Line data Source code
1 : /* tst_idna.c --- Self tests for idna_to_ascii().
2 : * Copyright (C) 2002-2020 Simon Josefsson
3 : *
4 : * This file is part of GNU Libidn.
5 : *
6 : * This program is free software: you can redistribute it and/or modify
7 : * it under the terms of the GNU General Public License as published by
8 : * the Free Software Foundation, either version 3 of the License, or
9 : * (at your option) any later version.
10 : *
11 : * This program is distributed in the hope that it will be useful,
12 : * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 : * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 : * GNU General Public License for more details.
15 : *
16 : * You should have received a copy of the GNU General Public License
17 : * along with this program. If not, see <http://www.gnu.org/licenses/>.
18 : *
19 : */
20 :
21 : #ifdef HAVE_CONFIG_H
22 : # include "config.h"
23 : #endif
24 :
25 : #include <stdio.h>
26 : #include <stdlib.h>
27 : #include <stdarg.h>
28 : #include <string.h>
29 :
30 : #include <stringprep.h>
31 : #include <idna.h>
32 : #include <idn-free.h>
33 :
34 : #include "utils.h"
35 :
36 : struct idna
37 : {
38 : const char *name;
39 : size_t inlen;
40 : uint32_t in[100];
41 : const char *out;
42 : int flags;
43 : int toasciirc;
44 : int tounicoderc;
45 : };
46 :
47 : static const struct idna idna[] = {
48 : {
49 : "Arabic (Egyptian)", 17,
50 : {
51 : 0x0644, 0x064A, 0x0647, 0x0645, 0x0627, 0x0628, 0x062A, 0x0643,
52 : 0x0644, 0x0645, 0x0648, 0x0634, 0x0639, 0x0631, 0x0628, 0x064A,
53 : 0x061F},
54 : IDNA_ACE_PREFIX "egbpdaj6bu4bxfgehfvwxn", 0, IDNA_SUCCESS, IDNA_SUCCESS},
55 : {
56 : "Chinese (simplified)", 9,
57 : {
58 : 0x4ED6, 0x4EEC, 0x4E3A, 0x4EC0, 0x4E48, 0x4E0D, 0x8BF4, 0x4E2D, 0x6587},
59 : IDNA_ACE_PREFIX "ihqwcrb4cv8a8dqg056pqjye", 0, IDNA_SUCCESS, IDNA_SUCCESS},
60 : {
61 : "Chinese (traditional)", 9,
62 : {
63 : 0x4ED6, 0x5011, 0x7232, 0x4EC0, 0x9EBD, 0x4E0D, 0x8AAA, 0x4E2D, 0x6587},
64 : IDNA_ACE_PREFIX "ihqwctvzc91f659drss3x8bo0yb", 0, IDNA_SUCCESS,
65 : IDNA_SUCCESS},
66 : {
67 : "Czech: Pro[CCARON]prost[ECARON]nemluv[IACUTE][CCARON]esky", 22,
68 : {
69 : 0x0050, 0x0072, 0x006F, 0x010D, 0x0070, 0x0072, 0x006F, 0x0073,
70 : 0x0074, 0x011B, 0x006E, 0x0065, 0x006D, 0x006C, 0x0075, 0x0076,
71 : 0x00ED, 0x010D, 0x0065, 0x0073, 0x006B, 0x0079},
72 : IDNA_ACE_PREFIX "Proprostnemluvesky-uyb24dma41a", 0, IDNA_SUCCESS,
73 : IDNA_SUCCESS},
74 : {
75 : "Hebrew", 22,
76 : {
77 : 0x05DC, 0x05DE, 0x05D4, 0x05D4, 0x05DD, 0x05E4, 0x05E9, 0x05D5,
78 : 0x05D8, 0x05DC, 0x05D0, 0x05DE, 0x05D3, 0x05D1, 0x05E8, 0x05D9,
79 : 0x05DD, 0x05E2, 0x05D1, 0x05E8, 0x05D9, 0x05EA},
80 : IDNA_ACE_PREFIX "4dbcagdahymbxekheh6e0a7fei0b", 0, IDNA_SUCCESS,
81 : IDNA_SUCCESS},
82 : {
83 : "Hindi (Devanagari)", 30,
84 : {
85 : 0x092F, 0x0939, 0x0932, 0x094B, 0x0917, 0x0939, 0x093F, 0x0928,
86 : 0x094D, 0x0926, 0x0940, 0x0915, 0x094D, 0x092F, 0x094B, 0x0902,
87 : 0x0928, 0x0939, 0x0940, 0x0902, 0x092C, 0x094B, 0x0932, 0x0938,
88 : 0x0915, 0x0924, 0x0947, 0x0939, 0x0948, 0x0902},
89 : IDNA_ACE_PREFIX "i1baa7eci9glrd9b2ae1bj0hfcgg6iyaf8o0a1dig0cd", 0,
90 : IDNA_SUCCESS},
91 : {
92 : "Japanese (kanji and hiragana)", 18,
93 : {
94 : 0x306A, 0x305C, 0x307F, 0x3093, 0x306A, 0x65E5, 0x672C, 0x8A9E,
95 : 0x3092, 0x8A71, 0x3057, 0x3066, 0x304F, 0x308C, 0x306A, 0x3044,
96 : 0x306E, 0x304B},
97 : IDNA_ACE_PREFIX "n8jok5ay5dzabd5bym9f0cm5685rrjetr6pdxa", 0, IDNA_SUCCESS},
98 : {
99 : "Korean (Hangul syllables)", 24,
100 : {
101 : 0xC138, 0xACC4, 0xC758, 0xBAA8, 0xB4E0, 0xC0AC, 0xB78C, 0xB4E4,
102 : 0xC774, 0xD55C, 0xAD6D, 0xC5B4, 0xB97C, 0xC774, 0xD574, 0xD55C,
103 : 0xB2E4, 0xBA74, 0xC5BC, 0xB9C8, 0xB098, 0xC88B, 0xC744, 0xAE4C},
104 : IDNA_ACE_PREFIX "989aomsvi5e83db1d2a355cv1e0vak1dwrv93d5xbh15a0dt"
105 : "30a5jpsd879ccm6fea98c", 0, IDNA_PUNYCODE_ERROR, IDNA_PUNYCODE_ERROR},
106 : /* too long output */
107 : {
108 : "Russian (Cyrillic)", 28,
109 : {
110 : 0x043F, 0x043E, 0x0447, 0x0435, 0x043C, 0x0443, 0x0436, 0x0435,
111 : 0x043E, 0x043D, 0x0438, 0x043D, 0x0435, 0x0433, 0x043E, 0x0432,
112 : 0x043E, 0x0440, 0x044F, 0x0442, 0x043F, 0x043E, 0x0440, 0x0443,
113 : 0x0441, 0x0441, 0x043A, 0x0438},
114 : IDNA_ACE_PREFIX "b1abfaaepdrnnbgefbadotcwatmq2g4l", 0,
115 : IDNA_SUCCESS, IDNA_SUCCESS},
116 : {
117 : "Spanish: Porqu[EACUTE]nopuedensimplementehablarenEspa[NTILDE]ol", 40,
118 : {
119 : 0x0050, 0x006F, 0x0072, 0x0071, 0x0075, 0x00E9, 0x006E, 0x006F,
120 : 0x0070, 0x0075, 0x0065, 0x0064, 0x0065, 0x006E, 0x0073, 0x0069,
121 : 0x006D, 0x0070, 0x006C, 0x0065, 0x006D, 0x0065, 0x006E, 0x0074,
122 : 0x0065, 0x0068, 0x0061, 0x0062, 0x006C, 0x0061, 0x0072, 0x0065,
123 : 0x006E, 0x0045, 0x0073, 0x0070, 0x0061, 0x00F1, 0x006F, 0x006C},
124 : IDNA_ACE_PREFIX "PorqunopuedensimplementehablarenEspaol-fmd56a", 0,
125 : IDNA_SUCCESS},
126 : {
127 : "Vietnamese", 31,
128 : {
129 : 0x0054, 0x1EA1, 0x0069, 0x0073, 0x0061, 0x006F, 0x0068, 0x1ECD,
130 : 0x006B, 0x0068, 0x00F4, 0x006E, 0x0067, 0x0074, 0x0068, 0x1EC3,
131 : 0x0063, 0x0068, 0x1EC9, 0x006E, 0x00F3, 0x0069, 0x0074, 0x0069,
132 : 0x1EBF, 0x006E, 0x0067, 0x0056, 0x0069, 0x1EC7, 0x0074},
133 : IDNA_ACE_PREFIX "TisaohkhngthchnitingVit-kjcr8268qyxafd2f1b9g", 0,
134 : IDNA_SUCCESS},
135 : {
136 : "Japanese 3[NEN]B[GUMI][KINPACHI][SENSEI]", 8,
137 : {
138 : 0x0033, 0x5E74, 0x0042, 0x7D44, 0x91D1, 0x516B, 0x5148, 0x751F},
139 : IDNA_ACE_PREFIX "3B-ww4c5e180e575a65lsy2b", 0, IDNA_SUCCESS,
140 : IDNA_SUCCESS},
141 : {
142 : "Japanese [AMURO][NAMIE]-with-SUPER-MONKEYS", 24,
143 : {
144 : 0x5B89, 0x5BA4, 0x5948, 0x7F8E, 0x6075, 0x002D, 0x0077, 0x0069,
145 : 0x0074, 0x0068, 0x002D, 0x0053, 0x0055, 0x0050, 0x0045, 0x0052,
146 : 0x002D, 0x004D, 0x004F, 0x004E, 0x004B, 0x0045, 0x0059, 0x0053},
147 : IDNA_ACE_PREFIX "-with-SUPER-MONKEYS-pc58ag80a8qai00g7n9n", 0,
148 : IDNA_SUCCESS},
149 : {
150 : "Japanese Hello-Another-Way-[SOREZORE][NO][BASHO]", 25,
151 : {
152 : 0x0048, 0x0065, 0x006C, 0x006C, 0x006F, 0x002D, 0x0041, 0x006E,
153 : 0x006F, 0x0074, 0x0068, 0x0065, 0x0072, 0x002D, 0x0057, 0x0061,
154 : 0x0079, 0x002D, 0x305D, 0x308C, 0x305E, 0x308C, 0x306E, 0x5834,
155 : 0x6240},
156 : IDNA_ACE_PREFIX "Hello-Another-Way--fc4qua05auwb3674vfr0b", 0,
157 : IDNA_SUCCESS},
158 : {
159 : "Japanese [HITOTSU][YANE][NO][SHITA]2", 8,
160 : {
161 : 0x3072, 0x3068, 0x3064, 0x5C4B, 0x6839, 0x306E, 0x4E0B, 0x0032},
162 : IDNA_ACE_PREFIX "2-u9tlzr9756bt3uc0v", 0, IDNA_SUCCESS, IDNA_SUCCESS},
163 : {
164 : "Japanese Maji[DE]Koi[SURU]5[BYOU][MAE]", 13,
165 : {
166 : 0x004D, 0x0061, 0x006A, 0x0069, 0x3067, 0x004B, 0x006F, 0x0069,
167 : 0x3059, 0x308B, 0x0035, 0x79D2, 0x524D},
168 : IDNA_ACE_PREFIX "MajiKoi5-783gue6qz075azm5e", 0, IDNA_SUCCESS,
169 : IDNA_SUCCESS},
170 : {
171 : "Japanese [PAFII]de[RUNBA]", 9,
172 : {
173 : 0x30D1, 0x30D5, 0x30A3, 0x30FC, 0x0064, 0x0065, 0x30EB, 0x30F3, 0x30D0},
174 : IDNA_ACE_PREFIX "de-jg4avhby1noc0d", 0, IDNA_SUCCESS, IDNA_SUCCESS},
175 : {
176 : "Japanese [SONO][SUPIIDO][DE]", 7,
177 : {
178 : 0x305D, 0x306E, 0x30B9, 0x30D4, 0x30FC, 0x30C9, 0x3067},
179 : IDNA_ACE_PREFIX "d9juau41awczczp", 0, IDNA_SUCCESS, IDNA_SUCCESS},
180 : {
181 : "Greek", 8,
182 : {
183 : 0x03b5, 0x03bb, 0x03bb, 0x03b7, 0x03bd, 0x03b9, 0x03ba, 0x03ac},
184 : IDNA_ACE_PREFIX "hxargifdar", 0, IDNA_SUCCESS, IDNA_SUCCESS},
185 : {
186 : "Maltese (Malti)", 10,
187 : {
188 : 0x0062, 0x006f, 0x006e, 0x0121, 0x0075, 0x0073, 0x0061, 0x0127,
189 : 0x0127, 0x0061},
190 : IDNA_ACE_PREFIX "bonusaa-5bb1da", 0, IDNA_SUCCESS, IDNA_SUCCESS},
191 : {
192 : "Russian (Cyrillic)", 28,
193 : {
194 : 0x043f, 0x043e, 0x0447, 0x0435, 0x043c, 0x0443, 0x0436, 0x0435,
195 : 0x043e, 0x043d, 0x0438, 0x043d, 0x0435, 0x0433, 0x043e, 0x0432,
196 : 0x043e, 0x0440, 0x044f, 0x0442, 0x043f, 0x043e, 0x0440, 0x0443,
197 : 0x0441, 0x0441, 0x043a, 0x0438},
198 : IDNA_ACE_PREFIX "b1abfaaepdrnnbgefbadotcwatmq2g4l", 0,
199 : IDNA_SUCCESS, IDNA_SUCCESS},
200 : #if 0
201 : {
202 : "(S) -> $1.00 <-", 11,
203 : {
204 : 0x002D, 0x003E, 0x0020, 0x0024, 0x0031, 0x002E, 0x0030, 0x0030,
205 : 0x0020, 0x003C, 0x002D},
206 : IDNA_ACE_PREFIX "-> $1.00 <--", 0, IDNA_SUCCESS, IDNA_SUCCESS},
207 : #endif
208 : { /* XXX depends on IDNA_ACE_PREFIX */
209 : "ToASCII() with ACE prefix", 4 + 3,
210 : {
211 : 'x', 'n', '-', '-', 'f', 'o', 0x3067},
212 : IDNA_ACE_PREFIX "too long too long too long too long too long too "
213 : "long too long too long too long too long ", 0,
214 : IDNA_CONTAINS_ACE_PREFIX, IDNA_INVALID_LENGTH}
215 : };
216 :
217 : void
218 1 : doit (void)
219 : {
220 : char label[100];
221 1 : uint32_t *ucs4label = NULL;
222 : uint32_t tmp[100];
223 : size_t len, len2;
224 : int rc;
225 : unsigned i;
226 :
227 23 : for (i = 0; i < sizeof (idna) / sizeof (idna[0]); i++)
228 : {
229 22 : if (debug)
230 0 : printf ("IDNA entry %u: %s\n", i, idna[i].name);
231 :
232 22 : if (debug)
233 : {
234 0 : printf ("in:\n");
235 0 : ucs4print (idna[i].in, idna[i].inlen);
236 : }
237 :
238 22 : rc = idna_to_ascii_4i (idna[i].in, idna[i].inlen, label, idna[i].flags);
239 22 : if (rc != idna[i].toasciirc)
240 : {
241 0 : fail ("IDNA entry %u failed: %d\n", i, rc);
242 0 : if (debug)
243 0 : printf ("FATAL\n");
244 0 : continue;
245 : }
246 :
247 22 : if (debug && rc == IDNA_SUCCESS)
248 : {
249 0 : printf ("computed out: %s\n", label);
250 0 : printf ("expected out: %s\n", idna[i].out);
251 : }
252 22 : else if (debug)
253 0 : printf ("returned %d expected %d\n", rc, idna[i].toasciirc);
254 :
255 22 : if (rc == IDNA_SUCCESS)
256 : {
257 20 : if (strlen (idna[i].out) != strlen (label) ||
258 20 : strcasecmp (idna[i].out, label) != 0)
259 : {
260 0 : fail ("IDNA entry %u failed\n", i);
261 0 : if (debug)
262 0 : printf ("ERROR\n");
263 : }
264 20 : else if (debug)
265 0 : printf ("OK\n");
266 : }
267 2 : else if (debug)
268 0 : printf ("OK\n");
269 :
270 22 : if (ucs4label)
271 21 : idn_free (ucs4label);
272 :
273 22 : ucs4label = stringprep_utf8_to_ucs4 (idna[i].out, -1, &len);
274 :
275 22 : if (debug)
276 : {
277 0 : printf ("in: %s (%d==%d)\n", idna[i].out,
278 0 : (int) strlen (idna[i].out), (int) len);
279 0 : ucs4print (ucs4label, len);
280 : }
281 :
282 22 : len2 = sizeof (tmp) / sizeof (tmp[0]);
283 22 : rc = idna_to_unicode_44i (ucs4label, len, tmp, &len2, idna[i].flags);
284 22 : if (debug)
285 : {
286 0 : printf ("expected out (%lu):\n",
287 : rc == IDNA_SUCCESS ? idna[i].inlen : len);
288 0 : if (rc == IDNA_SUCCESS)
289 0 : ucs4print (idna[i].in, idna[i].inlen);
290 : else
291 0 : ucs4print (ucs4label, len);
292 :
293 0 : printf ("computed out (%d):\n", (int) len2);
294 0 : ucs4print (tmp, len2);
295 : }
296 :
297 22 : if (rc != idna[i].tounicoderc)
298 : {
299 0 : fail ("IDNA entry %u failed: %d\n", i, rc);
300 0 : if (debug)
301 0 : printf ("FATAL\n");
302 0 : continue;
303 : }
304 :
305 22 : if ((rc == IDNA_SUCCESS && (len2 != idna[i].inlen ||
306 22 : memcmp (idna[i].in, tmp, len2) != 0)) ||
307 2 : (rc != IDNA_SUCCESS && (len2 != len ||
308 2 : memcmp (ucs4label, tmp, len) != 0)))
309 : {
310 0 : if (debug)
311 : {
312 0 : if (rc == IDNA_SUCCESS)
313 0 : printf ("len=%d len2=%d\n", (int) len2, (int) idna[i].inlen);
314 : else
315 0 : printf ("len=%d len2=%d\n", (int) len, (int) len2);
316 : }
317 0 : fail ("IDNA entry %u failed\n", i);
318 0 : if (debug)
319 0 : printf ("ERROR\n");
320 : }
321 22 : else if (debug)
322 0 : printf ("OK\n\n");
323 : }
324 :
325 1 : if (ucs4label)
326 1 : idn_free (ucs4label);
327 1 : }
|