usr.bin/cut/cut.c

   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  * Adam S. Moskowitz of Menlo Consulting and Marciano Pitargue.
   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  * 3. 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  * @(#) Copyright (c) 1989, 1993 The Regents of the University of California.  All rights reserved.
  33  * @(#)cut.c    8.3 (Berkeley) 5/4/95
  34  * $FreeBSD: head/usr.bin/cut/cut.c 243474 2012-11-24 04:15:25Z andrew $
  35  */
  36
  37 #include <ctype.h>
  38 #include <err.h>
  39 #include <errno.h>
  40 #include <limits.h>
  41 #include <locale.h>
  42 #include <stdio.h>
  43 #include <stdlib.h>
  44 #include <string.h>
  45 #include <unistd.h>
  46 #include <wchar.h>
  47
  48 static int      bflag;
  49 static int      cflag;
  50 static wchar_t  dchar;
  51 static char     dcharmb[MB_LEN_MAX + 1];
  52 static int      dflag;
  53 static int      fflag;
  54 static int      nflag;
  55 static int      sflag;
  56 static int      wflag;
  57
  58 static size_t   autostart, autostop, maxval;
  59 static char *   positions;
  60
  61 static int      b_cut(FILE *, const char *);
  62 static int      b_n_cut(FILE *, const char *);
  63 static int      c_cut(FILE *, const char *);
  64 static int      f_cut(FILE *, const char *);
  65 static void     get_list(char *);
  66 static int      is_delim(wchar_t);
  67 static void     needpos(size_t);
  68 static void     usage(void);
  69
  70 int
  71 main(int argc, char *argv[])
  72 {
  73         FILE *fp;
  74         int (*fcn)(FILE *, const char *);
  75         int ch, rval;
  76         size_t n;
  77
  78         setlocale(LC_ALL, "");
  79
  80         fcn = NULL;
  81         dchar = '\t';                   /* default delimiter is \t */
  82         strcpy(dcharmb, "\t");
  83
  84         while ((ch = getopt(argc, argv, "b:c:d:f:snw")) != -1)
  85                 switch(ch) {
  86                 case 'b':
  87                         get_list(optarg);
  88                         bflag = 1;
  89                         break;
  90                 case 'c':
  91                         get_list(optarg);
  92                         cflag = 1;
  93                         break;
  94                 case 'd':
  95                         n = mbrtowc(&dchar, optarg, MB_LEN_MAX, NULL);
  96                         if (dchar == '\0' || n != strlen(optarg))
  97                                 errx(1, "bad delimiter");
  98                         strcpy(dcharmb, optarg);
  99                         dflag = 1;
 100                         break;
 101                 case 'f':
 102                         get_list(optarg);
 103                         fflag = 1;
 104                         break;
 105                 case 's':
 106                         sflag = 1;
 107                         break;
 108                 case 'n':
 109                         nflag = 1;
 110                         break;
 111                 case 'w':
 112                         wflag = 1;
 113                         break;
 114                 case '?':
 115                 default:
 116                         usage();
 117                 }
 118         argc -= optind;
 119         argv += optind;
 120
 121         if (fflag) {
 122                 if (bflag || cflag || nflag || (wflag && dflag))
 123                         usage();
 124         } else if (!(bflag || cflag) || dflag || sflag || wflag)
 125                 usage();
 126         else if (!bflag && nflag)
 127                 usage();
 128
 129         if (fflag)
 130                 fcn = f_cut;
 131         else if (cflag)
 132                 fcn = MB_CUR_MAX > 1 ? c_cut : b_cut;
 133         else if (bflag)
 134                 fcn = nflag && MB_CUR_MAX > 1 ? b_n_cut : b_cut;
 135
 136         rval = 0;
 137         if (*argv)
 138                 for (; *argv; ++argv) {
 139                         if (strcmp(*argv, "-") == 0)
 140                                 rval |= fcn(stdin, "stdin");
 141                         else {
 142                                 if (!(fp = fopen(*argv, "r"))) {
 143                                         warn("%s", *argv);
 144                                         rval = 1;
 145                                         continue;
 146                                 }
 147                                 fcn(fp, *argv);
 148                                 (void)fclose(fp);
 149                         }
 150                 }
 151         else
 152                 rval = fcn(stdin, "stdin");
 153         exit(rval);
 154 }
 155
 156 static void
 157 get_list(char *list)
 158 {
 159         size_t setautostart, start, stop;
 160         char *pos;
 161         char *p;
 162
 163         /*
 164          * set a byte in the positions array to indicate if a field or
 165          * column is to be selected; use +1, it's 1-based, not 0-based.
 166          * Numbers and number ranges may be overlapping, repeated, and in
 167          * any order. We handle "-3-5" although there's no real reason to.
 168          */
 169         for (; (p = strsep(&list, ", \t")) != NULL;) {
 170                 setautostart = start = stop = 0;
 171                 if (*p == '-') {
 172                         ++p;
 173                         setautostart = 1;
 174                 }
 175                 if (isdigit((unsigned char)*p)) {
 176                         start = stop = strtol(p, &p, 10);
 177                         if (setautostart && start > autostart)
 178                                 autostart = start;
 179                 }
 180                 if (*p == '-') {
 181                         if (isdigit((unsigned char)p[1]))
 182                                 stop = strtol(p + 1, &p, 10);
 183                         if (*p == '-') {
 184                                 ++p;
 185                                 if (!autostop || autostop > stop)
 186                                         autostop = stop;
 187                         }
 188                 }
 189                 if (*p)
 190                         errx(1, "[-bcf] list: illegal list value");
 191                 if (!stop || !start)
 192                         errx(1, "[-bcf] list: values may not include zero");
 193                 if (maxval < stop) {
 194                         maxval = stop;
 195                         needpos(maxval + 1);
 196                 }
 197                 for (pos = positions + start; start++ <= stop; *pos++ = 1);
 198         }
 199
 200         /* overlapping ranges */
 201         if (autostop && maxval > autostop) {
 202                 maxval = autostop;
 203                 needpos(maxval + 1);
 204         }
 205
 206         /* set autostart */
 207         if (autostart)
 208                 memset(positions + 1, '1', autostart);
 209 }
 210
 211 static void
 212 needpos(size_t n)
 213 {
 214         static size_t npos;
 215         size_t oldnpos;
 216
 217         /* Grow the positions array to at least the specified size. */
 218         if (n > npos) {
 219                 oldnpos = npos;
 220                 if (npos == 0)
 221                         npos = n;
 222                 while (n > npos)
 223                         npos *= 2;
 224                 if ((positions = realloc(positions, npos)) == NULL)
 225                         err(1, "realloc");
 226                 memset((char *)positions + oldnpos, 0, npos - oldnpos);
 227         }
 228 }
 229
 230 static int
 231 b_cut(FILE *fp, const char *fname __unused)
 232 {
 233         int ch, col;
 234         char *pos;
 235
 236         ch = 0;
 237         for (;;) {
 238                 pos = positions + 1;
 239                 for (col = maxval; col; --col) {
 240                         if ((ch = getc(fp)) == EOF)
 241                                 return (0);
 242                         if (ch == '\n')
 243                                 break;
 244                         if (*pos++)
 245                                 (void)putchar(ch);
 246                 }
 247                 if (ch != '\n') {
 248                         if (autostop)
 249                                 while ((ch = getc(fp)) != EOF && ch != '\n')
 250                                         (void)putchar(ch);
 251                         else
 252                                 while ((ch = getc(fp)) != EOF && ch != '\n');
 253                 }
 254                 (void)putchar('\n');
 255         }
 256         return (0);
 257 }
 258
 259 /*
 260  * Cut based on byte positions, taking care not to split multibyte characters.
 261  * Although this function also handles the case where -n is not specified,
 262  * b_cut() ought to be much faster.
 263  */
 264 static int
 265 b_n_cut(FILE *fp, const char *fname)
 266 {
 267         size_t col, i, lbuflen;
 268         char *lbuf;
 269         int canwrite, clen, warned;
 270         mbstate_t mbs;
 271
 272         memset(&mbs, 0, sizeof(mbs));
 273         warned = 0;
 274         while ((lbuf = fgetln(fp, &lbuflen)) != NULL) {
 275                 for (col = 0; lbuflen > 0; col += clen) {
 276                         if ((clen = mbrlen(lbuf, lbuflen, &mbs)) < 0) {
 277                                 if (!warned) {
 278                                         warn("%s", fname);
 279                                         warned = 1;
 280                                 }
 281                                 memset(&mbs, 0, sizeof(mbs));
 282                                 clen = 1;
 283                         }
 284                         if (clen == 0 || *lbuf == '\n')
 285                                 break;
 286                         if (col < maxval && !positions[1 + col]) {
 287                                 /*
 288                                  * Print the character if (1) after an initial
 289                                  * segment of un-selected bytes, the rest of
 290                                  * it is selected, and (2) the last byte is
 291                                  * selected.
 292                                  */
 293                                 i = col;
 294                                 while (i < col + clen && i < maxval &&
 295                                     !positions[1 + i])
 296                                         i++;
 297                                 canwrite = i < col + clen;
 298                                 for (; i < col + clen && i < maxval; i++)
 299                                         canwrite &= positions[1 + i];
 300                                 if (canwrite)
 301                                         fwrite(lbuf, 1, clen, stdout);
 302                         } else {
 303                                 /*
 304                                  * Print the character if all of it has
 305                                  * been selected.
 306                                  */
 307                                 canwrite = 1;
 308                                 for (i = col; i < col + clen; i++)
 309                                         if ((i >= maxval && !autostop) ||
 310                                             (i < maxval && !positions[1 + i])) {
 311                                                 canwrite = 0;
 312                                                 break;
 313                                         }
 314                                 if (canwrite)
 315                                         fwrite(lbuf, 1, clen, stdout);
 316                         }
 317                         lbuf += clen;
 318                         lbuflen -= clen;
 319                 }
 320                 if (lbuflen > 0)
 321                         putchar('\n');
 322         }
 323         return (warned);
 324 }
 325
 326 static int
 327 c_cut(FILE *fp, const char *fname)
 328 {
 329         wint_t ch;
 330         int col;
 331         char *pos;
 332
 333         ch = 0;
 334         for (;;) {
 335                 pos = positions + 1;
 336                 for (col = maxval; col; --col) {
 337                         if ((ch = getwc(fp)) == WEOF)
 338                                 goto out;
 339                         if (ch == '\n')
 340                                 break;
 341                         if (*pos++)
 342                                 (void)putwchar(ch);
 343                 }
 344                 if (ch != '\n') {
 345                         if (autostop)
 346                                 while ((ch = getwc(fp)) != WEOF && ch != '\n')
 347                                         (void)putwchar(ch);
 348                         else
 349                                 while ((ch = getwc(fp)) != WEOF && ch != '\n');
 350                 }
 351                 (void)putwchar('\n');
 352         }
 353 out:
 354         if (ferror(fp)) {
 355                 warn("%s", fname);
 356                 return (1);
 357         }
 358         return (0);
 359 }
 360
 361 static int
 362 is_delim(wchar_t ch)
 363 {
 364         if (wflag) {
 365                 if (ch == ' ' || ch == '\t')
 366                         return 1;
 367         } else {
 368                 if (ch == dchar)
 369                         return 1;
 370         }
 371         return 0;
 372 }
 373
 374 static int
 375 f_cut(FILE *fp, const char *fname)
 376 {
 377         wchar_t ch;
 378         int field, i, isdelim;
 379         char *pos, *p;
 380         int output;
 381         char *lbuf, *mlbuf;
 382         size_t clen, lbuflen, reallen;
 383
 384         mlbuf = NULL;
 385         while ((lbuf = fgetln(fp, &lbuflen)) != NULL) {
 386                 reallen = lbuflen;
 387                 /* Assert EOL has a newline. */
 388                 if (*(lbuf + lbuflen - 1) != '\n') {
 389                         /* Can't have > 1 line with no trailing newline. */
 390                         mlbuf = malloc(lbuflen + 1);
 391                         if (mlbuf == NULL)
 392                                 err(1, "malloc");
 393                         memcpy(mlbuf, lbuf, lbuflen);
 394                         *(mlbuf + lbuflen) = '\n';
 395                         lbuf = mlbuf;
 396                         reallen++;
 397                 }
 398                 output = 0;
 399                 for (isdelim = 0, p = lbuf;; p += clen) {
 400                         clen = mbrtowc(&ch, p, lbuf + reallen - p, NULL);
 401                         if (clen == (size_t)-1 || clen == (size_t)-2) {
 402                                 warnc(EILSEQ, "%s", fname);
 403                                 free(mlbuf);
 404                                 return (1);
 405                         }
 406                         if (clen == 0)
 407                                 clen = 1;
 408                         /* this should work if newline is delimiter */
 409                         if (is_delim(ch))
 410                                 isdelim = 1;
 411                         if (ch == '\n') {
 412                                 if (!isdelim && !sflag)
 413                                         (void)fwrite(lbuf, lbuflen, 1, stdout);
 414                                 break;
 415                         }
 416                 }
 417                 if (!isdelim)
 418                         continue;
 419
 420                 pos = positions + 1;
 421                 for (field = maxval, p = lbuf; field; --field, ++pos) {
 422                         if (*pos && output++)
 423                                 for (i = 0; dcharmb[i] != '\0'; i++)
 424                                         putchar(dcharmb[i]);
 425                         for (;;) {
 426                                 clen = mbrtowc(&ch, p, lbuf + reallen - p,
 427                                     NULL);
 428                                 if (clen == (size_t)-1 || clen == (size_t)-2) {
 429                                         warnc(EILSEQ, "%s", fname);
 430                                         free(mlbuf);
 431                                         return (1);
 432                                 }
 433                                 if (clen == 0)
 434                                         clen = 1;
 435                                 p += clen;
 436                                 if (ch == '\n' || is_delim(ch)) {
 437                                         /* compress whitespace */
 438                                         if (wflag && ch != '\n')
 439                                                 while (is_delim(*p))
 440                                                         p++;
 441                                         break;
 442                                 }
 443                                 if (*pos)
 444                                         for (i = 0; i < (int)clen; i++)
 445                                                 putchar(p[i - clen]);
 446                         }
 447                         if (ch == '\n')
 448                                 break;
 449                 }
 450                 if (ch != '\n') {
 451                         if (autostop) {
 452                                 if (output)
 453                                         for (i = 0; dcharmb[i] != '\0'; i++)
 454                                                 putchar(dcharmb[i]);
 455                                 for (; (ch = *p) != '\n'; ++p)
 456                                         (void)putchar(ch);
 457                         } else
 458                                 for (; (ch = *p) != '\n'; ++p);
 459                 }
 460                 (void)putchar('\n');
 461         }
 462         free(mlbuf);
 463         return (0);
 464 }
 465
 466 static void
 467 usage(void)
 468 {
 469         (void)fprintf(stderr, "%s\n%s\n%s\n",
 470                 "usage: cut -b list [-n] [file ...]",
 471                 "       cut -c list [file ...]",
 472                 "       cut -f list [-s] [-w | -d delim] [file ...]");
 473         exit(1);
 474 }