works need not be licensed under
50 similar terms.
51
52 C. Punycode sample implementation
53
54 punycode-sample.c 2.0.0 (2004-Mar-21-Sun)
55 http://www.nicemice.net/idn/
56 Adam M. Costello
57 http://www.nicemice.net/amc/
58
59 This is ANSI C code (C89) implementing Punycode 1.0.x.
60 */
61
62 #include <config.h>
63
64 #include "idn2.h" /* IDN2_OK, ... */
65
66 /* Re-definitions to avoid modifying code below too much. */
67 #define punycode_uint uint32_t
68 #define punycode_success IDN2_OK
69 #define punycode_overflow IDN2_PUNYCODE_OVERFLOW
70 #define punycode_big_output IDN2_PUNYCODE_BIG_OUTPUT
71 #define punycode_bad_input IDN2_PUNYCODE_BAD_INPUT
72 #define punycode_decode _idn2_punycode_decode_internal
73
74 /**********************************************************/
75 /* Implementation (would normally go in its own .c file): */
76
77 #include <stdio.h>
78 #include <string.h>
79
80 #include "punycode.h"
81
82 /*** Bootstring parameters for Punycode ***/
83
84 enum
85 { base = 36, tmin = 1, tmax = 26, skew = 38, damp = 700,
86 initial_bias = 72, initial_n = 0x80, delimiter = 0x2D
87 };
88
89 /* basic(cp) tests whether cp is a basic code point: */
90 #define basic(cp) ((cp >= 'a' && cp <= 'z') || (cp >= '0' && cp <='9') || (cp >= 'A' && cp <='Z') || cp == '-' || cp == '_')
91
92 /* decode_digit(cp) returns the numeric value of a basic code */
93 /* point (for use in representing integers) in the range 0 to */
94 /* base-1, or base if cp does not represent a value. */
95
96 static unsigned
decode_digit(int cp)97 decode_digit (int cp)
98 {
99 if (cp >= 'a' && cp <= 'z')
100 return cp - 'a';
101 if (cp >= '0' && cp <= '9')
102 return cp - '0' + 26;
103 if (cp >= 'A' && cp <= 'Z')
104 return cp - 'A';
105
106 return 0;
107 }
108
109 /*** Platform-specific constants ***/
110
111 /* maxint is the maximum value of a punycode_uint variable: */
112 static const punycode_uint maxint = -1;
113 /* Because maxint is unsigned, -1 becomes the maximum value. */
114
115 /*** Bias adaptation function ***/
116
117 static punycode_uint
118 adapt (punycode_uint delta, punycode_uint numpoints, int firsttime)
119 _GL_ATTRIBUTE_CONST;
120
adapt(punycode_uint delta,punycode_uint numpoints,int firsttime)121 static punycode_uint adapt (punycode_uint delta, punycode_uint numpoints,
122 int firsttime)
123 {
124 punycode_uint k;
125
126 delta = firsttime ? delta / damp : delta >> 1;
127 /* delta >> 1 is a faster way of doing delta / 2 */
128 delta += delta / numpoints;
129
130 for (k = 0; delta > ((base - tmin) * tmax) / 2; k += base)
131 {
132 delta /= base - tmin;
133 }
134
135 return k + (base - tmin + 1) * delta / (delta + skew);
136 }
137
138 /*** Main decode function ***/
139
140 int
punycode_decode(size_t input_length,const char input[],size_t * output_length,punycode_uint output[])141 punycode_decode (size_t input_length,
142 const char input[],
143 size_t *output_length, punycode_uint output[])
144 {
145 punycode_uint n, out = 0, i, max_out, bias, oldi, w, k, digit, t;
146 size_t b = 0, j, in;
147
148 if (!input_length)
149 return punycode_bad_input;
150
151 /* Check that all chars are basic */
152 for (j = 0; j < input_length; ++j)
153 {
154 if (!basic (input[j]))
155 return punycode_bad_input;
156 if (input[j] == delimiter)
157 b = j;
158 }
159
160 max_out =
161 *output_length > maxint ? maxint : (punycode_uint) * output_length;
162
163 if (input[b] == delimiter)
164 {
165 /* do not accept leading or trailing delimiter
166 * - leading delim must be omitted if there is no ASCII char in u-label
167 * - trailing delim means there where no non-ASCII chars in u-label
168 */
169 if (!b || b == input_length - 1)
170 return punycode_bad_input;
171
172 if (b >= max_out)
173 return punycode_big_output;
174
175 /* Check that all chars before last delimiter are basic chars */
176 /* and copy the first b code points to the output. */
177 for (j = 0; j < b; j++)
178 output[out++] = input[j];
179
180 b += 1; /* advance to non-basic char encoding */
181 }
182
183 /* Initialize the state: */
184 n = initial_n;
185 i = 0;
186 bias = initial_bias;
187
188 /* Main decoding loop: Start just after the last delimiter if any */
189 /* basic code points were copied; start at the beginning otherwise. */
190 for (in = b; in < input_length; ++out)
191 {
192
193 /* in is the index of the next ASCII code point to be consumed, */
194 /* and out is the number of code points in the output array. */
195
196 /* Decode a generalized variable-length integer into delta, */
197 /* which gets added to i. The overflow checking is easier */
198 /* if we increase i as we go, then subtract off its starting */
199 /* value at the end to obtain delta. */
200 for (oldi = i, w = 1, k = base;; k += base)
201 {
202 if (in >= input_length)
203 return punycode_bad_input;
204 digit = decode_digit (input[in++]);
205 if (digit >= base)
206 return punycode_bad_input;
207 if (digit > (maxint - i) / w)
208 return punycode_overflow;
209 i += digit * w;
210 t = k <= bias /* + tmin */ ? tmin : /* +tmin not needed */
211 k >= bias + tmax ? tmax : k - bias;
212 if (digit < t)
213 break;
214 if (w > maxint / (base - t))
215 return punycode_overflow;
216 w *= (base - t);
217 }
218
219 bias = adapt (i - oldi, out + 1, oldi == 0);
220
221 /* i was supposed to wrap around from out+1 to 0, */
222 /* incrementing n each time, so we'll fix that now: */
223 if (i / (out + 1) > maxint - n)
224 return punycode_overflow;
225 n += i / (out + 1);
226 if (n > 0x10FFFF || (n >= 0xD800 && n <= 0xDBFF))
227 return punycode_bad_input;
228 i %= (out + 1);
229
230 /* Insert n at position i of the output: */
231
232 /* not needed for Punycode: */
233 /* if (basic(n)) return punycode_bad_input; */
234 if (out >= max_out)
235 return punycode_big_output;
236
237 memmove (output + i + 1, output + i, (out - i) * sizeof *output);
238 output[i++] = n;
239 }
240
241 *output_length = (size_t) out;
242 /* cannot overflow because out <= old value of *output_length */
243 return punycode_success;
244 }
245
246 /* Create a compatibility symbol if supported. Hidden references make
247 the target symbol hidden, hence the alias. */
248 #ifdef HAVE_SYMVER_ALIAS_SUPPORT
249 __typeof__ (_idn2_punycode_decode_internal) _idn2_punycode_decode
250 __attribute__((visibility ("default"),
251 alias ("_idn2_punycode_decode_internal")));
252 __asm__ (".symver _idn2_punycode_decode, _idn2_punycode_decode@IDN2_0.0.0");
253 #endif
254