xargs/xargs.c

   1 /*-
   2  * Copyright (c) 1990, 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  * John B. Roll Jr.
   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. All advertising materials mentioning features or use of this software
  17  *    must display the following acknowledgement:
  18  *      This product includes software developed by the University of
  19  *      California, Berkeley and its contributors.
  20  * 4. Neither the name of the University nor the names of its contributors
  21  *    may be used to endorse or promote products derived from this software
  22  *    without specific prior written permission.
  23  *
  24  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  25  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  27  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  28  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  29  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  30  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  31  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  32  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  33  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  34  * SUCH DAMAGE.
  35  *
  36  * $xMach: xargs.c,v 1.6 2002/02/23 05:27:47 tim Exp $
  37  */
  38
  39 #ifndef lint
  40 static const char copyright[] =
  41 "@(#) Copyright (c) 1990, 1993\n\
  42         The Regents of the University of California.  All rights reserved.\n";
  43 #endif /* not lint */
  44
  45 #if 0
  46 #ifndef lint
  47 static char sccsid[] = "@(#)xargs.c     8.1 (Berkeley) 6/6/93";
  48 #endif /* not lint */
  49 #endif
  50
  51 #include <sys/cdefs.h>
  52 __RCSID("$FreeBSD: src/usr.bin/xargs/xargs.c,v 1.41 2002/07/01 03:21:05 tjr Exp $");
  53
  54 #include <sys/types.h>
  55 #include <sys/wait.h>
  56
  57 #include <err.h>
  58 #include <errno.h>
  59 #include <locale.h>
  60 #include <paths.h>
  61 #include <regex.h>
  62 #include <stdio.h>
  63 #include <stdlib.h>
  64 #include <string.h>
  65 #include <unistd.h>
  66
  67 #include "pathnames.h"
  68
  69 static void     parse_input(int, char *[]);
  70 static void     prerun(int, char *[]);
  71 static int      prompt(void);
  72 static void     run(char **);
  73 static void     usage(void);
  74 void            strnsubst(char **, const char *, const char *, size_t);
  75
  76 static char echo[] = _PATH_ECHO;
  77 static char **av, **bxp, **ep, **exp, **xp;
  78 static char *argp, *bbp, *ebp, *inpline, *p, *replstr;
  79 static const char *eofstr;
  80 static int count, insingle, indouble, pflag, tflag, Rflag, rval, zflag;
  81 static int cnt, Iflag, jfound, Lflag, wasquoted, xflag;
  82
  83 extern char **environ;
  84
  85 int
  86 main(int argc, char *argv[])
  87 {
  88         long arg_max;
  89         int ch, Jflag, nargs, nflag, nline;
  90         size_t linelen;
  91
  92         inpline = replstr = NULL;
  93         ep = environ;
  94         eofstr = "";
  95         Jflag = nflag = 0;
  96
  97         /*
  98          * POSIX.2 limits the exec line length to ARG_MAX - 2K.  Running that
  99          * caused some E2BIG errors, so it was changed to ARG_MAX - 4K.  Given
 100          * that the smallest argument is 2 bytes in length, this means that
 101          * the number of arguments is limited to:
 102          *
 103          *       (ARG_MAX - 4K - LENGTH(utility + arguments)) / 2.
 104          *
 105          * We arbitrarily limit the number of arguments to 5000.  This is
 106          * allowed by POSIX.2 as long as the resulting minimum exec line is
 107          * at least LINE_MAX.  Realloc'ing as necessary is possible, but
 108          * probably not worthwhile.
 109          */
 110         nargs = 5000;
 111         if ((arg_max = sysconf(_SC_ARG_MAX)) == -1)
 112                 errx(1, "sysconf(_SC_ARG_MAX) failed");
 113         nline = arg_max - 4 * 1024;
 114         while (*ep != NULL) {
 115                 /* 1 byte for each '\0' */
 116                 nline -= strlen(*ep++) + 1 + sizeof(*ep);
 117         }
 118         while ((ch = getopt(argc, argv, "0E:I:J:L:n:pR:s:tx")) != -1)
 119                 switch(ch) {
 120                 case 'E':
 121                         eofstr = optarg;
 122                         break;
 123                 case 'I':
 124                         Jflag = 0;
 125                         Iflag = 1;
 126                         Lflag = 1;
 127                         replstr = optarg;
 128                         break;
 129                 case 'J':
 130                         Iflag = 0;
 131                         Jflag = 1;
 132                         replstr = optarg;
 133                         break;
 134                 case 'L':
 135                         Lflag = atoi(optarg);
 136                         break;
 137                 case 'n':
 138                         nflag = 1;
 139                         if ((nargs = atoi(optarg)) <= 0)
 140                                 errx(1, "illegal argument count");
 141                         break;
 142                 case 'p':
 143                         pflag = 1;
 144                         break;
 145                 case 'R':
 146                         if ((Rflag = atoi(optarg)) <= 0)
 147                                 errx(1, "illegal number of replacements");
 148                         break;
 149                 case 's':
 150                         nline = atoi(optarg);
 151                         break;
 152                 case 't':
 153                         tflag = 1;
 154                         break;
 155                 case 'x':
 156                         xflag = 1;
 157                         break;
 158                 case '0':
 159                         zflag = 1;
 160                         break;
 161                 case '?':
 162                 default:
 163                         usage();
 164         }
 165         argc -= optind;
 166         argv += optind;
 167
 168         if (!Iflag && Rflag)
 169                 usage();
 170         if (Iflag && !Rflag)
 171                 Rflag = 5;
 172         if (xflag && !nflag)
 173                 usage();
 174         if (Iflag || Lflag)
 175                 xflag = 1;
 176         if (replstr != NULL && *replstr == '\0')
 177                 errx(1, "replstr may not be empty");
 178
 179         /*
 180          * Allocate pointers for the utility name, the utility arguments,
 181          * the maximum arguments to be read from stdin and the trailing
 182          * NULL.
 183          */
 184         linelen = 1 + argc + nargs + 1;
 185         if ((av = bxp = malloc(linelen * sizeof(char **))) == NULL)
 186                 errx(1, "malloc failed");
 187
 188         /*
 189          * Use the user's name for the utility as argv[0], just like the
 190          * shell.  Echo is the default.  Set up pointers for the user's
 191          * arguments.
 192          */
 193         if (*argv == NULL)
 194                 cnt = strlen(*bxp++ = echo);
 195         else {
 196                 do {
 197                         if (Jflag && strcmp(*argv, replstr) == 0) {
 198                                 char **avj;
 199                                 jfound = 1;
 200                                 argv++;
 201                                 for (avj = argv; *avj; avj++)
 202                                         cnt += strlen(*avj) + 1;
 203                                 break;
 204                         }
 205                         cnt += strlen(*bxp++ = *argv) + 1;
 206                 } while (*++argv != NULL);
 207         }
 208
 209         /*
 210          * Set up begin/end/traversing pointers into the array.  The -n
 211          * count doesn't include the trailing NULL pointer, so the malloc
 212          * added in an extra slot.
 213          */
 214         exp = (xp = bxp) + nargs;
 215
 216         /*
 217          * Allocate buffer space for the arguments read from stdin and the
 218          * trailing NULL.  Buffer space is defined as the default or specified
 219          * space, minus the length of the utility name and arguments.  Set up
 220          * begin/end/traversing pointers into the array.  The -s count does
 221          * include the trailing NULL, so the malloc didn't add in an extra
 222          * slot.
 223          */
 224         nline -= cnt;
 225         if (nline <= 0)
 226                 errx(1, "insufficient space for command");
 227
 228         if ((bbp = malloc((size_t)(nline + 1))) == NULL)
 229                 errx(1, "malloc failed");
 230         ebp = (argp = p = bbp) + nline - 1;
 231         for (;;)
 232                 parse_input(argc, argv);
 233 }
 234
 235 static void
 236 parse_input(int argc, char *argv[])
 237 {
 238         int ch, foundeof;
 239         char **avj;
 240
 241         foundeof = 0;
 242
 243         switch(ch = getchar()) {
 244         case EOF:
 245                 /* No arguments since last exec. */
 246                 if (p == bbp)
 247                         exit(rval);
 248                 goto arg1;
 249         case ' ':
 250         case '\t':
 251                 /* Quotes escape tabs and spaces. */
 252                 if (insingle || indouble || zflag)
 253                         goto addch;
 254                 goto arg2;
 255         case '\0':
 256                 if (zflag)
 257                         goto arg2;
 258                 goto addch;
 259         case '\n':
 260                 count++;
 261                 if (zflag)
 262                         goto addch;
 263
 264                 /* Quotes do not escape newlines. */
 265 arg1:           if (insingle || indouble)
 266                         errx(1, "unterminated quote");
 267 arg2:
 268                 foundeof = *eofstr != '\0' &&
 269                     strcmp(argp, eofstr) == 0;
 270
 271                 /* Do not make empty args unless they are quoted */
 272                 if ((argp != p || wasquoted) && !foundeof) {
 273                         *p++ = '\0';
 274                         *xp++ = argp;
 275                         if (Iflag) {
 276                                 size_t curlen;
 277
 278                                 if (inpline == NULL)
 279                                         curlen = 0;
 280                                 else {
 281                                         /*
 282                                          * If this string is not zero
 283                                          * length, append a space for
 284                                          * seperation before the next
 285                                          * argument.
 286                                          */
 287                                         if ((curlen = strlen(inpline)))
 288                                                 strcat(inpline, " ");
 289                                 }
 290                                 curlen++;
 291                                 /*
 292                                  * Allocate enough to hold what we will
 293                                  * be holding in a second, and to append
 294                                  * a space next time through, if we have
 295                                  * to.
 296                                  */
 297                                 inpline = realloc(inpline, curlen + 2 +
 298                                     strlen(argp));
 299                                 if (inpline == NULL)
 300                                         errx(1, "realloc failed");
 301                                 if (curlen == 1)
 302                                         strcpy(inpline, argp);
 303                                 else
 304                                         strcat(inpline, argp);
 305                         }
 306                 }
 307
 308                 /*
 309                  * If max'd out on args or buffer, or reached EOF,
 310                  * run the command.  If xflag and max'd out on buffer
 311                  * but not on args, object.  Having reached the limit
 312                  * of input lines, as specified by -L is the same as
 313                  * maxing out on arguments.
 314                  */
 315                 if (xp == exp || p > ebp || ch == EOF ||
 316                     (Lflag <= count && xflag) || foundeof) {
 317                         if (xflag && xp != exp && p > ebp)
 318                                 errx(1, "insufficient space for arguments");
 319                         if (jfound) {
 320                                 for (avj = argv; *avj; avj++)
 321                                         *xp++ = *avj;
 322                         }
 323                         prerun(argc, av);
 324                         if (ch == EOF || foundeof)
 325                                 exit(rval);
 326                         p = bbp;
 327                         xp = bxp;
 328                         count = 0;
 329                 }
 330                 argp = p;
 331                 wasquoted = 0;
 332                 break;
 333         case '\'':
 334                 if (indouble || zflag)
 335                         goto addch;
 336                 insingle = !insingle;
 337                 wasquoted = 1;
 338                 break;
 339         case '"':
 340                 if (insingle || zflag)
 341                         goto addch;
 342                 indouble = !indouble;
 343                 wasquoted = 1;
 344                 break;
 345         case '\\':
 346                 if (zflag)
 347                         goto addch;
 348                 /* Backslash escapes anything, is escaped by quotes. */
 349                 if (!insingle && !indouble && (ch = getchar()) == EOF)
 350                         errx(1, "backslash at EOF");
 351                 /* FALLTHROUGH */
 352         default:
 353 addch:          if (p < ebp) {
 354                         *p++ = ch;
 355                         break;
 356                 }
 357
 358                 /* If only one argument, not enough buffer space. */
 359                 if (bxp == xp)
 360                         errx(1, "insufficient space for argument");
 361                 /* Didn't hit argument limit, so if xflag object. */
 362                 if (xflag)
 363                         errx(1, "insufficient space for arguments");
 364
 365                 if (jfound) {
 366                         for (avj = argv; *avj; avj++)
 367                                 *xp++ = *avj;
 368                 }
 369                 prerun(argc, av);
 370                 xp = bxp;
 371                 cnt = ebp - argp;
 372                 memcpy(bbp, argp, (size_t)cnt);
 373                 p = (argp = bbp) + cnt;
 374                 *p++ = ch;
 375                 break;
 376         }
 377         return;
 378 }
 379
 380 /*
 381  * Do things necessary before run()'ing, such as -I substitution,
 382  * and then call run().
 383  */
 384 static void
 385 prerun(int argc, char *argv[])
 386 {
 387         char **tmp, **tmp2, **avj;
 388         int repls;
 389
 390         repls = Rflag;
 391
 392         if (argc == 0 || repls == 0) {
 393                 *xp = NULL;
 394                 run(argv);
 395                 return;
 396         }
 397
 398         avj = argv;
 399
 400         /*
 401          * Allocate memory to hold the argument list, and
 402          * a NULL at the tail.
 403          */
 404         tmp = malloc((argc + 1) * sizeof(char**));
 405         if (tmp == NULL)
 406                 errx(1, "malloc failed");
 407         tmp2 = tmp;
 408
 409         /*
 410          * Save the first argument and iterate over it, we
 411          * cannot do strnsubst() to it.
 412          */
 413         if ((*tmp++ = strdup(*avj++)) == NULL)
 414                 errx(1, "strdup failed");
 415
 416         /*
 417          * For each argument to utility, if we have not used up
 418          * the number of replacements we are allowed to do, and
 419          * if the argument contains at least one occurance of
 420          * replstr, call strnsubst(), else just save the string.
 421          * Iterations over elements of avj and tmp are done
 422          * where appropriate.
 423          */
 424         while (--argc) {
 425                 *tmp = *avj++;
 426                 if (repls && strstr(*tmp, replstr) != NULL) {
 427                         strnsubst(tmp++, replstr, inpline, (size_t)255);
 428                         repls--;
 429                 } else {
 430                         if ((*tmp = strdup(*tmp)) == NULL)
 431                                 errx(1, "strdup failed");
 432                         tmp++;
 433                 }
 434         }
 435
 436         /*
 437          * Run it.
 438          */
 439         *tmp = NULL;
 440         run(tmp2);
 441
 442         /*
 443          * Walk from the tail to the head, free along the way.
 444          */
 445         for (; tmp2 != tmp; tmp--)
 446                 free(*tmp);
 447         /*
 448          * Now free the list itself.
 449          */
 450         free(tmp2);
 451
 452         /*
 453          * Free the input line buffer, if we have one.
 454          */
 455         if (inpline != NULL) {
 456                 free(inpline);
 457                 inpline = NULL;
 458         }
 459 }
 460
 461 static void
 462 run(char **argv)
 463 {
 464         volatile int childerr;
 465         char **avec;
 466         pid_t pid;
 467         int status;
 468
 469         /*
 470          * If the user wants to be notified of each command before it is
 471          * executed, notify them.  If they want the notification to be
 472          * followed by a prompt, then prompt them.
 473          */
 474         if (tflag || pflag) {
 475                 (void)fprintf(stderr, "%s", *argv);
 476                 for (avec = argv + 1; *avec != NULL; ++avec)
 477                         (void)fprintf(stderr, " %s", *avec);
 478                 /*
 479                  * If the user has asked to be prompted, do so.
 480                  */
 481                 if (pflag)
 482                         /*
 483                          * If they asked not to exec, return without execution
 484                          * but if they asked to, go to the execution.  If we
 485                          * could not open their tty, break the switch and drop
 486                          * back to -t behaviour.
 487                          */
 488                         switch (prompt()) {
 489                         case 0:
 490                                 return;
 491                         case 1:
 492                                 goto exec;
 493                         case 2:
 494                                 break;
 495                         }
 496                 (void)fprintf(stderr, "\n");
 497                 (void)fflush(stderr);
 498         }
 499 exec:
 500         childerr = 0;
 501         switch(pid = vfork()) {
 502         case -1:
 503                 err(1, "vfork");
 504         case 0:
 505                 execvp(argv[0], argv);
 506                 childerr = errno;
 507                 _exit(1);
 508         }
 509         pid = waitpid(pid, &status, 0);
 510         if (pid == -1)
 511                 err(1, "waitpid");
 512         /* If we couldn't invoke the utility, exit. */
 513         if (childerr != 0)
 514                 err(childerr == ENOENT ? 127 : 126, "%s", *argv);
 515         /* If utility signaled or exited with a value of 255, exit 1-125. */
 516         if (WIFSIGNALED(status) || WEXITSTATUS(status) == 255)
 517                 exit(1);
 518         if (WEXITSTATUS(status))
 519                 rval = 1;
 520 }
 521
 522 /*
 523  * Prompt the user about running a command.
 524  */
 525 static int
 526 prompt(void)
 527 {
 528         regex_t cre;
 529         size_t rsize;
 530         int match;
 531         char *response;
 532         FILE *ttyfp;
 533
 534         if ((ttyfp = fopen(_PATH_TTY, "r")) == NULL)
 535                 return (2);     /* Indicate that the TTY failed to open. */
 536         (void)fprintf(stderr, "?...");
 537         (void)fflush(stderr);
 538         if ((response = fgetln(ttyfp, &rsize)) == NULL ||
 539             regcomp(&cre,
 540                 "^[yY]",
 541                 REG_BASIC) != 0) {
 542                 (void)fclose(ttyfp);
 543                 return (0);
 544         }
 545         match = regexec(&cre, response, 0, NULL, 0);
 546         (void)fclose(ttyfp);
 547         regfree(&cre);
 548         return (match == 0);
 549 }
 550
 551 static void
 552 usage(void)
 553 {
 554         fprintf(stderr,
 555 "usage: xargs [-0pt] [-E eofstr] [-I replstr [-R replacements]] [-J replstr]\n"
 556 "             [-L number] [-n number [-x] [-s size] [utility [argument ...]]\n");
 557         exit(1);
 558 }