1 /* 2 * Copyright (c) 1989, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * This code is derived from software contributed to Berkeley by 6 * Case Larsen. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 4. Neither the name of the University nor the names of its contributors 17 * may be used to endorse or promote products derived from this software 18 * without specific prior written permission. 19 * 20 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 23 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 30 * SUCH DAMAGE. 31 */ 32 33 #ifndef lint 34 static const char copyright[] = 35 "@(#) Copyright (c) 1989, 1993\n\ 36 The Regents of the University of California. All rights reserved.\n"; 37 #endif /* not lint */ 38 39 #ifndef lint 40 #if 0 41 static char sccsid[] = "@(#)uniq.c 8.3 (Berkeley) 5/4/95"; 42 #endif 43 static const char rcsid[] = 44 "$FreeBSD$"; 45 #endif /* not lint */ 46 47 #include <sys/capsicum.h> 48 49 #include <ctype.h> 50 #include <err.h> 51 #include <errno.h> 52 #include <limits.h> 53 #include <locale.h> 54 #include <nl_types.h> 55 #include <stdint.h> 56 #include <stdio.h> 57 #include <stdlib.h> 58 #include <string.h> 59 #include <termios.h> 60 #include <unistd.h> 61 #include <wchar.h> 62 #include <wctype.h> 63 64 static int cflag, dflag, uflag, iflag; 65 static int numchars, numfields, repeats; 66 67 static FILE *file(const char *, const char *); 68 static wchar_t *convert(const char *); 69 static int inlcmp(const char *, const char *); 70 static void show(FILE *, const char *); 71 static wchar_t *skip(wchar_t *); 72 static void obsolete(char *[]); 73 static void usage(void); 74 75 static void 76 strerror_init(void) 77 { 78 79 /* 80 * Cache NLS data before entering capability mode. 81 * XXXPJD: There should be strerror_init() and strsignal_init() in libc. 82 */ 83 (void)catopen("libc", NL_CAT_LOCALE); 84 } 85 86 int 87 main (int argc, char *argv[]) 88 { 89 wchar_t *tprev, *tthis; 90 FILE *ifp, *ofp; 91 int ch, comp; 92 size_t prevbuflen, thisbuflen, b1; 93 char *prevline, *thisline, *p; 94 const char *ifn; 95 cap_rights_t rights; 96 97 (void) setlocale(LC_ALL, ""); 98 99 obsolete(argv); 100 while ((ch = getopt(argc, argv, "cdif:s:u")) != -1) 101 switch (ch) { 102 case 'c': 103 cflag = 1; 104 break; 105 case 'd': 106 dflag = 1; 107 break; 108 case 'i': 109 iflag = 1; 110 break; 111 case 'f': 112 numfields = strtol(optarg, &p, 10); 113 if (numfields < 0 || *p) 114 errx(1, "illegal field skip value: %s", optarg); 115 break; 116 case 's': 117 numchars = strtol(optarg, &p, 10); 118 if (numchars < 0 || *p) 119 errx(1, "illegal character skip value: %s", optarg); 120 break; 121 case 'u': 122 uflag = 1; 123 break; 124 case '?': 125 default: 126 usage(); 127 } 128 129 argc -= optind; 130 argv += optind; 131 132 /* If no flags are set, default is -d -u. */ 133 if (cflag) { 134 if (dflag || uflag) 135 usage(); 136 } else if (!dflag && !uflag) 137 dflag = uflag = 1; 138 139 if (argc > 2) 140 usage(); 141 142 ifp = stdin; 143 ifn = "stdin"; 144 ofp = stdout; 145 if (argc > 0 && strcmp(argv[0], "-") != 0) 146 ifp = file(ifn = argv[0], "r"); 147 cap_rights_init(&rights, CAP_FSTAT, CAP_READ); 148 if (cap_rights_limit(fileno(ifp), &rights) < 0 && errno != ENOSYS) 149 err(1, "unable to limit rights for %s", ifn); 150 cap_rights_init(&rights, CAP_FSTAT, CAP_WRITE); 151 if (argc > 1) 152 ofp = file(argv[1], "w"); 153 else 154 cap_rights_set(&rights, CAP_IOCTL); 155 if (cap_rights_limit(fileno(ofp), &rights) < 0 && errno != ENOSYS) { 156 err(1, "unable to limit rights for %s", 157 argc > 1 ? argv[1] : "stdout"); 158 } 159 if (cap_rights_is_set(&rights, CAP_IOCTL)) { 160 unsigned long cmd; 161 162 cmd = TIOCGETA; /* required by isatty(3) in printf(3) */ 163 164 if (cap_ioctls_limit(fileno(ofp), &cmd, 1) < 0 && 165 errno != ENOSYS) { 166 err(1, "unable to limit ioctls for %s", 167 argc > 1 ? argv[1] : "stdout"); 168 } 169 } 170 171 strerror_init(); 172 if (cap_enter() < 0 && errno != ENOSYS) 173 err(1, "unable to enter capability mode"); 174 175 prevbuflen = thisbuflen = 0; 176 prevline = thisline = NULL; 177 178 if (getline(&prevline, &prevbuflen, ifp) < 0) { 179 if (ferror(ifp)) 180 err(1, "%s", ifn); 181 exit(0); 182 } 183 tprev = convert(prevline); 184 185 if (!cflag && uflag && dflag) 186 show(ofp, prevline); 187 188 tthis = NULL; 189 while (getline(&thisline, &thisbuflen, ifp) >= 0) { 190 if (tthis != NULL) 191 free(tthis); 192 tthis = convert(thisline); 193 194 if (tthis == NULL && tprev == NULL) 195 comp = inlcmp(thisline, prevline); 196 else if (tthis == NULL || tprev == NULL) 197 comp = 1; 198 else 199 comp = wcscoll(tthis, tprev); 200 201 if (comp) { 202 /* If different, print; set previous to new value. */ 203 if (cflag || !dflag || !uflag) 204 show(ofp, prevline); 205 p = prevline; 206 b1 = prevbuflen; 207 prevline = thisline; 208 prevbuflen = thisbuflen; 209 if (tprev != NULL) 210 free(tprev); 211 tprev = tthis; 212 if (!cflag && uflag && dflag) 213 show(ofp, prevline); 214 thisline = p; 215 thisbuflen = b1; 216 tthis = NULL; 217 repeats = 0; 218 } else 219 ++repeats; 220 } 221 if (ferror(ifp)) 222 err(1, "%s", ifn); 223 if (cflag || !dflag || !uflag) 224 show(ofp, prevline); 225 exit(0); 226 } 227 228 static wchar_t * 229 convert(const char *str) 230 { 231 size_t n; 232 wchar_t *buf, *ret, *p; 233 234 if ((n = mbstowcs(NULL, str, 0)) == (size_t)-1) 235 return (NULL); 236 if (SIZE_MAX / sizeof(*buf) < n + 1) 237 errx(1, "conversion buffer length overflow"); 238 if ((buf = malloc((n + 1) * sizeof(*buf))) == NULL) 239 err(1, "malloc"); 240 if (mbstowcs(buf, str, n + 1) != n) 241 errx(1, "internal mbstowcs() error"); 242 /* The last line may not end with \n. */ 243 if (n > 0 && buf[n - 1] == L'\n') 244 buf[n - 1] = L'\0'; 245 246 /* If requested get the chosen fields + character offsets. */ 247 if (numfields || numchars) { 248 if ((ret = wcsdup(skip(buf))) == NULL) 249 err(1, "wcsdup"); 250 free(buf); 251 } else 252 ret = buf; 253 254 if (iflag) { 255 for (p = ret; *p != L'\0'; p++) 256 *p = towlower(*p); 257 } 258 259 return (ret); 260 } 261 262 static int 263 inlcmp(const char *s1, const char *s2) 264 { 265 int c1, c2; 266 267 while (*s1 == *s2++) 268 if (*s1++ == '\0') 269 return (0); 270 c1 = (unsigned char)*s1; 271 c2 = (unsigned char)*(s2 - 1); 272 /* The last line may not end with \n. */ 273 if (c1 == '\n') 274 c1 = '\0'; 275 if (c2 == '\n') 276 c2 = '\0'; 277 return (c1 - c2); 278 } 279 280 /* 281 * show -- 282 * Output a line depending on the flags and number of repetitions 283 * of the line. 284 */ 285 static void 286 show(FILE *ofp, const char *str) 287 { 288 289 if (cflag) 290 (void)fprintf(ofp, "%4d %s", repeats + 1, str); 291 if ((dflag && repeats) || (uflag && !repeats)) 292 (void)fprintf(ofp, "%s", str); 293 } 294 295 static wchar_t * 296 skip(wchar_t *str) 297 { 298 int nchars, nfields; 299 300 for (nfields = 0; *str != L'\0' && nfields++ != numfields; ) { 301 while (iswblank(*str)) 302 str++; 303 while (*str != L'\0' && !iswblank(*str)) 304 str++; 305 } 306 for (nchars = numchars; nchars-- && *str != L'\0'; ++str) 307 ; 308 return(str); 309 } 310 311 static FILE * 312 file(const char *name, const char *mode) 313 { 314 FILE *fp; 315 316 if ((fp = fopen(name, mode)) == NULL) 317 err(1, "%s", name); 318 return(fp); 319 } 320 321 static void 322 obsolete(char *argv[]) 323 { 324 int len; 325 char *ap, *p, *start; 326 327 while ((ap = *++argv)) { 328 /* Return if "--" or not an option of any form. */ 329 if (ap[0] != '-') { 330 if (ap[0] != '+') 331 return; 332 } else if (ap[1] == '-') 333 return; 334 if (!isdigit((unsigned char)ap[1])) 335 continue; 336 /* 337 * Digit signifies an old-style option. Malloc space for dash, 338 * new option and argument. 339 */ 340 len = strlen(ap); 341 if ((start = p = malloc(len + 3)) == NULL) 342 err(1, "malloc"); 343 *p++ = '-'; 344 *p++ = ap[0] == '+' ? 's' : 'f'; 345 (void)strcpy(p, ap + 1); 346 *argv = start; 347 } 348 } 349 350 static void 351 usage(void) 352 { 353 (void)fprintf(stderr, 354 "usage: uniq [-c | -d | -u] [-i] [-f fields] [-s chars] [input [output]]\n"); 355 exit(1); 356 } 357