xref: /openssl/test/punycode_test.c (revision da1c088f)
1 /*
2  * Copyright 2022-2023 The OpenSSL Project Authors. All Rights Reserved.
3  *
4  * Licensed under the Apache License 2.0 (the "License").  You may not use
5  * this file except in compliance with the License.  You can obtain a copy
6  * in the file LICENSE in the source distribution or at
7  * https://www.openssl.org/source/license.html
8  */
9 
10 #include <openssl/crypto.h>
11 #include <string.h>
12 
13 #include "crypto/punycode.h"
14 #include "internal/nelem.h"
15 #include "internal/packet.h"
16 #include "testutil.h"
17 
18 
19 static const struct puny_test {
20     unsigned int raw[50];
21     const char *encoded;
22 } puny_cases[] = {
23     { /* Test of 4 byte codepoint using smileyface emoji */
24         { 0x1F600
25         },
26         "e28h"
27     },
28     /* Test cases from RFC 3492 */
29     {   /* Arabic (Egyptian) */
30         { 0x0644, 0x064A, 0x0647, 0x0645, 0x0627, 0x0628, 0x062A, 0x0643, 0x0644,
31           0x0645, 0x0648, 0x0634, 0x0639, 0x0631, 0x0628, 0x064A, 0x061F
32         },
33         "egbpdaj6bu4bxfgehfvwxn"
34     },
35     {   /* Chinese (simplified) */
36         { 0x4ED6, 0x4EEC, 0x4E3A, 0x4EC0, 0x4E48, 0x4E0D, 0x8BF4, 0x4E2D, 0x6587
37         },
38         "ihqwcrb4cv8a8dqg056pqjye"
39     },
40     {   /* Chinese (traditional) */
41         { 0x4ED6, 0x5011, 0x7232, 0x4EC0, 0x9EBD, 0x4E0D, 0x8AAA, 0x4E2D, 0x6587
42         },
43         "ihqwctvzc91f659drss3x8bo0yb"
44     },
45     {    /* Czech: Pro<ccaron>prost<ecaron>nemluv<iacute><ccaron>esky */
46         { 0x0050, 0x0072, 0x006F, 0x010D, 0x0070, 0x0072, 0x006F, 0x0073, 0x0074,
47           0x011B, 0x006E, 0x0065, 0x006D, 0x006C, 0x0075, 0x0076, 0x00ED, 0x010D,
48           0x0065, 0x0073, 0x006B, 0x0079
49        },
50         "Proprostnemluvesky-uyb24dma41a"
51     },
52     {   /* Hebrew */
53         { 0x05DC, 0x05DE, 0x05D4, 0x05D4, 0x05DD, 0x05E4, 0x05E9, 0x05D5, 0x05D8,
54           0x05DC, 0x05D0, 0x05DE, 0x05D3, 0x05D1, 0x05E8, 0x05D9, 0x05DD, 0x05E2,
55           0x05D1, 0x05E8, 0x05D9, 0x05EA
56         },
57         "4dbcagdahymbxekheh6e0a7fei0b"
58     },
59     {   /* Hindi (Devanagari) */
60         { 0x092F, 0x0939, 0x0932, 0x094B, 0x0917, 0x0939, 0x093F, 0x0928, 0x094D,
61           0x0926, 0x0940, 0x0915, 0x094D, 0x092F, 0x094B, 0x0902, 0x0928, 0x0939,
62           0x0940, 0x0902, 0x092C, 0x094B, 0x0932, 0x0938, 0x0915, 0x0924, 0x0947,
63           0x0939, 0x0948, 0x0902
64         },
65         "i1baa7eci9glrd9b2ae1bj0hfcgg6iyaf8o0a1dig0cd"
66     },
67     {   /* Japanese (kanji and hiragana) */
68         { 0x306A, 0x305C, 0x307F, 0x3093, 0x306A, 0x65E5, 0x672C, 0x8A9E, 0x3092,
69           0x8A71, 0x3057, 0x3066, 0x304F, 0x308C, 0x306A, 0x3044, 0x306E, 0x304B
70         },
71         "n8jok5ay5dzabd5bym9f0cm5685rrjetr6pdxa"
72     },
73     {   /* Korean (Hangul syllables) */
74         { 0xC138, 0xACC4, 0xC758, 0xBAA8, 0xB4E0, 0xC0AC, 0xB78C, 0xB4E4, 0xC774,
75           0xD55C, 0xAD6D, 0xC5B4, 0xB97C, 0xC774, 0xD574, 0xD55C, 0xB2E4, 0xBA74,
76           0xC5BC, 0xB9C8, 0xB098, 0xC88B, 0xC744, 0xAE4C
77         },
78         "989aomsvi5e83db1d2a355cv1e0vak1dwrv93d5xbh15a0dt30a5jpsd879ccm6fea98c"
79     },
80     {   /* Russian (Cyrillic) */
81         { 0x043F, 0x043E, 0x0447, 0x0435, 0x043C, 0x0443, 0x0436, 0x0435, 0x043E,
82           0x043D, 0x0438, 0x043D, 0x0435, 0x0433, 0x043E, 0x0432, 0x043E, 0x0440,
83           0x044F, 0x0442, 0x043F, 0x043E, 0x0440, 0x0443, 0x0441, 0x0441, 0x043A,
84           0x0438
85         },
86         "b1abfaaepdrnnbgefbaDotcwatmq2g4l"
87     },
88     {   /* Spanish */
89         { 0x0050, 0x006F, 0x0072, 0x0071, 0x0075, 0x00E9, 0x006E, 0x006F, 0x0070,
90           0x0075, 0x0065, 0x0064, 0x0065, 0x006E, 0x0073, 0x0069, 0x006D, 0x0070,
91           0x006C, 0x0065, 0x006D, 0x0065, 0x006E, 0x0074, 0x0065, 0x0068, 0x0061,
92           0x0062, 0x006C, 0x0061, 0x0072, 0x0065, 0x006E, 0x0045, 0x0073, 0x0070,
93           0x0061, 0x00F1, 0x006F, 0x006C
94         },
95         "PorqunopuedensimplementehablarenEspaol-fmd56a"
96     },
97     {   /* Vietnamese */
98         { 0x0054, 0x1EA1, 0x0069, 0x0073, 0x0061, 0x006F, 0x0068, 0x1ECD, 0x006B,
99           0x0068, 0x00F4, 0x006E, 0x0067, 0x0074, 0x0068, 0x1EC3, 0x0063, 0x0068,
100           0x1EC9, 0x006E, 0x00F3, 0x0069, 0x0074, 0x0069, 0x1EBF, 0x006E, 0x0067,
101           0x0056, 0x0069, 0x1EC7, 0x0074
102         },
103         "TisaohkhngthchnitingVit-kjcr8268qyxafd2f1b9g"
104     },
105     {   /* Japanese: 3<nen>B<gumi><kinpachi><sensei> */
106         { 0x0033, 0x5E74, 0x0042, 0x7D44, 0x91D1, 0x516B, 0x5148, 0x751F
107         },
108         "3B-ww4c5e180e575a65lsy2b"
109     },
110     {   /* Japanese: <amuro><namie>-with-SUPER-MONKEYS */
111         { 0x5B89, 0x5BA4, 0x5948, 0x7F8E, 0x6075, 0x002D, 0x0077, 0x0069, 0x0074,
112           0x0068, 0x002D, 0x0053, 0x0055, 0x0050, 0x0045, 0x0052, 0x002D, 0x004D,
113           0x004F, 0x004E, 0x004B, 0x0045, 0x0059, 0x0053
114         },
115         "-with-SUPER-MONKEYS-pc58ag80a8qai00g7n9n"
116     },
117     {   /* Japanese: Hello-Another-Way-<sorezore><no><basho> */
118         { 0x0048, 0x0065, 0x006C, 0x006C, 0x006F, 0x002D, 0x0041, 0x006E, 0x006F,
119           0x0074, 0x0068, 0x0065, 0x0072, 0x002D, 0x0057, 0x0061, 0x0079, 0x002D,
120           0x305D, 0x308C, 0x305E, 0x308C, 0x306E, 0x5834, 0x6240
121         },
122         "Hello-Another-Way--fc4qua05auwb3674vfr0b"
123     },
124     {   /* Japanese: <hitotsu><yane><no><shita>2 */
125         { 0x3072, 0x3068, 0x3064, 0x5C4B, 0x6839, 0x306E, 0x4E0B, 0x0032
126         },
127         "2-u9tlzr9756bt3uc0v"
128     },
129     {   /* Japanese: Maji<de>Koi<suru>5<byou><mae> */
130         { 0x004D, 0x0061, 0x006A, 0x0069, 0x3067, 0x004B, 0x006F, 0x0069, 0x3059,
131           0x308B, 0x0035, 0x79D2, 0x524D
132         },
133         "MajiKoi5-783gue6qz075azm5e"
134     },
135     {   /* Japanese: <pafii>de<runba> */
136         { 0x30D1, 0x30D5, 0x30A3, 0x30FC, 0x0064, 0x0065, 0x30EB, 0x30F3, 0x30D0
137         },
138         "de-jg4avhby1noc0d"
139     },
140     {   /* Japanese: <sono><supiido><de> */
141         { 0x305D, 0x306E, 0x30B9, 0x30D4, 0x30FC, 0x30C9, 0x3067
142         },
143         "d9juau41awczczp"
144     },
145     {   /* -> $1.00 <- */
146         { 0x002D, 0x003E, 0x0020, 0x0024, 0x0031, 0x002E, 0x0030, 0x0030, 0x0020,
147           0x003C, 0x002D
148         },
149         "-> $1.00 <--"
150     }
151 };
152 
test_punycode(int n)153 static int test_punycode(int n)
154 {
155     const struct puny_test *tc = puny_cases + n;
156     unsigned int buffer[50];
157     unsigned int bsize = OSSL_NELEM(buffer);
158     size_t i;
159 
160     if (!TEST_true(ossl_punycode_decode(tc->encoded, strlen(tc->encoded),
161                                         buffer, &bsize)))
162         return 0;
163     for (i = 0; i < OSSL_NELEM(tc->raw); i++)
164         if (tc->raw[i] == 0)
165             break;
166     if (!TEST_mem_eq(buffer, bsize * sizeof(*buffer),
167                      tc->raw, i * sizeof(*tc->raw)))
168         return 0;
169     return 1;
170 }
171 
172 static const struct bad_decode_test {
173     size_t outlen;
174     const char input[20];
175 } bad_decode_tests[] = {
176     { 20, "xn--e-*" },   /* bad digit '*' */
177     { 10, "xn--e-999" }, /* loop > enc_len */
178     { 20, "xn--e-999999999" }, /* Too big */
179     { 20, {'x', 'n', '-', '-', (char)0x80, '-' } }, /* Not basic */
180     { 20, "xn--e-Oy65t" }, /* codepoint > 0x10FFFF */
181 };
182 
test_a2ulabel_bad_decode(int tst)183 static int test_a2ulabel_bad_decode(int tst)
184 {
185     char out[20];
186 
187     return TEST_int_eq(ossl_a2ulabel(bad_decode_tests[tst].input, out, bad_decode_tests[tst].outlen), -1);
188 }
189 
test_a2ulabel(void)190 static int test_a2ulabel(void)
191 {
192     char out[50];
193     char in[530] = { 0 };
194 
195     /*
196      * The punycode being passed in and parsed is malformed but we're not
197      * verifying that behaviour here.
198      */
199     if (!TEST_int_eq(ossl_a2ulabel("xn--a.b.c", out, 1), 0)
200             || !TEST_int_eq(ossl_a2ulabel("xn--a.b.c", out, 7), 1))
201         return 0;
202     /* Test for an off by one on the buffer size works */
203     if (!TEST_int_eq(ossl_a2ulabel("xn--a.b.c", out, 6), 0)
204             || !TEST_int_eq(ossl_a2ulabel("xn--a.b.c", out, 7), 1)
205             || !TEST_str_eq(out,"\xc2\x80.b.c"))
206         return 0;
207 
208     /* Test 4 byte smiley face */
209     if (!TEST_int_eq(ossl_a2ulabel("xn--e28h.com", out, 10), 1))
210         return 0;
211 
212     /* Test that we dont overflow the fixed internal buffer of 512 bytes when the starting bytes are copied */
213     strcpy(in, "xn--");
214     memset(in + 4, 'e', 513);
215     memcpy(in + 517, "-3ya", 4);
216     if (!TEST_int_eq(ossl_a2ulabel(in, out, 50), -1))
217         return 0;
218 
219     return 1;
220 }
221 
test_puny_overrun(void)222 static int test_puny_overrun(void)
223 {
224     static const unsigned int out[] = {
225         0x0033, 0x5E74, 0x0042, 0x7D44, 0x91D1, 0x516B, 0x5148, 0x751F
226     };
227     static const char *in = "3B-ww4c5e180e575a65lsy2b";
228     unsigned int buf[OSSL_NELEM(out)];
229     unsigned int bsize = OSSL_NELEM(buf) - 1;
230 
231     if (!TEST_false(ossl_punycode_decode(in, strlen(in), buf, &bsize))) {
232         if (TEST_mem_eq(buf, bsize * sizeof(*buf), out, sizeof(out)))
233             TEST_error("CRITICAL: buffer overrun detected!");
234         return 0;
235     }
236     return 1;
237 }
238 
test_dotted_overflow(void)239 static int test_dotted_overflow(void)
240 {
241     static const char string[] = "a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a";
242     const size_t num_reps = OSSL_NELEM(string) / 2;
243     WPACKET p;
244     BUF_MEM *in;
245     char *out = NULL;
246     size_t i;
247     int res = 0;
248 
249     /* Create out input punycode string */
250     if (!TEST_ptr(in = BUF_MEM_new()))
251         return 0;
252     if (!TEST_true(WPACKET_init_len(&p, in, 0))) {
253         BUF_MEM_free(in);
254         return 0;
255     }
256     for (i = 0; i < num_reps; i++) {
257         if (i > 1 && !TEST_true(WPACKET_put_bytes_u8(&p, '.')))
258             goto err;
259         if (!TEST_true(WPACKET_memcpy(&p, "xn--a", sizeof("xn--a") - 1)))
260             goto err;
261     }
262     if (!TEST_true(WPACKET_put_bytes_u8(&p, '\0')))
263             goto err;
264     if (!TEST_ptr(out = OPENSSL_malloc(in->length)))
265         goto err;
266 
267     /* Test the decode into an undersized buffer */
268     memset(out, 0x7f, in->length - 1);
269     if (!TEST_int_le(ossl_a2ulabel(in->data, out, num_reps), 0)
270             || !TEST_int_eq(out[num_reps], 0x7f))
271         goto err;
272 
273     /* Test the decode works into a full size buffer */
274     if (!TEST_int_gt(ossl_a2ulabel(in->data, out, in->length), 0)
275             || !TEST_size_t_eq(strlen(out), num_reps * 3))
276         goto err;
277 
278     res = 1;
279  err:
280     WPACKET_cleanup(&p);
281     BUF_MEM_free(in);
282     OPENSSL_free(out);
283     return res;
284 }
285 
setup_tests(void)286 int setup_tests(void)
287 {
288     ADD_ALL_TESTS(test_punycode, OSSL_NELEM(puny_cases));
289     ADD_TEST(test_dotted_overflow);
290     ADD_TEST(test_a2ulabel);
291     ADD_TEST(test_puny_overrun);
292     ADD_ALL_TESTS(test_a2ulabel_bad_decode, OSSL_NELEM(bad_decode_tests));
293     return 1;
294 }
295