1 /*- 2 * Copyright (c) 2011 Michihiro NAKAJIMA 3 * All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 8 * 1. Redistributions of source code must retain the above copyright 9 * notice, this list of conditions and the following disclaimer. 10 * 2. Redistributions in binary form must reproduce the above copyright 11 * notice, this list of conditions and the following disclaimer in the 12 * documentation and/or other materials provided with the distribution. 13 * 14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR 15 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 16 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 17 * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT, 18 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 19 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 20 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 21 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 22 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 23 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 24 */ 25 #include "test.h" 26 __FBSDID("$FreeBSD"); 27 28 #include <locale.h> 29 30 static void 31 test_read_format_ustar_filename_eucJP_UTF8(const char *refname) 32 { 33 struct archive *a; 34 struct archive_entry *ae; 35 36 /* 37 * Read eucJP filename in en_US.UTF-8 with "hdrcharset=eucJP" option. 38 */ 39 if (NULL == setlocale(LC_ALL, "en_US.UTF-8")) { 40 skipping("en_US.UTF-8 locale not available on this system."); 41 return; 42 } 43 44 assert((a = archive_read_new()) != NULL); 45 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 46 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 47 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=eucJP")) { 48 skipping("This system cannot convert character-set" 49 " from eucJP to UTF-8."); 50 goto cleanup; 51 } 52 assertEqualIntA(a, ARCHIVE_OK, 53 archive_read_open_filename(a, refname, 10240)); 54 55 /* Verify regular file. */ 56 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 57 assertEqualString("\xe6\xbc\xa2\xe5\xad\x97.txt", 58 archive_entry_pathname(ae)); 59 assertEqualInt(8, archive_entry_size(ae)); 60 61 /* Verify regular file. */ 62 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 63 assertEqualString("\xe8\xa1\xa8.txt", archive_entry_pathname(ae)); 64 assertEqualInt(4, archive_entry_size(ae)); 65 66 67 /* End of archive. */ 68 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 69 70 /* Verify archive format. */ 71 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 72 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 73 74 /* Close the archive. */ 75 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 76 cleanup: 77 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 78 } 79 80 static void 81 test_read_format_ustar_filename_CP866_KOI8R(const char *refname) 82 { 83 struct archive *a; 84 struct archive_entry *ae; 85 86 /* 87 * Read CP866 filename in ru_RU.KOI8-R with "hdrcharset=CP866" option. 88 */ 89 if (NULL == setlocale(LC_ALL, "Russian_Russia.20866") && 90 NULL == setlocale(LC_ALL, "ru_RU.KOI8-R")) { 91 skipping("ru_RU.KOI8-R locale not available on this system."); 92 return; 93 } 94 95 assert((a = archive_read_new()) != NULL); 96 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 97 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 98 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=CP866")) { 99 skipping("This system cannot convert character-set" 100 " from CP866 to KOI8-R."); 101 goto cleanup; 102 } 103 assertEqualIntA(a, ARCHIVE_OK, 104 archive_read_open_filename(a, refname, 10240)); 105 106 /* Verify regular file. */ 107 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 108 assertEqualString("\xf0\xf2\xe9\xf7\xe5\xf4", 109 archive_entry_pathname(ae)); 110 assertEqualInt(6, archive_entry_size(ae)); 111 112 /* Verify regular file. */ 113 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 114 assertEqualString("\xd0\xd2\xc9\xd7\xc5\xd4", 115 archive_entry_pathname(ae)); 116 assertEqualInt(6, archive_entry_size(ae)); 117 118 119 /* End of archive. */ 120 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 121 122 /* Verify archive format. */ 123 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 124 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 125 126 /* Close the archive. */ 127 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 128 cleanup: 129 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 130 } 131 132 static void 133 test_read_format_ustar_filename_CP866_UTF8(const char *refname) 134 { 135 struct archive *a; 136 struct archive_entry *ae; 137 138 /* 139 * Read CP866 filename in en_US.UTF-8 with "hdrcharset=CP866" option. 140 */ 141 if (NULL == setlocale(LC_ALL, "en_US.UTF-8")) { 142 skipping("en_US.UTF-8 locale not available on this system."); 143 return; 144 } 145 146 assert((a = archive_read_new()) != NULL); 147 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 148 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 149 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=CP866")) { 150 skipping("This system cannot convert character-set" 151 " from CP866 to UTF-8."); 152 goto cleanup; 153 } 154 assertEqualIntA(a, ARCHIVE_OK, 155 archive_read_open_filename(a, refname, 10240)); 156 157 /* Verify regular file. */ 158 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 159 assertEqualString("\xd0\x9f\xd0\xa0\xd0\x98\xd0\x92\xd0\x95\xd0\xa2", 160 archive_entry_pathname(ae)); 161 assertEqualInt(6, archive_entry_size(ae)); 162 163 /* Verify regular file. */ 164 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 165 assertEqualString("\xd0\xbf\xd1\x80\xd0\xb8\xd0\xb2\xd0\xb5\xd1\x82", 166 archive_entry_pathname(ae)); 167 assertEqualInt(6, archive_entry_size(ae)); 168 169 170 /* End of archive. */ 171 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 172 173 /* Verify archive format. */ 174 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 175 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 176 177 /* Close the archive. */ 178 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 179 cleanup: 180 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 181 } 182 183 static void 184 test_read_format_ustar_filename_KOI8R_CP866(const char *refname) 185 { 186 struct archive *a; 187 struct archive_entry *ae; 188 189 /* 190 * Read KOI8-R filename in ru_RU.CP866 with "hdrcharset=KOI8-R" option. 191 */ 192 if (NULL == setlocale(LC_ALL, "Russian_Russia.866") && 193 NULL == setlocale(LC_ALL, "ru_RU.CP866")) { 194 skipping("ru_RU.CP866 locale not available on this system."); 195 return; 196 } 197 198 assert((a = archive_read_new()) != NULL); 199 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 200 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 201 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=KOI8-R")) { 202 skipping("This system cannot convert character-set" 203 " from KOI8-R to CP866."); 204 goto cleanup; 205 } 206 assertEqualIntA(a, ARCHIVE_OK, 207 archive_read_open_filename(a, refname, 10240)); 208 209 /* Verify regular file. */ 210 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 211 assertEqualString("\xaf\xe0\xa8\xa2\xa5\xe2", 212 archive_entry_pathname(ae)); 213 assertEqualInt(6, archive_entry_size(ae)); 214 215 /* Verify regular file. */ 216 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 217 assertEqualString("\x8f\x90\x88\x82\x85\x92", 218 archive_entry_pathname(ae)); 219 assertEqualInt(6, archive_entry_size(ae)); 220 221 222 /* End of archive. */ 223 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 224 225 /* Verify archive format. */ 226 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 227 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 228 229 /* Close the archive. */ 230 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 231 cleanup: 232 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 233 } 234 235 static void 236 test_read_format_ustar_filename_KOI8R_UTF8(const char *refname) 237 { 238 struct archive *a; 239 struct archive_entry *ae; 240 241 /* 242 * Read KOI8-R filename in en_US.UTF-8 with "hdrcharset=KOI8-R" option. 243 */ 244 if (NULL == setlocale(LC_ALL, "en_US.UTF-8")) { 245 skipping("en_US.UTF-8 locale not available on this system."); 246 return; 247 } 248 249 assert((a = archive_read_new()) != NULL); 250 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 251 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 252 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=KOI8-R")) { 253 skipping("This system cannot convert character-set" 254 " from KOI8-R to UTF-8."); 255 goto cleanup; 256 } 257 assertEqualIntA(a, ARCHIVE_OK, 258 archive_read_open_filename(a, refname, 10240)); 259 260 /* Verify regular file. */ 261 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 262 assertEqualString("\xd0\xbf\xd1\x80\xd0\xb8\xd0\xb2\xd0\xb5\xd1\x82", 263 archive_entry_pathname(ae)); 264 assertEqualInt(6, archive_entry_size(ae)); 265 266 /* Verify regular file. */ 267 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 268 assertEqualString("\xd0\x9f\xd0\xa0\xd0\x98\xd0\x92\xd0\x95\xd0\xa2", 269 archive_entry_pathname(ae)); 270 assertEqualInt(6, archive_entry_size(ae)); 271 272 273 /* End of archive. */ 274 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 275 276 /* Verify archive format. */ 277 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 278 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 279 280 /* Close the archive. */ 281 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 282 cleanup: 283 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 284 } 285 286 static void 287 test_read_format_ustar_filename_eucJP_CP932(const char *refname) 288 { 289 struct archive *a; 290 struct archive_entry *ae; 291 292 /* 293 * Read eucJP filename in CP932/SJIS with "hdrcharset=eucJP" option. 294 */ 295 if (NULL == setlocale(LC_ALL, "Japanese_Japan") && 296 NULL == setlocale(LC_ALL, "ja_JP.SJIS")) { 297 skipping("CP932 locale not available on this system."); 298 return; 299 } 300 301 assert((a = archive_read_new()) != NULL); 302 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 303 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 304 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=eucJP")) { 305 skipping("This system cannot convert character-set" 306 " from eucJP."); 307 goto cleanup; 308 } 309 assertEqualIntA(a, ARCHIVE_OK, 310 archive_read_open_filename(a, refname, 10240)); 311 312 /* Verify regular file. */ 313 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 314 assertEqualString("\x8a\xbf\x8e\x9a.txt", archive_entry_pathname(ae)); 315 assertEqualInt(8, archive_entry_size(ae)); 316 317 /* Verify regular file. */ 318 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 319 assertEqualString("\x95\x5c.txt", archive_entry_pathname(ae)); 320 assertEqualInt(4, archive_entry_size(ae)); 321 322 323 /* End of archive. */ 324 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 325 326 /* Verify archive format. */ 327 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 328 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 329 330 /* Close the archive. */ 331 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 332 cleanup: 333 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 334 } 335 336 static void 337 test_read_format_ustar_filename_CP866_CP1251(const char *refname) 338 { 339 struct archive *a; 340 struct archive_entry *ae; 341 342 /* 343 * Read CP866 filename in CP1251 with "hdrcharset=CP866" option. 344 */ 345 if (NULL == setlocale(LC_ALL, "Russian_Russia") && 346 NULL == setlocale(LC_ALL, "ru_RU.CP1251")) { 347 skipping("CP1251 locale not available on this system."); 348 return; 349 } 350 351 assert((a = archive_read_new()) != NULL); 352 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 353 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 354 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=CP866")) { 355 skipping("This system cannot convert character-set" 356 " from CP866 to CP1251."); 357 goto cleanup; 358 } 359 assertEqualIntA(a, ARCHIVE_OK, 360 archive_read_open_filename(a, refname, 10240)); 361 362 /* Verify regular file. */ 363 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 364 assertEqualString("\xcf\xd0\xc8\xc2\xc5\xd2", 365 archive_entry_pathname(ae)); 366 assertEqualInt(6, archive_entry_size(ae)); 367 368 /* Verify regular file. */ 369 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 370 assertEqualString("\xef\xf0\xe8\xe2\xe5\xf2", 371 archive_entry_pathname(ae)); 372 assertEqualInt(6, archive_entry_size(ae)); 373 374 375 /* End of archive. */ 376 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 377 378 /* Verify archive format. */ 379 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 380 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 381 382 /* Close the archive. */ 383 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 384 cleanup: 385 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 386 } 387 388 /* 389 * This test only for Windows platform because other archiver 390 * applications on Windows translate CP1251 filenames into CP866 391 * filenames and store it in the ustar file and so we should read 392 * it by default on Windows. 393 */ 394 static void 395 test_read_format_ustar_filename_CP866_CP1251_win(const char *refname) 396 { 397 struct archive *a; 398 struct archive_entry *ae; 399 400 /* 401 * Read CP866 filename in CP1251 without "hdrcharset=CP866" option. 402 */ 403 if (NULL == setlocale(LC_ALL, "Russian_Russia")) { 404 skipping("Russian_Russia locale not available on this system."); 405 return; 406 } 407 408 assert((a = archive_read_new()) != NULL); 409 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 410 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 411 assertEqualIntA(a, ARCHIVE_OK, 412 archive_read_open_filename(a, refname, 10240)); 413 414 /* Verify regular file. */ 415 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 416 assertEqualString("\xcf\xd0\xc8\xc2\xc5\xd2", 417 archive_entry_pathname(ae)); 418 assertEqualInt(6, archive_entry_size(ae)); 419 420 /* Verify regular file. */ 421 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 422 assertEqualString("\xef\xf0\xe8\xe2\xe5\xf2", 423 archive_entry_pathname(ae)); 424 assertEqualInt(6, archive_entry_size(ae)); 425 426 427 /* End of archive. */ 428 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 429 430 /* Verify archive format. */ 431 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 432 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 433 434 /* Close the archive. */ 435 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 436 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 437 } 438 439 static void 440 test_read_format_ustar_filename_KOI8R_CP1251(const char *refname) 441 { 442 struct archive *a; 443 struct archive_entry *ae; 444 445 /* 446 * Read KOI8-R filename in CP1251 with "hdrcharset=KOI8-R" option. 447 */ 448 if (NULL == setlocale(LC_ALL, "Russian_Russia") && 449 NULL == setlocale(LC_ALL, "ru_RU.CP1251")) { 450 skipping("CP1251 locale not available on this system."); 451 return; 452 } 453 454 assert((a = archive_read_new()) != NULL); 455 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_filter_all(a)); 456 assertEqualIntA(a, ARCHIVE_OK, archive_read_support_format_all(a)); 457 if (ARCHIVE_OK != archive_read_set_options(a, "hdrcharset=KOI8-R")) { 458 skipping("This system cannot convert character-set" 459 " from KOI8-R to CP1251."); 460 goto cleanup; 461 } 462 assertEqualIntA(a, ARCHIVE_OK, 463 archive_read_open_filename(a, refname, 10240)); 464 465 /* Verify regular file. */ 466 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 467 assertEqualString("\xef\xf0\xe8\xe2\xe5\xf2", 468 archive_entry_pathname(ae)); 469 assertEqualInt(6, archive_entry_size(ae)); 470 471 /* Verify regular file. */ 472 assertEqualIntA(a, ARCHIVE_OK, archive_read_next_header(a, &ae)); 473 assertEqualString("\xcf\xd0\xc8\xc2\xc5\xd2", 474 archive_entry_pathname(ae)); 475 assertEqualInt(6, archive_entry_size(ae)); 476 477 478 /* End of archive. */ 479 assertEqualIntA(a, ARCHIVE_EOF, archive_read_next_header(a, &ae)); 480 481 /* Verify archive format. */ 482 assertEqualIntA(a, ARCHIVE_FILTER_COMPRESS, archive_filter_code(a, 0)); 483 assertEqualIntA(a, ARCHIVE_FORMAT_TAR_USTAR, archive_format(a)); 484 485 /* Close the archive. */ 486 assertEqualInt(ARCHIVE_OK, archive_read_close(a)); 487 cleanup: 488 assertEqualInt(ARCHIVE_OK, archive_read_free(a)); 489 } 490 491 492 DEFINE_TEST(test_read_format_ustar_filename) 493 { 494 const char *refname1 = "test_read_format_ustar_filename_eucjp.tar.Z"; 495 const char *refname2 = "test_read_format_ustar_filename_cp866.tar.Z"; 496 const char *refname3 = "test_read_format_ustar_filename_koi8r.tar.Z"; 497 498 extract_reference_file(refname1); 499 test_read_format_ustar_filename_eucJP_UTF8(refname1); 500 test_read_format_ustar_filename_eucJP_CP932(refname1); 501 502 extract_reference_file(refname2); 503 test_read_format_ustar_filename_CP866_KOI8R(refname2); 504 test_read_format_ustar_filename_CP866_UTF8(refname2); 505 test_read_format_ustar_filename_CP866_CP1251(refname2); 506 test_read_format_ustar_filename_CP866_CP1251_win(refname2); 507 508 extract_reference_file(refname3); 509 test_read_format_ustar_filename_KOI8R_CP866(refname3); 510 test_read_format_ustar_filename_KOI8R_UTF8(refname3); 511 test_read_format_ustar_filename_KOI8R_CP1251(refname3); 512 } 513