| 1 | /* Test of conversion of wide character to multibyte character.
|
|---|
| 2 | Copyright (C) 2008-2012 Free Software Foundation, Inc.
|
|---|
| 3 |
|
|---|
| 4 | This program is free software: you can redistribute it and/or modify
|
|---|
| 5 | it under the terms of the GNU General Public License as published by
|
|---|
| 6 | the Free Software Foundation; either version 3 of the License, or
|
|---|
| 7 | (at your option) any later version.
|
|---|
| 8 |
|
|---|
| 9 | This program is distributed in the hope that it will be useful,
|
|---|
| 10 | but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 11 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 12 | GNU General Public License for more details.
|
|---|
| 13 |
|
|---|
| 14 | You should have received a copy of the GNU General Public License
|
|---|
| 15 | along with this program. If not, see <http://www.gnu.org/licenses/>. */
|
|---|
| 16 |
|
|---|
| 17 | #include <config.h>
|
|---|
| 18 |
|
|---|
| 19 | #include <wchar.h>
|
|---|
| 20 |
|
|---|
| 21 | #include <locale.h>
|
|---|
| 22 | #include <stdio.h>
|
|---|
| 23 | #include <stdlib.h>
|
|---|
| 24 | #include <string.h>
|
|---|
| 25 |
|
|---|
| 26 | #include "macros.h"
|
|---|
| 27 |
|
|---|
| 28 | #if (defined _WIN32 || defined __WIN32__) && !defined __CYGWIN__
|
|---|
| 29 |
|
|---|
| 30 | static int
|
|---|
| 31 | test_one_locale (const char *name, int codepage)
|
|---|
| 32 | {
|
|---|
| 33 | char buf[64];
|
|---|
| 34 | size_t ret;
|
|---|
| 35 |
|
|---|
| 36 | # if 1
|
|---|
| 37 | /* Portable code to set the locale. */
|
|---|
| 38 | {
|
|---|
| 39 | char name_with_codepage[1024];
|
|---|
| 40 |
|
|---|
| 41 | sprintf (name_with_codepage, "%s.%d", name, codepage);
|
|---|
| 42 |
|
|---|
| 43 | /* Set the locale. */
|
|---|
| 44 | if (setlocale (LC_ALL, name_with_codepage) == NULL)
|
|---|
| 45 | return 77;
|
|---|
| 46 | }
|
|---|
| 47 | # else
|
|---|
| 48 | /* Hacky way to set a locale.codepage combination that setlocale() refuses
|
|---|
| 49 | to set. */
|
|---|
| 50 | {
|
|---|
| 51 | /* Codepage of the current locale, set with setlocale().
|
|---|
| 52 | Not necessarily the same as GetACP(). */
|
|---|
| 53 | extern __declspec(dllimport) unsigned int __lc_codepage;
|
|---|
| 54 |
|
|---|
| 55 | /* Set the locale. */
|
|---|
| 56 | if (setlocale (LC_ALL, name) == NULL)
|
|---|
| 57 | return 77;
|
|---|
| 58 |
|
|---|
| 59 | /* Clobber the codepage and MB_CUR_MAX, both set by setlocale(). */
|
|---|
| 60 | __lc_codepage = codepage;
|
|---|
| 61 | switch (codepage)
|
|---|
| 62 | {
|
|---|
| 63 | case 1252:
|
|---|
| 64 | case 1256:
|
|---|
| 65 | MB_CUR_MAX = 1;
|
|---|
| 66 | break;
|
|---|
| 67 | case 932:
|
|---|
| 68 | case 950:
|
|---|
| 69 | case 936:
|
|---|
| 70 | MB_CUR_MAX = 2;
|
|---|
| 71 | break;
|
|---|
| 72 | case 54936:
|
|---|
| 73 | case 65001:
|
|---|
| 74 | MB_CUR_MAX = 4;
|
|---|
| 75 | break;
|
|---|
| 76 | }
|
|---|
| 77 |
|
|---|
| 78 | /* Test whether the codepage is really available. */
|
|---|
| 79 | {
|
|---|
| 80 | mbstate_t state;
|
|---|
| 81 | wchar_t wc;
|
|---|
| 82 |
|
|---|
| 83 | memset (&state, '\0', sizeof (mbstate_t));
|
|---|
| 84 | if (mbrtowc (&wc, " ", 1, &state) == (size_t)(-1))
|
|---|
| 85 | return 77;
|
|---|
| 86 | }
|
|---|
| 87 | }
|
|---|
| 88 | # endif
|
|---|
| 89 |
|
|---|
| 90 | /* Test NUL character. */
|
|---|
| 91 | {
|
|---|
| 92 | buf[0] = 'x';
|
|---|
| 93 | ret = wcrtomb (buf, 0, NULL);
|
|---|
| 94 | ASSERT (ret == 1);
|
|---|
| 95 | ASSERT (buf[0] == '\0');
|
|---|
| 96 | }
|
|---|
| 97 |
|
|---|
| 98 | /* Test single bytes. */
|
|---|
| 99 | {
|
|---|
| 100 | int c;
|
|---|
| 101 |
|
|---|
| 102 | for (c = 0; c < 0x100; c++)
|
|---|
| 103 | switch (c)
|
|---|
| 104 | {
|
|---|
| 105 | case '\t': case '\v': case '\f':
|
|---|
| 106 | case ' ': case '!': case '"': case '#': case '%':
|
|---|
| 107 | case '&': case '\'': case '(': case ')': case '*':
|
|---|
| 108 | case '+': case ',': case '-': case '.': case '/':
|
|---|
| 109 | case '0': case '1': case '2': case '3': case '4':
|
|---|
| 110 | case '5': case '6': case '7': case '8': case '9':
|
|---|
| 111 | case ':': case ';': case '<': case '=': case '>':
|
|---|
| 112 | case '?':
|
|---|
| 113 | case 'A': case 'B': case 'C': case 'D': case 'E':
|
|---|
| 114 | case 'F': case 'G': case 'H': case 'I': case 'J':
|
|---|
| 115 | case 'K': case 'L': case 'M': case 'N': case 'O':
|
|---|
| 116 | case 'P': case 'Q': case 'R': case 'S': case 'T':
|
|---|
| 117 | case 'U': case 'V': case 'W': case 'X': case 'Y':
|
|---|
| 118 | case 'Z':
|
|---|
| 119 | case '[': case '\\': case ']': case '^': case '_':
|
|---|
| 120 | case 'a': case 'b': case 'c': case 'd': case 'e':
|
|---|
| 121 | case 'f': case 'g': case 'h': case 'i': case 'j':
|
|---|
| 122 | case 'k': case 'l': case 'm': case 'n': case 'o':
|
|---|
| 123 | case 'p': case 'q': case 'r': case 's': case 't':
|
|---|
| 124 | case 'u': case 'v': case 'w': case 'x': case 'y':
|
|---|
| 125 | case 'z': case '{': case '|': case '}': case '~':
|
|---|
| 126 | /* c is in the ISO C "basic character set". */
|
|---|
| 127 | ret = wcrtomb (buf, btowc (c), NULL);
|
|---|
| 128 | ASSERT (ret == 1);
|
|---|
| 129 | ASSERT (buf[0] == (char) c);
|
|---|
| 130 | break;
|
|---|
| 131 | }
|
|---|
| 132 | }
|
|---|
| 133 |
|
|---|
| 134 | /* Test special calling convention, passing a NULL pointer. */
|
|---|
| 135 | {
|
|---|
| 136 | ret = wcrtomb (NULL, '\0', NULL);
|
|---|
| 137 | ASSERT (ret == 1);
|
|---|
| 138 | ret = wcrtomb (NULL, btowc ('x'), NULL);
|
|---|
| 139 | ASSERT (ret == 1);
|
|---|
| 140 | }
|
|---|
| 141 |
|
|---|
| 142 | switch (codepage)
|
|---|
| 143 | {
|
|---|
| 144 | case 1252:
|
|---|
| 145 | /* Locale encoding is CP1252, an extension of ISO-8859-1. */
|
|---|
| 146 | {
|
|---|
| 147 | /* Convert "B\374\337er": "Büßer" */
|
|---|
| 148 | memset (buf, 'x', 8);
|
|---|
| 149 | ret = wcrtomb (buf, 0x00FC, NULL);
|
|---|
| 150 | ASSERT (ret == 1);
|
|---|
| 151 | ASSERT (memcmp (buf, "\374", 1) == 0);
|
|---|
| 152 | ASSERT (buf[1] == 'x');
|
|---|
| 153 |
|
|---|
| 154 | memset (buf, 'x', 8);
|
|---|
| 155 | ret = wcrtomb (buf, 0x00DF, NULL);
|
|---|
| 156 | ASSERT (ret == 1);
|
|---|
| 157 | ASSERT (memcmp (buf, "\337", 1) == 0);
|
|---|
| 158 | ASSERT (buf[1] == 'x');
|
|---|
| 159 | }
|
|---|
| 160 | return 0;
|
|---|
| 161 |
|
|---|
| 162 | case 1256:
|
|---|
| 163 | /* Locale encoding is CP1256, not the same as ISO-8859-6. */
|
|---|
| 164 | {
|
|---|
| 165 | /* Convert "x\302\341\346y": "xآلوy" */
|
|---|
| 166 | memset (buf, 'x', 8);
|
|---|
| 167 | ret = wcrtomb (buf, 0x0622, NULL);
|
|---|
| 168 | ASSERT (ret == 1);
|
|---|
| 169 | ASSERT (memcmp (buf, "\302", 1) == 0);
|
|---|
| 170 | ASSERT (buf[1] == 'x');
|
|---|
| 171 |
|
|---|
| 172 | memset (buf, 'x', 8);
|
|---|
| 173 | ret = wcrtomb (buf, 0x0644, NULL);
|
|---|
| 174 | ASSERT (ret == 1);
|
|---|
| 175 | ASSERT (memcmp (buf, "\341", 1) == 0);
|
|---|
| 176 | ASSERT (buf[1] == 'x');
|
|---|
| 177 |
|
|---|
| 178 | memset (buf, 'x', 8);
|
|---|
| 179 | ret = wcrtomb (buf, 0x0648, NULL);
|
|---|
| 180 | ASSERT (ret == 1);
|
|---|
| 181 | ASSERT (memcmp (buf, "\346", 1) == 0);
|
|---|
| 182 | ASSERT (buf[1] == 'x');
|
|---|
| 183 | }
|
|---|
| 184 | return 0;
|
|---|
| 185 |
|
|---|
| 186 | case 932:
|
|---|
| 187 | /* Locale encoding is CP932, similar to Shift_JIS. */
|
|---|
| 188 | {
|
|---|
| 189 | /* Convert "<\223\372\226\173\214\352>": "<日本語>" */
|
|---|
| 190 | memset (buf, 'x', 8);
|
|---|
| 191 | ret = wcrtomb (buf, 0x65E5, NULL);
|
|---|
| 192 | ASSERT (ret == 2);
|
|---|
| 193 | ASSERT (memcmp (buf, "\223\372", 2) == 0);
|
|---|
| 194 | ASSERT (buf[2] == 'x');
|
|---|
| 195 |
|
|---|
| 196 | memset (buf, 'x', 8);
|
|---|
| 197 | ret = wcrtomb (buf, 0x672C, NULL);
|
|---|
| 198 | ASSERT (ret == 2);
|
|---|
| 199 | ASSERT (memcmp (buf, "\226\173", 2) == 0);
|
|---|
| 200 | ASSERT (buf[2] == 'x');
|
|---|
| 201 |
|
|---|
| 202 | memset (buf, 'x', 8);
|
|---|
| 203 | ret = wcrtomb (buf, 0x8A9E, NULL);
|
|---|
| 204 | ASSERT (ret == 2);
|
|---|
| 205 | ASSERT (memcmp (buf, "\214\352", 2) == 0);
|
|---|
| 206 | ASSERT (buf[2] == 'x');
|
|---|
| 207 | }
|
|---|
| 208 | return 0;
|
|---|
| 209 |
|
|---|
| 210 | case 950:
|
|---|
| 211 | /* Locale encoding is CP950, similar to Big5. */
|
|---|
| 212 | {
|
|---|
| 213 | /* Convert "<\244\351\245\273\273\171>": "<日本語>" */
|
|---|
| 214 | memset (buf, 'x', 8);
|
|---|
| 215 | ret = wcrtomb (buf, 0x65E5, NULL);
|
|---|
| 216 | ASSERT (ret == 2);
|
|---|
| 217 | ASSERT (memcmp (buf, "\244\351", 2) == 0);
|
|---|
| 218 | ASSERT (buf[2] == 'x');
|
|---|
| 219 |
|
|---|
| 220 | memset (buf, 'x', 8);
|
|---|
| 221 | ret = wcrtomb (buf, 0x672C, NULL);
|
|---|
| 222 | ASSERT (ret == 2);
|
|---|
| 223 | ASSERT (memcmp (buf, "\245\273", 2) == 0);
|
|---|
| 224 | ASSERT (buf[2] == 'x');
|
|---|
| 225 |
|
|---|
| 226 | memset (buf, 'x', 8);
|
|---|
| 227 | ret = wcrtomb (buf, 0x8A9E, NULL);
|
|---|
| 228 | ASSERT (ret == 2);
|
|---|
| 229 | ASSERT (memcmp (buf, "\273\171", 2) == 0);
|
|---|
| 230 | ASSERT (buf[2] == 'x');
|
|---|
| 231 | }
|
|---|
| 232 | return 0;
|
|---|
| 233 |
|
|---|
| 234 | case 936:
|
|---|
| 235 | /* Locale encoding is CP936 = GBK, an extension of GB2312. */
|
|---|
| 236 | {
|
|---|
| 237 | /* Convert "<\310\325\261\276\325\132>": "<日本語>" */
|
|---|
| 238 | memset (buf, 'x', 8);
|
|---|
| 239 | ret = wcrtomb (buf, 0x65E5, NULL);
|
|---|
| 240 | ASSERT (ret == 2);
|
|---|
| 241 | ASSERT (memcmp (buf, "\310\325", 2) == 0);
|
|---|
| 242 | ASSERT (buf[2] == 'x');
|
|---|
| 243 |
|
|---|
| 244 | memset (buf, 'x', 8);
|
|---|
| 245 | ret = wcrtomb (buf, 0x672C, NULL);
|
|---|
| 246 | ASSERT (ret == 2);
|
|---|
| 247 | ASSERT (memcmp (buf, "\261\276", 2) == 0);
|
|---|
| 248 | ASSERT (buf[2] == 'x');
|
|---|
| 249 |
|
|---|
| 250 | memset (buf, 'x', 8);
|
|---|
| 251 | ret = wcrtomb (buf, 0x8A9E, NULL);
|
|---|
| 252 | ASSERT (ret == 2);
|
|---|
| 253 | ASSERT (memcmp (buf, "\325\132", 2) == 0);
|
|---|
| 254 | ASSERT (buf[2] == 'x');
|
|---|
| 255 | }
|
|---|
| 256 | return 0;
|
|---|
| 257 |
|
|---|
| 258 | case 54936:
|
|---|
| 259 | /* Locale encoding is CP54936 = GB18030. */
|
|---|
| 260 | {
|
|---|
| 261 | /* Convert "B\250\271\201\060\211\070er": "Büßer" */
|
|---|
| 262 | memset (buf, 'x', 8);
|
|---|
| 263 | ret = wcrtomb (buf, 0x00FC, NULL);
|
|---|
| 264 | ASSERT (ret == 2);
|
|---|
| 265 | ASSERT (memcmp (buf, "\250\271", 2) == 0);
|
|---|
| 266 | ASSERT (buf[2] == 'x');
|
|---|
| 267 |
|
|---|
| 268 | memset (buf, 'x', 8);
|
|---|
| 269 | ret = wcrtomb (buf, 0x00DF, NULL);
|
|---|
| 270 | ASSERT (ret == 4);
|
|---|
| 271 | ASSERT (memcmp (buf, "\201\060\211\070", 4) == 0);
|
|---|
| 272 | ASSERT (buf[4] == 'x');
|
|---|
| 273 | }
|
|---|
| 274 | return 0;
|
|---|
| 275 |
|
|---|
| 276 | case 65001:
|
|---|
| 277 | /* Locale encoding is CP65001 = UTF-8. */
|
|---|
| 278 | {
|
|---|
| 279 | /* Convert "B\303\274\303\237er": "Büßer" */
|
|---|
| 280 | memset (buf, 'x', 8);
|
|---|
| 281 | ret = wcrtomb (buf, 0x00FC, NULL);
|
|---|
| 282 | ASSERT (ret == 2);
|
|---|
| 283 | ASSERT (memcmp (buf, "\303\274", 2) == 0);
|
|---|
| 284 | ASSERT (buf[2] == 'x');
|
|---|
| 285 |
|
|---|
| 286 | memset (buf, 'x', 8);
|
|---|
| 287 | ret = wcrtomb (buf, 0x00DF, NULL);
|
|---|
| 288 | ASSERT (ret == 2);
|
|---|
| 289 | ASSERT (memcmp (buf, "\303\237", 2) == 0);
|
|---|
| 290 | ASSERT (buf[2] == 'x');
|
|---|
| 291 | }
|
|---|
| 292 | return 0;
|
|---|
| 293 |
|
|---|
| 294 | default:
|
|---|
| 295 | return 1;
|
|---|
| 296 | }
|
|---|
| 297 | }
|
|---|
| 298 |
|
|---|
| 299 | int
|
|---|
| 300 | main (int argc, char *argv[])
|
|---|
| 301 | {
|
|---|
| 302 | int codepage = atoi (argv[argc - 1]);
|
|---|
| 303 | int result;
|
|---|
| 304 | int i;
|
|---|
| 305 |
|
|---|
| 306 | result = 77;
|
|---|
| 307 | for (i = 1; i < argc - 1; i++)
|
|---|
| 308 | {
|
|---|
| 309 | int ret = test_one_locale (argv[i], codepage);
|
|---|
| 310 |
|
|---|
| 311 | if (ret != 77)
|
|---|
| 312 | result = ret;
|
|---|
| 313 | }
|
|---|
| 314 |
|
|---|
| 315 | if (result == 77)
|
|---|
| 316 | {
|
|---|
| 317 | fprintf (stderr, "Skipping test: found no locale with codepage %d\n",
|
|---|
| 318 | codepage);
|
|---|
| 319 | }
|
|---|
| 320 | return result;
|
|---|
| 321 | }
|
|---|
| 322 |
|
|---|
| 323 | #else
|
|---|
| 324 |
|
|---|
| 325 | int
|
|---|
| 326 | main (int argc, char *argv[])
|
|---|
| 327 | {
|
|---|
| 328 | fputs ("Skipping test: not a native Windows system\n", stderr);
|
|---|
| 329 | return 77;
|
|---|
| 330 | }
|
|---|
| 331 |
|
|---|
| 332 | #endif
|
|---|