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, nflag, nline;
 104         size_t nargs;
 105         size_t linelen;
 106         char *endptr;
 107
 108         inpline = replstr = NULL;
 109         ep = environ;
 110         eofstr = "";
 111         Jflag = nflag = 0;
 112
 113         (void)setlocale(LC_ALL, "");
 114
 115         /*
 116          * POSIX.2 limits the exec line length to ARG_MAX - 2K.  Running that
 117          * caused some E2BIG errors, so it was changed to ARG_MAX - 4K.  Given
 118          * that the smallest argument is 2 bytes in length, this means that
 119          * the number of arguments is limited to:
 120          *
 121          *       (ARG_MAX - 4K - LENGTH(utility + arguments)) / 2.
 122          *
 123          * We arbitrarily limit the number of arguments to 5000.  This is
 124          * allowed by POSIX.2 as long as the resulting minimum exec line is
 125          * at least LINE_MAX.  Realloc'ing as necessary is possible, but
 126          * probably not worthwhile.
 127          */
 128         nargs = 5000;
 129         if ((arg_max = sysconf(_SC_ARG_MAX)) == -1)
 130                 errx(1, "sysconf(_SC_ARG_MAX) failed");
 131         nline = arg_max - 4 * 1024;
 132         while (*ep != NULL) {
 133                 /* 1 byte for each '\0' */
 134                 nline -= strlen(*ep++) + 1 + sizeof(*ep);
 135         }
 136         maxprocs = 1;
 137         while ((ch = getopt(argc, argv, "0E:I:J:L:n:oP:pR:s:tx")) != -1)
 138                 switch(ch) {
 139                 case 'E':
 140                         eofstr = optarg;
 141                         break;
 142                 case 'I':
 143                         Jflag = 0;
 144                         Iflag = 1;
 145                         Lflag = 1;
 146                         replstr = optarg;
 147                         break;
 148                 case 'J':
 149                         Iflag = 0;
 150                         Jflag = 1;
 151                         replstr = optarg;
 152                         break;
 153                 case 'L':
 154                         Lflag = atoi(optarg);
 155                         if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 156                                 nflag = 0; /* Override */
 157                                 nargs = 5000;
 158                         }
 159                         break;
 160                 case 'n':
 161                         nflag = 1;
 162                         if ((nargs = strtol(optarg, NULL, 10)) <= 0)
 163                                 errx(1, "illegal argument count");
 164                         if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 165                                 Lflag = 0; /* Override */
 166                         }
 167                         break;
 168                 case 'o':
 169                         oflag = 1;
 170                         break;
 171                 case 'P':
 172                         if ((maxprocs = atoi(optarg)) <= 0)
 173                                 errx(1, "max. processes must be >0");
 174                         break;
 175                 case 'p':
 176                         pflag = 1;
 177                         break;
 178                 case 'R':
 179                         Rflag = strtol(optarg, &endptr, 10);
 180                         if (*endptr != '\0')
 181                                 errx(1, "replacements must be a number");
 182                         break;
 183                 case 's':
 184                         nline = atoi(optarg);
 185                         break;
 186                 case 't':
 187                         tflag = 1;
 188                         break;
 189                 case 'x':
 190                         xflag = 1;
 191                         break;
 192                 case '0':
 193                         zflag = 1;
 194                         break;
 195                 case '?':
 196                 default:
 197                         usage();
 198         }
 199         argc -= optind;
 200         argv += optind;
 201
 202         if (!Iflag && Rflag)
 203                 usage();
 204         if (Iflag && !Rflag)
 205                 Rflag = 5;
 206         if (xflag && !nflag)
 207                 usage();
 208         if (Iflag || Lflag)
 209                 xflag = 1;
 210         if (replstr != NULL && *replstr == '\0')
 211                 errx(1, "replstr may not be empty");
 212
 213         /*
 214          * Allocate pointers for the utility name, the utility arguments,
 215          * the maximum arguments to be read from stdin and the trailing
 216          * NULL.
 217          */
 218         linelen = 1 + argc + nargs + 1;
 219         if ((av = bxp = malloc(linelen * sizeof(char **))) == NULL)
 220                 errx(1, "malloc failed");
 221
 222         /*
 223          * Use the user's name for the utility as argv[0], just like the
 224          * shell.  Echo is the default.  Set up pointers for the user's
 225          * arguments.
 226          */
 227         if (*argv == NULL)
 228                 cnt = strlen(*bxp++ = echo);
 229         else {
 230                 do {
 231                         if (Jflag && strcmp(*argv, replstr) == 0) {
 232                                 char **avj;
 233                                 jfound = 1;
 234                                 argv++;
 235                                 for (avj = argv; *avj; avj++)
 236                                         cnt += strlen(*avj) + 1;
 237                                 break;
 238                         }
 239                         cnt += strlen(*bxp++ = *argv) + 1;
 240                 } while (*++argv != NULL);
 241         }
 242
 243         /*
 244          * Set up begin/end/traversing pointers into the array.  The -n
 245          * count doesn't include the trailing NULL pointer, so the malloc
 246          * added in an extra slot.
 247          */
 248         endxp = (xp = bxp) + nargs;
 249
 250         /*
 251          * Allocate buffer space for the arguments read from stdin and the
 252          * trailing NULL.  Buffer space is defined as the default or specified
 253          * space, minus the length of the utility name and arguments.  Set up
 254          * begin/end/traversing pointers into the array.  The -s count does
 255          * include the trailing NULL, so the malloc didn't add in an extra
 256          * slot.
 257          */
 258         nline -= cnt;
 259         if (nline <= 0)
 260                 errx(1, "insufficient space for command");
 261
 262         if ((bbp = malloc((size_t)(nline + 1))) == NULL)
 263                 errx(1, "malloc failed");
 264         ebp = (argp = p = bbp) + nline - 1;
 265         for (;;)
 266                 parse_input(argc, argv);
 267 }
 268
 269 static void
 270 parse_input(int argc, char *argv[])
 271 {
 272         int ch, foundeof;
 273         char **avj;
 274         int last_was_backslashed = 0;
 275
 276         foundeof = 0;
 277
 278         switch(ch = getchar()) {
 279         case EOF:
 280                 /* No arguments since last exec. */
 281                 if (p == bbp) {
 282                         waitchildren(*argv, 1);
 283                         exit(rval);
 284                 }
 285                 goto arg1;
 286         case ' ':
 287                 last_was_blank = 1;
 288         case '\t':
 289                 /* Quotes escape tabs and spaces. */
 290                 if (insingle || indouble || zflag)
 291                         goto addch;
 292                 goto arg2;
 293         case '\0':
 294                 if (zflag) {
 295                         /*
 296                          * Increment 'count', so that nulls will be treated
 297                          * as end-of-line, as well as end-of-argument.  This
 298                          * is needed so -0 works properly with -I and -L.
 299                          */
 300                         count++;
 301                         goto arg2;
 302                 }
 303                 goto addch;
 304         case '\n':
 305                 if (zflag)
 306                         goto addch;
 307                 if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 308                         if (last_was_newline) {
 309                                 /* don't count empty line */
 310                                 break;
 311                         }
 312                         if (!last_was_blank ) {
 313                                 /* only count if NOT continuation line */
 314                                 count++;
 315                         }
 316                 } else {
 317                         count++;
 318                 }
 319                 last_was_newline = 1;
 320
 321                 /* Quotes do not escape newlines. */
 322 arg1:           if (insingle || indouble)
 323                         errx(1, "unterminated quote");
 324 arg2:
 325                 foundeof = *eofstr != '\0' &&
 326                     strcmp(argp, eofstr) == 0;
 327
 328 #ifdef __APPLE__
 329                 /* 6591323: -I specifies that it processes the entire line,
 330                  * so only recognize eofstr at the end of a line. */
 331                 if (Iflag && !last_was_newline)
 332                         foundeof = 0;
 333
 334                 /* 6591323: Essentially the same as the EOF handling above. */
 335                 if (foundeof && (p - strlen(eofstr) == bbp)) {
 336                         waitchildren(*argv, 1);
 337                         exit(rval);
 338                 }
 339 #endif
 340
 341                 /* Do not make empty args unless they are quoted */
 342                 if ((argp != p || wasquoted) && !foundeof) {
 343                         *p++ = '\0';
 344                         *xp++ = argp;
 345                         if (Iflag) {
 346                                 size_t curlen;
 347
 348                                 if (inpline == NULL)
 349                                         curlen = 0;
 350                                 else {
 351                                         /*
 352                                          * If this string is not zero
 353                                          * length, append a space for
 354                                          * separation before the next
 355                                          * argument.
 356                                          */
 357                                         if ((curlen = strlen(inpline)))
 358                                                 strcat(inpline, " ");
 359                                 }
 360                                 curlen++;
 361                                 /*
 362                                  * Allocate enough to hold what we will
 363                                  * be holding in a second, and to append
 364                                  * a space next time through, if we have
 365                                  * to.
 366                                  */
 367                                 inpline = realloc(inpline, curlen + 2 +
 368                                     strlen(argp));
 369                                 if (inpline == NULL)
 370                                         errx(1, "realloc failed");
 371                                 if (curlen == 1)
 372                                         strcpy(inpline, argp);
 373                                 else
 374                                         strcat(inpline, argp);
 375                         }
 376                 }
 377
 378                 /*
 379                  * If max'd out on args or buffer, or reached EOF,
 380                  * run the command.  If xflag and max'd out on buffer
 381                  * but not on args, object.  Having reached the limit
 382                  * of input lines, as specified by -L is the same as
 383                  * maxing out on arguments.
 384                  */
 385                 if (xp == endxp || p > ebp || ch == EOF ||
 386                     (Lflag <= count && xflag) || foundeof) {
 387                         if (xflag && xp != endxp && p > ebp)
 388                                 errx(1, "insufficient space for arguments");
 389                         if (jfound) {
 390                                 for (avj = argv; *avj; avj++)
 391                                         *xp++ = *avj;
 392                         }
 393                         prerun(argc, av);
 394                         if (ch == EOF || foundeof) {
 395                                 waitchildren(*argv, 1);
 396                                 exit(rval);
 397                         }
 398                         p = bbp;
 399                         xp = bxp;
 400                         count = 0;
 401                 }
 402                 argp = p;
 403                 wasquoted = 0;
 404                 break;
 405         case '\'':
 406                 if (indouble || zflag)
 407                         goto addch;
 408                 insingle = !insingle;
 409                 wasquoted = 1;
 410                 break;
 411         case '"':
 412                 if (insingle || zflag)
 413                         goto addch;
 414                 indouble = !indouble;
 415                 wasquoted = 1;
 416                 break;
 417         case '\\':
 418                 last_was_backslashed = 1;
 419                 if (zflag)
 420                         goto addch;
 421                 /* Backslash escapes anything, is escaped by quotes. */
 422                 if (!insingle && !indouble && (ch = getchar()) == EOF)
 423                         errx(1, "backslash at EOF");
 424                 /* FALLTHROUGH */
 425         default:
 426 addch:          if (p < ebp) {
 427                         *p++ = ch;
 428                         break;
 429                 }
 430
 431                 /* If only one argument, not enough buffer space. */
 432                 if (bxp == xp)
 433                         errx(1, "insufficient space for argument");
 434                 /* Didn't hit argument limit, so if xflag object. */
 435                 if (xflag)
 436                         errx(1, "insufficient space for arguments");
 437
 438                 if (jfound) {
 439                         for (avj = argv; *avj; avj++)
 440                                 *xp++ = *avj;
 441                 }
 442                 prerun(argc, av);
 443                 xp = bxp;
 444                 cnt = ebp - argp;
 445                 memcpy(bbp, argp, (size_t)cnt);
 446                 p = (argp = bbp) + cnt;
 447                 *p++ = ch;
 448                 break;
 449         }
 450         if (ch != ' ')
 451                 last_was_blank = 0;
 452         if (ch != '\n' || last_was_backslashed)
 453                 last_was_newline = 0;
 454 }
 455
 456 /*
 457  * Do things necessary before run()'ing, such as -I substitution,
 458  * and then call run().
 459  */
 460 static void
 461 prerun(int argc, char *argv[])
 462 {
 463         char **tmp, **tmp2, **avj;
 464         int repls;
 465
 466         repls = Rflag;
 467
 468         if (argc == 0 || repls == 0) {
 469                 *xp = NULL;
 470                 run(argv);
 471                 return;
 472         }
 473
 474         avj = argv;
 475
 476         /*
 477          * Allocate memory to hold the argument list, and
 478          * a NULL at the tail.
 479          */
 480         tmp = malloc((argc + 1) * sizeof(char**));
 481         if (tmp == NULL)
 482                 errx(1, "malloc failed");
 483         tmp2 = tmp;
 484
 485         /*
 486          * Save the first argument and iterate over it, we
 487          * cannot do strnsubst() to it.
 488          */
 489         if ((*tmp++ = strdup(*avj++)) == NULL)
 490                 errx(1, "strdup failed");
 491
 492         /*
 493          * For each argument to utility, if we have not used up
 494          * the number of replacements we are allowed to do, and
 495          * if the argument contains at least one occurrence of
 496          * replstr, call strnsubst(), else just save the string.
 497          * Iterations over elements of avj and tmp are done
 498          * where appropriate.
 499          */
 500         while (--argc) {
 501                 *tmp = *avj++;
 502                 if (repls && strstr(*tmp, replstr) != NULL) {
 503                         strnsubst(tmp++, replstr, inpline, (size_t)255);
 504                         if (repls > 0)
 505                                 repls--;
 506                 } else {
 507                         if ((*tmp = strdup(*tmp)) == NULL)
 508                                 errx(1, "strdup failed");
 509                         tmp++;
 510                 }
 511         }
 512
 513         /*
 514          * Run it.
 515          */
 516         *tmp = NULL;
 517         run(tmp2);
 518
 519         /*
 520          * Walk from the tail to the head, free along the way.
 521          */
 522         for (; tmp2 != tmp; tmp--)
 523                 free(*tmp);
 524         /*
 525          * Now free the list itself.
 526          */
 527         free(tmp2);
 528
 529         /*
 530          * Free the input line buffer, if we have one.
 531          */
 532         if (inpline != NULL) {
 533                 free(inpline);
 534                 inpline = NULL;
 535         }
 536 }
 537
 538 static void
 539 run(char **argv)
 540 {
 541         pid_t pid;
 542         int fd;
 543         char **avec;
 544
 545         /*
 546          * If the user wants to be notified of each command before it is
 547          * executed, notify them.  If they want the notification to be
 548          * followed by a prompt, then prompt them.
 549          */
 550         if (tflag || pflag) {
 551                 (void)fprintf(stderr, "%s", *argv);
 552                 for (avec = argv + 1; *avec != NULL; ++avec)
 553                         (void)fprintf(stderr, " %s", *avec);
 554                 /*
 555                  * If the user has asked to be prompted, do so.
 556                  */
 557                 if (pflag)
 558                         /*
 559                          * If they asked not to exec, return without execution
 560                          * but if they asked to, go to the execution.  If we
 561                          * could not open their tty, break the switch and drop
 562                          * back to -t behaviour.
 563                          */
 564                         switch (prompt()) {
 565                         case 0:
 566                                 return;
 567                         case 1:
 568                                 goto exec;
 569                         case 2:
 570                                 break;
 571                         }
 572                 (void)fprintf(stderr, "\n");
 573                 (void)fflush(stderr);
 574         }
 575 exec:
 576         childerr = 0;
 577         switch(pid = vfork()) {
 578         case -1:
 579                 err(1, "vfork");
 580         case 0:
 581                 if (oflag) {
 582                         if ((fd = open(_PATH_TTY, O_RDONLY)) == -1)
 583                                 err(1, "can't open /dev/tty");
 584                 } else {
 585                         fd = open(_PATH_DEVNULL, O_RDONLY);
 586                 }
 587                 if (fd > STDIN_FILENO) {
 588                         if (dup2(fd, STDIN_FILENO) != 0)
 589                                 err(1, "can't dup2 to stdin");
 590                         close(fd);
 591                 }
 592                 execvp(argv[0], argv);
 593                 childerr = errno;
 594                 _exit(1);
 595         }
 596         curprocs++;
 597         waitchildren(*argv, 0);
 598 }
 599
 600 static void
 601 waitchildren(const char *name, int waitall)
 602 {
 603         pid_t pid;
 604         int status;
 605
 606         while ((pid = waitpid(-1, &status, !waitall && curprocs < maxprocs ?
 607             WNOHANG : 0)) > 0) {
 608                 curprocs--;
 609                 /* If we couldn't invoke the utility, exit. */
 610                 if (childerr != 0) {
 611                         errno = childerr;
 612                         err(errno == ENOENT ? 127 : 126, "%s", name);
 613                 }
 614                 /*
 615                  * If utility signaled or exited with a value of 255,
 616                  * exit 1-125.
 617                  */
 618                 if (WIFSIGNALED(status) || WEXITSTATUS(status) == 255)
 619                         exit(1);
 620                 if (WEXITSTATUS(status))
 621                         rval = 1;
 622         }
 623         if (pid == -1 && errno != ECHILD)
 624                 err(1, "wait3");
 625 }
 626
 627 /*
 628  * Prompt the user about running a command.
 629  */
 630 static int
 631 prompt(void)
 632 {
 633         regex_t cre;
 634         size_t rsize;
 635         int match;
 636         char *response;
 637         FILE *ttyfp;
 638
 639         if ((ttyfp = fopen(_PATH_TTY, "r")) == NULL)
 640                 return (2);     /* Indicate that the TTY failed to open. */
 641         (void)fprintf(stderr, "?...");
 642         (void)fflush(stderr);
 643         if ((response = fgetln(ttyfp, &rsize)) == NULL ||
 644             regcomp(&cre, nl_langinfo(YESEXPR), REG_BASIC) != 0) {
 645                 (void)fclose(ttyfp);
 646                 return (0);
 647         }
 648         match = regexec(&cre, response, 0, NULL, 0);
 649         (void)fclose(ttyfp);
 650         regfree(&cre);
 651         return (match == 0);
 652 }
 653
 654 static void
 655 usage(void)
 656 {
 657         fprintf(stderr,
 658 "usage: xargs [-0opt] [-E eofstr] [-I replstr [-R replacements]] [-J replstr]\n"
 659 "             [-L number] [-n number [-x]] [-P maxprocs] [-s size]\n"
 660 "             [utility [argument ...]]\n");
 661         exit(1);
 662 }