t_c16rtomb.c revision 1.6 1 /* $NetBSD: t_c16rtomb.c,v 1.6 2024/08/19 16:22:10 riastradh Exp $ */
2
3 /*-
4 * Copyright (c) 2002 Tim J. Robbins
5 * All rights reserved.
6 *
7 * Copyright (c) 2013 Ed Schouten <ed (at) FreeBSD.org>
8 * All rights reserved.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 */
31 /*
32 * Test program for c16rtomb() as specified by ISO/IEC 9899:2011.
33 */
34
35 #include <sys/cdefs.h>
36 __RCSID("$NetBSD: t_c16rtomb.c,v 1.6 2024/08/19 16:22:10 riastradh Exp $");
37
38 #include <errno.h>
39 #include <limits.h>
40 #include <locale.h>
41 #include <stdio.h>
42 #include <string.h>
43 #include <uchar.h>
44
45 #include <atf-c.h>
46
47 static void
48 require_lc_ctype(const char *locale_name)
49 {
50 char *lc_ctype_set;
51
52 lc_ctype_set = setlocale(LC_CTYPE, locale_name);
53 if (lc_ctype_set == NULL)
54 atf_tc_fail("setlocale(LC_CTYPE, \"%s\") failed; errno=%d",
55 locale_name, errno);
56
57 ATF_REQUIRE_EQ_MSG(strcmp(lc_ctype_set, locale_name), 0,
58 "lc_ctype_set=%s locale_name=%s", lc_ctype_set, locale_name);
59 }
60
61 static mbstate_t s;
62 static char buf[7*MB_LEN_MAX + 1];
63
64 ATF_TC_WITHOUT_HEAD(c16rtomb_c_locale_test);
65 ATF_TC_BODY(c16rtomb_c_locale_test, tc)
66 {
67 size_t n;
68
69 require_lc_ctype("C");
70
71 /*
72 * If the buffer argument is NULL, c16 is implicitly 0,
73 * c16rtomb() resets its internal state.
74 */
75 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, L'\0', NULL)), 1, "n=%zu", n);
76 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, 0xdc00, NULL)), 1, "n=%zu", n);
77
78 /* Null wide character. */
79 memset(&s, 0, sizeof(s));
80 memset(buf, 0xcc, sizeof(buf));
81 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0, &s)), 1, "n=%zu", n);
82 ATF_CHECK_MSG(((unsigned char)buf[0] == 0 &&
83 (unsigned char)buf[1] == 0xcc),
84 "buf=[%02x %02x]", buf[0], buf[1]);
85
86 /* Latin letter A, internal state. */
87 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, L'\0', NULL)), 1, "n=%zu", n);
88 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, L'A', NULL)), 1, "n=%zu", n);
89
90 /* Latin letter A. */
91 memset(&s, 0, sizeof(s));
92 memset(buf, 0xcc, sizeof(buf));
93 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, L'A', &s)), 1, "n=%zu", n);
94 ATF_CHECK_MSG(((unsigned char)buf[0] == 'A' &&
95 (unsigned char)buf[1] == 0xcc),
96 "buf=[%02x %02x]", buf[0], buf[1]);
97
98 /* Unicode character 'Pile of poo'. */
99 memset(&s, 0, sizeof(s));
100 memset(buf, 0xcc, sizeof(buf));
101 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xd83d, &s)), 0, "n=%zu", n);
102 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xdca9, &s)), (size_t)-1,
103 "n=%zu", n);
104 ATF_CHECK_EQ_MSG(errno, EILSEQ, "errno=%d", errno);
105 ATF_CHECK_EQ_MSG((unsigned char)buf[0], 0xcc, "buf=[%02x]", buf[0]);
106
107 /* Incomplete Unicode character 'Pile of poo', interrupted by NUL. */
108 memset(&s, 0, sizeof(s));
109 memset(buf, 0xcc, sizeof(buf));
110 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xd83d, &s)), 0, "n=%zu", n);
111 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, L'\0', &s)), 1, "n=%zu", n);
112 ATF_CHECK_MSG(((unsigned char)buf[0] == '\0' &&
113 (unsigned char)buf[1] == 0xcc),
114 "buf=[%02x %02x]", buf[0], buf[1]);
115 }
116
117 ATF_TC_WITHOUT_HEAD(c16rtomb_iso2022jp_locale_test);
118 ATF_TC_BODY(c16rtomb_iso2022jp_locale_test, tc)
119 {
120 char *p;
121 size_t n;
122
123 require_lc_ctype("ja_JP.ISO-2022-JP");
124
125 /*
126 * If the buffer argument is NULL, c16 is implicitly 0,
127 * c16rtomb() resets its internal state.
128 */
129 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, L'\0', NULL)), 1, "n=%zu", n);
130 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, 0xdc00, NULL)), 1, "n=%zu", n);
131
132 /* Null wide character. */
133 memset(&s, 0, sizeof(s));
134 memset(buf, 0xcc, sizeof(buf));
135 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0, &s)), 1, "n=%zu", n);
136 ATF_CHECK_MSG(((unsigned char)buf[0] == 0 &&
137 (unsigned char)buf[1] == 0xcc),
138 "buf=[%02x %02x]", buf[0], buf[1]);
139
140 /* Latin letter A, internal state. */
141 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, L'\0', NULL)), 1, "n=%zu", n);
142 ATF_CHECK_EQ_MSG((n = c16rtomb(NULL, L'A', NULL)), 1, "n=%zu", n);
143
144 /*
145 * 1. U+0042 LATIN CAPITAL LETTER A
146 * 2. U+00A5 YEN SIGN
147 * 3. U+00A5 YEN SIGN (again, no shift needed)
148 * 4. U+30A2 KATAKANA LETTER A
149 * 5. U+30A2 KATAKANA LETTER A (again, no shift needed)
150 * 6. incomplete UTF-16 surrogate pair -- no output
151 * 7. U+0000 NUL (plus shift sequence to initial state)
152 */
153 memset(&s, 0, sizeof(s));
154 memset(buf, 0xcc, sizeof(buf));
155 p = buf;
156 ATF_CHECK_EQ_MSG((n = c16rtomb(p, L'A', &s)), 1, "n=%zu", n); /* 1 */
157 p += 1;
158 ATF_CHECK_EQ_MSG((n = c16rtomb(p, 0xa5, &s)), 4, "n=%zu", n); /* 2 */
159 p += 4;
160 ATF_CHECK_EQ_MSG((n = c16rtomb(p, 0xa5, &s)), 1, "n=%zu", n); /* 3 */
161 p += 1;
162 ATF_CHECK_EQ_MSG((n = c16rtomb(p, 0x30a2, &s)), 5, "n=%zu", n); /* 4 */
163 p += 5;
164 ATF_CHECK_EQ_MSG((n = c16rtomb(p, 0x30a2, &s)), 2, "n=%zu", n); /* 5 */
165 p += 2;
166 ATF_CHECK_EQ_MSG((n = c16rtomb(p, 0xd800, &s)), 0, "n=%zu", n); /* 6 */
167 ATF_CHECK_EQ_MSG((n = c16rtomb(p, L'\0', &s)), 4, "n=%zu", n); /* 7 */
168 p += 4;
169 ATF_CHECK_MSG(((unsigned char)buf[0] == 'A' &&
170 (unsigned char)buf[1] == 0x1b && /* shift ISO/IEC 646:JP */
171 (unsigned char)buf[2] == '(' &&
172 (unsigned char)buf[3] == 'J' &&
173 (unsigned char)buf[4] == 0x5c && /* YEN SIGN */
174 (unsigned char)buf[5] == 0x5c && /* YEN SIGN */
175 (unsigned char)buf[6] == 0x1b && /* shift JIS X 0208 */
176 (unsigned char)buf[7] == '$' &&
177 (unsigned char)buf[8] == 'B' &&
178 (unsigned char)buf[9] == 0x25 && /* KATAKANA LETTER A */
179 (unsigned char)buf[10] == 0x22 &&
180 (unsigned char)buf[11] == 0x25 && /* KATAKANA LETTER A */
181 (unsigned char)buf[12] == 0x22 &&
182 (unsigned char)buf[13] == 0x1b && /* shift US-ASCII */
183 (unsigned char)buf[14] == '(' &&
184 (unsigned char)buf[15] == 'B' &&
185 (unsigned char)buf[16] == '\0' &&
186 (unsigned char)buf[17] == 0xcc),
187 "buf=[%02x %02x %02x %02x %02x %02x %02x %02x "
188 " %02x %02x %02x %02x %02x %02x %02x %02x "
189 " %02x %02x]",
190 buf[0], buf[1], buf[2], buf[3],
191 buf[4], buf[5], buf[6], buf[7],
192 buf[8], buf[9], buf[10], buf[11],
193 buf[12], buf[13], buf[14], buf[15],
194 buf[16], buf[17]);
195 }
196
197 ATF_TC_WITHOUT_HEAD(c16rtomb_iso_8859_1_test);
198 ATF_TC_BODY(c16rtomb_iso_8859_1_test, tc)
199 {
200 size_t n;
201
202 require_lc_ctype("en_US.ISO8859-1");
203
204 /* Unicode character 'Euro sign'. */
205 memset(&s, 0, sizeof(s));
206 memset(buf, 0xcc, sizeof(buf));
207 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0x20ac, &s)), (size_t)-1,
208 "n=%zu", n);
209 ATF_CHECK_EQ_MSG(errno, EILSEQ, "errno=%d", errno);
210 ATF_CHECK_EQ_MSG((unsigned char)buf[0], 0xcc, "buf=[%02x]", buf[0]);
211 }
212
213 ATF_TC_WITHOUT_HEAD(c16rtomb_iso_8859_15_test);
214 ATF_TC_BODY(c16rtomb_iso_8859_15_test, tc)
215 {
216 size_t n;
217
218 require_lc_ctype("en_US.ISO8859-15");
219
220 /* Unicode character 'Euro sign'. */
221 memset(&s, 0, sizeof(s));
222 memset(buf, 0xcc, sizeof(buf));
223 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0x20ac, &s)), 1, "n=%zu", n);
224 ATF_CHECK_MSG(((unsigned char)buf[0] == 0xa4 &&
225 (unsigned char)buf[1] == 0xcc),
226 "buf=[%02x %02x]", buf[0], buf[1]);
227 }
228
229 ATF_TC_WITHOUT_HEAD(c16rtomb_utf_8_test);
230 ATF_TC_BODY(c16rtomb_utf_8_test, tc)
231 {
232 size_t n;
233
234 require_lc_ctype("en_US.UTF-8");
235
236 /* Unicode character 'Pile of poo'. */
237 memset(&s, 0, sizeof(s));
238 memset(buf, 0xcc, sizeof(buf));
239 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xd83d, &s)), 0, "n=%zu", n);
240 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xdca9, &s)), 4, "n=%zu", n);
241 ATF_CHECK_MSG(((unsigned char)buf[0] == 0xf0 &&
242 (unsigned char)buf[1] == 0x9f &&
243 (unsigned char)buf[2] == 0x92 &&
244 (unsigned char)buf[3] == 0xa9 &&
245 (unsigned char)buf[4] == 0xcc),
246 "buf=[%02x %02x %02x %02x %02x]",
247 buf[0], buf[1], buf[2], buf[3], buf[4]);
248
249 /* Invalid code; 'Pile of poo' without the trail surrogate. */
250 memset(&s, 0, sizeof(s));
251 memset(buf, 0xcc, sizeof(buf));
252 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xd83d, &s)), 0, "n=%zu", n);
253 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, L'A', &s)), (size_t)-1,
254 "n=%zu", n);
255 ATF_CHECK_EQ_MSG(errno, EILSEQ, "errno=%d", errno);
256 ATF_CHECK_EQ_MSG((unsigned char)buf[0], 0xcc, "buf=[%02x]", buf[0]);
257
258 /* Invalid code; 'Pile of poo' without the lead surrogate. */
259 memset(&s, 0, sizeof(s));
260 memset(buf, 0xcc, sizeof(buf));
261 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xdca9, &s)), (size_t)-1,
262 "n=%zu", n);
263 ATF_CHECK_EQ_MSG(errno, EILSEQ, "errno=%d", errno);
264 ATF_CHECK_EQ_MSG((unsigned char)buf[0], 0xcc, "buf=[%02x]", buf[0]);
265
266 /* Incomplete Unicode character 'Pile of poo', interrupted by NUL. */
267 memset(&s, 0, sizeof(s));
268 memset(buf, 0xcc, sizeof(buf));
269 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, 0xd83d, &s)), 0, "n=%zu", n);
270 ATF_CHECK_EQ_MSG((n = c16rtomb(buf, L'\0', &s)), 1,
271 "n=%zu", n);
272 ATF_CHECK_MSG(((unsigned char)buf[0] == '\0' &&
273 (unsigned char)buf[1] == 0xcc),
274 "buf=[%02x %02x]", buf[0], buf[1]);
275 }
276
277 ATF_TP_ADD_TCS(tp)
278 {
279
280 ATF_TP_ADD_TC(tp, c16rtomb_c_locale_test);
281 ATF_TP_ADD_TC(tp, c16rtomb_iso2022jp_locale_test);
282 ATF_TP_ADD_TC(tp, c16rtomb_iso_8859_1_test);
283 ATF_TP_ADD_TC(tp, c16rtomb_iso_8859_15_test);
284 ATF_TP_ADD_TC(tp, c16rtomb_utf_8_test);
285
286 return (atf_no_error());
287 }
288