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 #if 0
  40 #ifndef lint
  41 static const char copyright[] =
  42 "@(#) Copyright (c) 1990, 1993\n\
  43         The Regents of the University of California.  All rights reserved.\n";
  44 #endif /* not lint */
  45
  46 #ifndef lint
  47 static char sccsid[] = "@(#)xargs.c     8.1 (Berkeley) 6/6/93";
  48 #endif /* not lint */
  49 #endif
  50 #include <sys/cdefs.h>
  51 __FBSDID("$FreeBSD: src/usr.bin/xargs/xargs.c,v 1.57 2005/02/27 02:01:31 gad Exp $");
  52
  53 #include <sys/param.h>
  54 #include <sys/wait.h>
  55
  56 #include <err.h>
  57 #include <errno.h>
  58 #include <fcntl.h>
  59 #include <langinfo.h>
  60 #include <locale.h>
  61 #include <paths.h>
  62 #include <regex.h>
  63 #include <stdio.h>
  64 #include <stdlib.h>
  65 #include <string.h>
  66 #include <unistd.h>
  67
  68 #include "pathnames.h"
  69
  70 #ifdef __APPLE__
  71 #include <get_compat.h>
  72 #else
  73 #define COMPAT_MODE(a,b) (1)
  74 #endif /* __APPLE__ */
  75
  76 static void     parse_input(int, char *[]);
  77 static void     prerun(int, char *[]);
  78 static int      prompt(void);
  79 static void     run(char **);
  80 static void     usage(void);
  81 void            strnsubst(char **, const char *, const char *, size_t);
  82 static void     waitchildren(const char *, int);
  83
  84 static int last_was_newline = 1;
  85 static int last_was_blank = 0;
  86
  87 static char echo[] = _PATH_ECHO;
  88 static char **av, **bxp, **ep, **endxp, **xp;
  89 static char *argp, *bbp, *ebp, *inpline, *p, *replstr;
  90 static const char *eofstr;
  91 static int count, insingle, indouble, oflag, pflag, tflag, Rflag, rval, zflag;
  92 static int cnt, Iflag, jfound, Lflag, wasquoted, xflag;
  93 static int curprocs, maxprocs;
  94
  95 static volatile int childerr;
  96
  97 extern char **environ;
  98
  99 int
 100 main(int argc, char *argv[])
 101 {
 102         long arg_max;
 103         int ch, Jflag, nargs, nflag, nline;
 104         size_t linelen;
 105         char *endptr;
 106
 107         inpline = replstr = NULL;
 108         ep = environ;
 109         eofstr = "";
 110         Jflag = nflag = 0;
 111
 112         (void)setlocale(LC_ALL, "");
 113
 114         /*
 115          * POSIX.2 limits the exec line length to ARG_MAX - 2K.  Running that
 116          * caused some E2BIG errors, so it was changed to ARG_MAX - 4K.  Given
 117          * that the smallest argument is 2 bytes in length, this means that
 118          * the number of arguments is limited to:
 119          *
 120          *       (ARG_MAX - 4K - LENGTH(utility + arguments)) / 2.
 121          *
 122          * We arbitrarily limit the number of arguments to 5000.  This is
 123          * allowed by POSIX.2 as long as the resulting minimum exec line is
 124          * at least LINE_MAX.  Realloc'ing as necessary is possible, but
 125          * probably not worthwhile.
 126          */
 127         nargs = 5000;
 128         if ((arg_max = sysconf(_SC_ARG_MAX)) == -1)
 129                 errx(1, "sysconf(_SC_ARG_MAX) failed");
 130         nline = arg_max - 4 * 1024;
 131         while (*ep != NULL) {
 132                 /* 1 byte for each '\0' */
 133                 nline -= strlen(*ep++) + 1 + sizeof(*ep);
 134         }
 135         maxprocs = 1;
 136         while ((ch = getopt(argc, argv, "0E:I:J:L:n:oP:pR:s:tx")) != -1)
 137                 switch(ch) {
 138                 case 'E':
 139                         eofstr = optarg;
 140                         break;
 141                 case 'I':
 142                         Jflag = 0;
 143                         Iflag = 1;
 144                         Lflag = 1;
 145                         replstr = optarg;
 146                         break;
 147                 case 'J':
 148                         Iflag = 0;
 149                         Jflag = 1;
 150                         replstr = optarg;
 151                         break;
 152                 case 'L':
 153                         Lflag = atoi(optarg);
 154                         if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 155                                 nflag = 0; /* Override */
 156                                 nargs = 5000;
 157                         }
 158                         break;
 159                 case 'n':
 160                         nflag = 1;
 161                         if ((nargs = atoi(optarg)) <= 0)
 162                                 errx(1, "illegal argument count");
 163                         if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 164                                 Lflag = 0; /* Override */
 165                         }
 166                         break;
 167                 case 'o':
 168                         oflag = 1;
 169                         break;
 170                 case 'P':
 171                         if ((maxprocs = atoi(optarg)) <= 0)
 172                                 errx(1, "max. processes must be >0");
 173                         break;
 174                 case 'p':
 175                         pflag = 1;
 176                         break;
 177                 case 'R':
 178                         Rflag = strtol(optarg, &endptr, 10);
 179                         if (*endptr != '\0')
 180                                 errx(1, "replacements must be a number");
 181                         break;
 182                 case 's':
 183                         nline = atoi(optarg);
 184                         break;
 185                 case 't':
 186                         tflag = 1;
 187                         break;
 188                 case 'x':
 189                         xflag = 1;
 190                         break;
 191                 case '0':
 192                         zflag = 1;
 193                         break;
 194                 case '?':
 195                 default:
 196                         usage();
 197         }
 198         argc -= optind;
 199         argv += optind;
 200
 201         if (!Iflag && Rflag)
 202                 usage();
 203         if (Iflag && !Rflag)
 204                 Rflag = 5;
 205         if (xflag && !nflag)
 206                 usage();
 207         if (Iflag || Lflag)
 208                 xflag = 1;
 209         if (replstr != NULL && *replstr == '\0')
 210                 errx(1, "replstr may not be empty");
 211
 212         /*
 213          * Allocate pointers for the utility name, the utility arguments,
 214          * the maximum arguments to be read from stdin and the trailing
 215          * NULL.
 216          */
 217         linelen = 1 + argc + nargs + 1;
 218         if ((av = bxp = malloc(linelen * sizeof(char **))) == NULL)
 219                 errx(1, "malloc failed");
 220
 221         /*
 222          * Use the user's name for the utility as argv[0], just like the
 223          * shell.  Echo is the default.  Set up pointers for the user's
 224          * arguments.
 225          */
 226         if (*argv == NULL)
 227                 cnt = strlen(*bxp++ = echo);
 228         else {
 229                 do {
 230                         if (Jflag && strcmp(*argv, replstr) == 0) {
 231                                 char **avj;
 232                                 jfound = 1;
 233                                 argv++;
 234                                 for (avj = argv; *avj; avj++)
 235                                         cnt += strlen(*avj) + 1;
 236                                 break;
 237                         }
 238                         cnt += strlen(*bxp++ = *argv) + 1;
 239                 } while (*++argv != NULL);
 240         }
 241
 242         /*
 243          * Set up begin/end/traversing pointers into the array.  The -n
 244          * count doesn't include the trailing NULL pointer, so the malloc
 245          * added in an extra slot.
 246          */
 247         endxp = (xp = bxp) + nargs;
 248
 249         /*
 250          * Allocate buffer space for the arguments read from stdin and the
 251          * trailing NULL.  Buffer space is defined as the default or specified
 252          * space, minus the length of the utility name and arguments.  Set up
 253          * begin/end/traversing pointers into the array.  The -s count does
 254          * include the trailing NULL, so the malloc didn't add in an extra
 255          * slot.
 256          */
 257         nline -= cnt;
 258         if (nline <= 0)
 259                 errx(1, "insufficient space for command");
 260
 261         if ((bbp = malloc((size_t)(nline + 1))) == NULL)
 262                 errx(1, "malloc failed");
 263         ebp = (argp = p = bbp) + nline - 1;
 264         for (;;)
 265                 parse_input(argc, argv);
 266 }
 267
 268 static void
 269 parse_input(int argc, char *argv[])
 270 {
 271         int ch, foundeof;
 272         char **avj;
 273         int last_was_backslashed = 0;
 274
 275         foundeof = 0;
 276
 277         switch(ch = getchar()) {
 278         case EOF:
 279                 /* No arguments since last exec. */
 280                 if (p == bbp) {
 281                         waitchildren(*argv, 1);
 282                         exit(rval);
 283                 }
 284                 goto arg1;
 285         case ' ':
 286                 last_was_blank = 1;
 287         case '\t':
 288                 /* Quotes escape tabs and spaces. */
 289                 if (insingle || indouble || zflag)
 290                         goto addch;
 291                 goto arg2;
 292         case '\0':
 293                 if (zflag) {
 294                         /*
 295                          * Increment 'count', so that nulls will be treated
 296                          * as end-of-line, as well as end-of-argument.  This
 297                          * is needed so -0 works properly with -I and -L.
 298                          */
 299                         count++;
 300                         goto arg2;
 301                 }
 302                 goto addch;
 303         case '\n':
 304                 if (zflag)
 305                         goto addch;
 306                 if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 307                         if (last_was_newline) {
 308                                 /* don't count empty line */
 309                                 break;
 310                         }
 311                         if (!last_was_blank ) {
 312                                 /* only count if NOT continuation line */
 313                                 count++;
 314                         }
 315                 } else {
 316                         count++;
 317                 }
 318                 last_was_newline = 1;
 319
 320                 /* Quotes do not escape newlines. */
 321 arg1:           if (insingle || indouble)
 322                         errx(1, "unterminated quote");
 323 arg2:
 324                 foundeof = *eofstr != '\0' &&
 325                     strcmp(argp, eofstr) == 0;
 326
 327                 /* Do not make empty args unless they are quoted */
 328                 if ((argp != p || wasquoted) && !foundeof) {
 329                         *p++ = '\0';
 330                         *xp++ = argp;
 331                         if (Iflag) {
 332                                 size_t curlen;
 333
 334                                 if (inpline == NULL)
 335                                         curlen = 0;
 336                                 else {
 337                                         /*
 338                                          * If this string is not zero
 339                                          * length, append a space for
 340                                          * separation before the next
 341                                          * argument.
 342                                          */
 343                                         if ((curlen = strlen(inpline)))
 344                                                 strcat(inpline, " ");
 345                                 }
 346                                 curlen++;
 347                                 /*
 348                                  * Allocate enough to hold what we will
 349                                  * be holding in a second, and to append
 350                                  * a space next time through, if we have
 351                                  * to.
 352                                  */
 353                                 inpline = realloc(inpline, curlen + 2 +
 354                                     strlen(argp));
 355                                 if (inpline == NULL)
 356                                         errx(1, "realloc failed");
 357                                 if (curlen == 1)
 358                                         strcpy(inpline, argp);
 359                                 else
 360                                         strcat(inpline, argp);
 361                         }
 362                 }
 363
 364                 /*
 365                  * If max'd out on args or buffer, or reached EOF,
 366                  * run the command.  If xflag and max'd out on buffer
 367                  * but not on args, object.  Having reached the limit
 368                  * of input lines, as specified by -L is the same as
 369                  * maxing out on arguments.
 370                  */
 371                 if (xp == endxp || p > ebp || ch == EOF ||
 372                     (Lflag <= count && xflag) || foundeof) {
 373                         if (xflag && xp != endxp && p > ebp)
 374                                 errx(1, "insufficient space for arguments");
 375                         if (jfound) {
 376                                 for (avj = argv; *avj; avj++)
 377                                         *xp++ = *avj;
 378                         }
 379                         prerun(argc, av);
 380                         if (ch == EOF || foundeof) {
 381                                 waitchildren(*argv, 1);
 382                                 exit(rval);
 383                         }
 384                         p = bbp;
 385                         xp = bxp;
 386                         count = 0;
 387                 }
 388                 argp = p;
 389                 wasquoted = 0;
 390                 break;
 391         case '\'':
 392                 if (indouble || zflag)
 393                         goto addch;
 394                 insingle = !insingle;
 395                 wasquoted = 1;
 396                 break;
 397         case '"':
 398                 if (insingle || zflag)
 399                         goto addch;
 400                 indouble = !indouble;
 401                 wasquoted = 1;
 402                 break;
 403         case '\\':
 404                 last_was_backslashed = 1;
 405                 if (zflag)
 406                         goto addch;
 407                 /* Backslash escapes anything, is escaped by quotes. */
 408                 if (!insingle && !indouble && (ch = getchar()) == EOF)
 409                         errx(1, "backslash at EOF");
 410                 /* FALLTHROUGH */
 411         default:
 412 addch:          if (p < ebp) {
 413                         *p++ = ch;
 414                         break;
 415                 }
 416
 417                 /* If only one argument, not enough buffer space. */
 418                 if (bxp == xp)
 419                         errx(1, "insufficient space for argument");
 420                 /* Didn't hit argument limit, so if xflag object. */
 421                 if (xflag)
 422                         errx(1, "insufficient space for arguments");
 423
 424                 if (jfound) {
 425                         for (avj = argv; *avj; avj++)
 426                                 *xp++ = *avj;
 427                 }
 428                 prerun(argc, av);
 429                 xp = bxp;
 430                 cnt = ebp - argp;
 431                 memcpy(bbp, argp, (size_t)cnt);
 432                 p = (argp = bbp) + cnt;
 433                 *p++ = ch;
 434                 break;
 435         }
 436         if (ch != ' ')
 437                 last_was_blank = 0;
 438         if (ch != '\n' || last_was_backslashed)
 439                 last_was_newline = 0;
 440 }
 441
 442 /*
 443  * Do things necessary before run()'ing, such as -I substitution,
 444  * and then call run().
 445  */
 446 static void
 447 prerun(int argc, char *argv[])
 448 {
 449         char **tmp, **tmp2, **avj;
 450         int repls;
 451
 452         repls = Rflag;
 453
 454         if (argc == 0 || repls == 0) {
 455                 *xp = NULL;
 456                 run(argv);
 457                 return;
 458         }
 459
 460         avj = argv;
 461
 462         /*
 463          * Allocate memory to hold the argument list, and
 464          * a NULL at the tail.
 465          */
 466         tmp = malloc((argc + 1) * sizeof(char**));
 467         if (tmp == NULL)
 468                 errx(1, "malloc failed");
 469         tmp2 = tmp;
 470
 471         /*
 472          * Save the first argument and iterate over it, we
 473          * cannot do strnsubst() to it.
 474          */
 475         if ((*tmp++ = strdup(*avj++)) == NULL)
 476                 errx(1, "strdup failed");
 477
 478         /*
 479          * For each argument to utility, if we have not used up
 480          * the number of replacements we are allowed to do, and
 481          * if the argument contains at least one occurrence of
 482          * replstr, call strnsubst(), else just save the string.
 483          * Iterations over elements of avj and tmp are done
 484          * where appropriate.
 485          */
 486         while (--argc) {
 487                 *tmp = *avj++;
 488                 if (repls && strstr(*tmp, replstr) != NULL) {
 489                         strnsubst(tmp++, replstr, inpline, (size_t)255);
 490                         if (repls > 0)
 491                                 repls--;
 492                 } else {
 493                         if ((*tmp = strdup(*tmp)) == NULL)
 494                                 errx(1, "strdup failed");
 495                         tmp++;
 496                 }
 497         }
 498
 499         /*
 500          * Run it.
 501          */
 502         *tmp = NULL;
 503         run(tmp2);
 504
 505         /*
 506          * Walk from the tail to the head, free along the way.
 507          */
 508         for (; tmp2 != tmp; tmp--)
 509                 free(*tmp);
 510         /*
 511          * Now free the list itself.
 512          */
 513         free(tmp2);
 514
 515         /*
 516          * Free the input line buffer, if we have one.
 517          */
 518         if (inpline != NULL) {
 519                 free(inpline);
 520                 inpline = NULL;
 521         }
 522 }
 523
 524 static void
 525 run(char **argv)
 526 {
 527         pid_t pid;
 528         int fd;
 529         char **avec;
 530
 531         /*
 532          * If the user wants to be notified of each command before it is
 533          * executed, notify them.  If they want the notification to be
 534          * followed by a prompt, then prompt them.
 535          */
 536         if (tflag || pflag) {
 537                 (void)fprintf(stderr, "%s", *argv);
 538                 for (avec = argv + 1; *avec != NULL; ++avec)
 539                         (void)fprintf(stderr, " %s", *avec);
 540                 /*
 541                  * If the user has asked to be prompted, do so.
 542                  */
 543                 if (pflag)
 544                         /*
 545                          * If they asked not to exec, return without execution
 546                          * but if they asked to, go to the execution.  If we
 547                          * could not open their tty, break the switch and drop
 548                          * back to -t behaviour.
 549                          */
 550                         switch (prompt()) {
 551                         case 0:
 552                                 return;
 553                         case 1:
 554                                 goto exec;
 555                         case 2:
 556                                 break;
 557                         }
 558                 (void)fprintf(stderr, "\n");
 559                 (void)fflush(stderr);
 560         }
 561 exec:
 562         childerr = 0;
 563         switch(pid = vfork()) {
 564         case -1:
 565                 err(1, "vfork");
 566         case 0:
 567                 if (oflag) {
 568                         if ((fd = open(_PATH_TTY, O_RDONLY)) == -1)
 569                                 err(1, "can't open /dev/tty");
 570                 } else {
 571                         fd = open(_PATH_DEVNULL, O_RDONLY);
 572                 }
 573                 if (fd > STDIN_FILENO) {
 574                         if (dup2(fd, STDIN_FILENO) != 0)
 575                                 err(1, "can't dup2 to stdin");
 576                         close(fd);
 577                 }
 578                 execvp(argv[0], argv);
 579                 childerr = errno;
 580                 _exit(1);
 581         }
 582         curprocs++;
 583         waitchildren(*argv, 0);
 584 }
 585
 586 static void
 587 waitchildren(const char *name, int waitall)
 588 {
 589         pid_t pid;
 590         int status;
 591
 592         while ((pid = waitpid(-1, &status, !waitall && curprocs < maxprocs ?
 593             WNOHANG : 0)) > 0) {
 594                 curprocs--;
 595                 /* If we couldn't invoke the utility, exit. */
 596                 if (childerr != 0) {
 597                         errno = childerr;
 598                         err(errno == ENOENT ? 127 : 126, "%s", name);
 599                 }
 600                 /*
 601                  * If utility signaled or exited with a value of 255,
 602                  * exit 1-125.
 603                  */
 604                 if (WIFSIGNALED(status) || WEXITSTATUS(status) == 255)
 605                         exit(1);
 606                 if (WEXITSTATUS(status))
 607                         rval = 1;
 608         }
 609         if (pid == -1 && errno != ECHILD)
 610                 err(1, "wait3");
 611 }
 612
 613 /*
 614  * Prompt the user about running a command.
 615  */
 616 static int
 617 prompt(void)
 618 {
 619         regex_t cre;
 620         size_t rsize;
 621         int match;
 622         char *response;
 623         FILE *ttyfp;
 624
 625         if ((ttyfp = fopen(_PATH_TTY, "r")) == NULL)
 626                 return (2);     /* Indicate that the TTY failed to open. */
 627         (void)fprintf(stderr, "?...");
 628         (void)fflush(stderr);
 629         if ((response = fgetln(ttyfp, &rsize)) == NULL ||
 630             regcomp(&cre, nl_langinfo(YESEXPR), REG_BASIC) != 0) {
 631                 (void)fclose(ttyfp);
 632                 return (0);
 633         }
 634         match = regexec(&cre, response, 0, NULL, 0);
 635         (void)fclose(ttyfp);
 636         regfree(&cre);
 637         return (match == 0);
 638 }
 639
 640 static void
 641 usage(void)
 642 {
 643         fprintf(stderr,
 644 "usage: xargs [-0opt] [-E eofstr] [-I replstr [-R replacements]] [-J replstr]\n"
 645 "             [-L number] [-n number [-x]] [-P maxprocs] [-s size]\n"
 646 "             [utility [argument ...]]\n");
 647         exit(1);
 648 }