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 static size_t pad9314053;
  95
  96 static volatile int childerr;
  97
  98 extern char **environ;
  99
 100 int
 101 main(int argc, char *argv[])
 102 {
 103         long arg_max;
 104         int ch, Jflag, nflag, nline;
 105         size_t nargs;
 106         size_t linelen;
 107         char *endptr;
 108
 109         inpline = replstr = NULL;
 110         ep = environ;
 111         eofstr = "";
 112         Jflag = nflag = 0;
 113
 114         (void)setlocale(LC_ALL, "");
 115
 116         /*
 117          * POSIX.2 limits the exec line length to ARG_MAX - 2K.  Running that
 118          * caused some E2BIG errors, so it was changed to ARG_MAX - 4K.  Given
 119          * that the smallest argument is 2 bytes in length, this means that
 120          * the number of arguments is limited to:
 121          *
 122          *       (ARG_MAX - 4K - LENGTH(utility + arguments)) / 2.
 123          *
 124          * We arbitrarily limit the number of arguments to 5000.  This is
 125          * allowed by POSIX.2 as long as the resulting minimum exec line is
 126          * at least LINE_MAX.  Realloc'ing as necessary is possible, but
 127          * probably not worthwhile.
 128          */
 129         nargs = 5000;
 130         if ((arg_max = sysconf(_SC_ARG_MAX)) == -1)
 131                 errx(1, "sysconf(_SC_ARG_MAX) failed");
 132         nline = arg_max - MAXPATHLEN; /* for argv[0] from execvp() */
 133         pad9314053 = sizeof(char *); /* reserve for string area rounding */
 134         while (*ep != NULL) {
 135                 /* 1 byte for each '\0' */
 136                 nline -= strlen(*ep++) + 1 + sizeof(*ep);
 137         }
 138         nline -= pad9314053;
 139         maxprocs = 1;
 140         while ((ch = getopt(argc, argv, "0E:I:J:L:n:oP:pR:s:tx")) != -1)
 141                 switch(ch) {
 142                 case 'E':
 143                         eofstr = optarg;
 144                         break;
 145                 case 'I':
 146                         Jflag = 0;
 147                         Iflag = 1;
 148                         Lflag = 1;
 149                         replstr = optarg;
 150                         break;
 151                 case 'J':
 152                         Iflag = 0;
 153                         Jflag = 1;
 154                         replstr = optarg;
 155                         break;
 156                 case 'L':
 157                         Lflag = atoi(optarg);
 158                         if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 159                                 nflag = 0; /* Override */
 160                                 nargs = 5000;
 161                         }
 162                         break;
 163                 case 'n':
 164                         nflag = 1;
 165                         if ((nargs = strtol(optarg, NULL, 10)) <= 0)
 166                                 errx(1, "illegal argument count");
 167                         if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 168                                 Lflag = 0; /* Override */
 169                         }
 170                         break;
 171                 case 'o':
 172                         oflag = 1;
 173                         break;
 174                 case 'P':
 175                         if ((maxprocs = atoi(optarg)) <= 0)
 176                                 errx(1, "max. processes must be >0");
 177                         break;
 178                 case 'p':
 179                         pflag = 1;
 180                         break;
 181                 case 'R':
 182                         Rflag = strtol(optarg, &endptr, 10);
 183                         if (*endptr != '\0')
 184                                 errx(1, "replacements must be a number");
 185                         break;
 186                 case 's':
 187                         nline = atoi(optarg);
 188                         pad9314053 = 0; /* assume the -s value is valid */
 189                         break;
 190                 case 't':
 191                         tflag = 1;
 192                         break;
 193                 case 'x':
 194                         xflag = 1;
 195                         break;
 196                 case '0':
 197                         zflag = 1;
 198                         break;
 199                 case '?':
 200                 default:
 201                         usage();
 202         }
 203         argc -= optind;
 204         argv += optind;
 205
 206         if (!Iflag && Rflag)
 207                 usage();
 208         if (Iflag && !Rflag)
 209                 Rflag = 5;
 210         if (xflag && !nflag)
 211                 usage();
 212         if (Iflag || Lflag)
 213                 xflag = 1;
 214         if (replstr != NULL && *replstr == '\0')
 215                 errx(1, "replstr may not be empty");
 216
 217         /*
 218          * Allocate pointers for the utility name, the utility arguments,
 219          * the maximum arguments to be read from stdin and the trailing
 220          * NULL.
 221          */
 222         linelen = 1 + argc + nargs + 1;
 223         if ((av = bxp = malloc(linelen * sizeof(char **))) == NULL)
 224                 errx(1, "malloc failed");
 225
 226         /*
 227          * Use the user's name for the utility as argv[0], just like the
 228          * shell.  Echo is the default.  Set up pointers for the user's
 229          * arguments.
 230          */
 231         if (*argv == NULL)
 232                 cnt = strlen(*bxp++ = echo) + pad9314053;
 233         else {
 234                 do {
 235                         if (Jflag && strcmp(*argv, replstr) == 0) {
 236                                 char **avj;
 237                                 jfound = 1;
 238                                 argv++;
 239                                 for (avj = argv; *avj; avj++)
 240                                         cnt += strlen(*avj) + 1 + pad9314053;
 241                                 break;
 242                         }
 243                         cnt += strlen(*bxp++ = *argv) + 1 + pad9314053;
 244                 } while (*++argv != NULL);
 245         }
 246
 247         /*
 248          * Set up begin/end/traversing pointers into the array.  The -n
 249          * count doesn't include the trailing NULL pointer, so the malloc
 250          * added in an extra slot.
 251          */
 252         endxp = (xp = bxp) + nargs;
 253
 254         /*
 255          * Allocate buffer space for the arguments read from stdin and the
 256          * trailing NULL.  Buffer space is defined as the default or specified
 257          * space, minus the length of the utility name and arguments.  Set up
 258          * begin/end/traversing pointers into the array.  The -s count does
 259          * include the trailing NULL, so the malloc didn't add in an extra
 260          * slot.
 261          */
 262         nline -= cnt;
 263         if (nline <= 0)
 264                 errx(1, "insufficient space for command");
 265
 266         if ((bbp = malloc((size_t)(nline + 1))) == NULL)
 267                 errx(1, "malloc failed");
 268         ebp = (argp = p = bbp) + nline - 1;
 269         for (;;)
 270                 parse_input(argc, argv);
 271 }
 272
 273 static void
 274 parse_input(int argc, char *argv[])
 275 {
 276         int ch, foundeof;
 277         char **avj;
 278         int last_was_backslashed = 0;
 279
 280         foundeof = 0;
 281
 282         switch(ch = getchar()) {
 283         case EOF:
 284                 /* No arguments since last exec. */
 285                 if (p == bbp) {
 286                         waitchildren(*argv, 1);
 287                         exit(rval);
 288                 }
 289                 goto arg1;
 290         case ' ':
 291                 last_was_blank = 1;
 292         case '\t':
 293                 /* Quotes escape tabs and spaces. */
 294                 if (insingle || indouble || zflag)
 295                         goto addch;
 296                 goto arg2;
 297         case '\0':
 298                 if (zflag) {
 299                         /*
 300                          * Increment 'count', so that nulls will be treated
 301                          * as end-of-line, as well as end-of-argument.  This
 302                          * is needed so -0 works properly with -I and -L.
 303                          */
 304                         count++;
 305                         goto arg2;
 306                 }
 307                 goto addch;
 308         case '\n':
 309                 if (zflag)
 310                         goto addch;
 311                 if (COMPAT_MODE("bin/xargs", "Unix2003")) {
 312                         if (last_was_newline) {
 313                                 /* don't count empty line */
 314                                 break;
 315                         }
 316                         if (!last_was_blank ) {
 317                                 /* only count if NOT continuation line */
 318                                 count++;
 319                         }
 320                 } else {
 321                         count++;
 322                 }
 323                 last_was_newline = 1;
 324
 325                 /* Quotes do not escape newlines. */
 326 arg1:           if (insingle || indouble)
 327                         errx(1, "unterminated quote");
 328 arg2:
 329                 foundeof = *eofstr != '\0' &&
 330                     strcmp(argp, eofstr) == 0;
 331
 332 #ifdef __APPLE__
 333                 /* 6591323: -I specifies that it processes the entire line,
 334                  * so only recognize eofstr at the end of a line. */
 335                 if (Iflag && !last_was_newline)
 336                         foundeof = 0;
 337
 338                 /* 6591323: Essentially the same as the EOF handling above. */
 339                 if (foundeof && (p - strlen(eofstr) == bbp)) {
 340                         waitchildren(*argv, 1);
 341                         exit(rval);
 342                 }
 343 #endif
 344
 345                 /* Do not make empty args unless they are quoted */
 346                 if ((argp != p || wasquoted) && !foundeof) {
 347                         *p++ = '\0';
 348                         *xp++ = argp;
 349                         if (Iflag) {
 350                                 size_t curlen;
 351
 352                                 if (inpline == NULL)
 353                                         curlen = 0;
 354                                 else {
 355                                         /*
 356                                          * If this string is not zero
 357                                          * length, append a space for
 358                                          * separation before the next
 359                                          * argument.
 360                                          */
 361                                         if ((curlen = strlen(inpline)))
 362                                                 strcat(inpline, " ");
 363                                 }
 364                                 curlen++;
 365                                 /*
 366                                  * Allocate enough to hold what we will
 367                                  * be holding in a second, and to append
 368                                  * a space next time through, if we have
 369                                  * to.
 370                                  */
 371                                 inpline = realloc(inpline, curlen + 2 +
 372                                     strlen(argp));
 373                                 if (inpline == NULL)
 374                                         errx(1, "realloc failed");
 375                                 if (curlen == 1)
 376                                         strcpy(inpline, argp);
 377                                 else
 378                                         strcat(inpline, argp);
 379                         }
 380                 }
 381
 382                 /*
 383                  * If max'd out on args or buffer, or reached EOF,
 384                  * run the command.  If xflag and max'd out on buffer
 385                  * but not on args, object.  Having reached the limit
 386                  * of input lines, as specified by -L is the same as
 387                  * maxing out on arguments.
 388                  */
 389                 if (xp == endxp || p + (count * pad9314053) > ebp || ch == EOF ||
 390                     (Lflag <= count && xflag) || foundeof) {
 391                         if (xflag && xp != endxp && p + (count * pad9314053) > ebp)
 392                                 errx(1, "insufficient space for arguments");
 393                         if (jfound) {
 394                                 for (avj = argv; *avj; avj++)
 395                                         *xp++ = *avj;
 396                         }
 397                         prerun(argc, av);
 398                         if (ch == EOF || foundeof) {
 399                                 waitchildren(*argv, 1);
 400                                 exit(rval);
 401                         }
 402                         p = bbp;
 403                         xp = bxp;
 404                         count = 0;
 405                 }
 406                 argp = p;
 407                 wasquoted = 0;
 408                 break;
 409         case '\'':
 410                 if (indouble || zflag)
 411                         goto addch;
 412                 insingle = !insingle;
 413                 wasquoted = 1;
 414                 break;
 415         case '"':
 416                 if (insingle || zflag)
 417                         goto addch;
 418                 indouble = !indouble;
 419                 wasquoted = 1;
 420                 break;
 421         case '\\':
 422                 last_was_backslashed = 1;
 423                 if (zflag)
 424                         goto addch;
 425                 /* Backslash escapes anything, is escaped by quotes. */
 426                 if (!insingle && !indouble && (ch = getchar()) == EOF)
 427                         errx(1, "backslash at EOF");
 428                 /* FALLTHROUGH */
 429         default:
 430 addch:          if (p < ebp) {
 431                         *p++ = ch;
 432                         break;
 433                 }
 434
 435                 /* If only one argument, not enough buffer space. */
 436                 if (bxp == xp)
 437                         errx(1, "insufficient space for argument");
 438                 /* Didn't hit argument limit, so if xflag object. */
 439                 if (xflag)
 440                         errx(1, "insufficient space for arguments");
 441
 442                 if (jfound) {
 443                         for (avj = argv; *avj; avj++)
 444                                 *xp++ = *avj;
 445                 }
 446                 prerun(argc, av);
 447                 xp = bxp;
 448                 cnt = ebp - argp;
 449                 memcpy(bbp, argp, (size_t)cnt);
 450                 p = (argp = bbp) + cnt;
 451                 *p++ = ch;
 452                 break;
 453         }
 454         if (ch != ' ')
 455                 last_was_blank = 0;
 456         if (ch != '\n' || last_was_backslashed)
 457                 last_was_newline = 0;
 458 }
 459
 460 /*
 461  * Do things necessary before run()'ing, such as -I substitution,
 462  * and then call run().
 463  */
 464 static void
 465 prerun(int argc, char *argv[])
 466 {
 467         char **tmp, **tmp2, **avj;
 468         int repls;
 469
 470         repls = Rflag;
 471
 472         if (argc == 0 || repls == 0) {
 473                 *xp = NULL;
 474                 run(argv);
 475                 return;
 476         }
 477
 478         avj = argv;
 479
 480         /*
 481          * Allocate memory to hold the argument list, and
 482          * a NULL at the tail.
 483          */
 484         tmp = malloc((argc + 1) * sizeof(char**));
 485         if (tmp == NULL)
 486                 errx(1, "malloc failed");
 487         tmp2 = tmp;
 488
 489         /*
 490          * Save the first argument and iterate over it, we
 491          * cannot do strnsubst() to it.
 492          */
 493         if ((*tmp++ = strdup(*avj++)) == NULL)
 494                 errx(1, "strdup failed");
 495
 496         /*
 497          * For each argument to utility, if we have not used up
 498          * the number of replacements we are allowed to do, and
 499          * if the argument contains at least one occurrence of
 500          * replstr, call strnsubst(), else just save the string.
 501          * Iterations over elements of avj and tmp are done
 502          * where appropriate.
 503          */
 504         while (--argc) {
 505                 *tmp = *avj++;
 506                 if (repls && strstr(*tmp, replstr) != NULL) {
 507                         strnsubst(tmp++, replstr, inpline, (size_t)255);
 508                         if (repls > 0)
 509                                 repls--;
 510                 } else {
 511                         if ((*tmp = strdup(*tmp)) == NULL)
 512                                 errx(1, "strdup failed");
 513                         tmp++;
 514                 }
 515         }
 516
 517         /*
 518          * Run it.
 519          */
 520         *tmp = NULL;
 521         run(tmp2);
 522
 523         /*
 524          * Walk from the tail to the head, free along the way.
 525          */
 526         for (; tmp2 != tmp; tmp--)
 527                 free(*tmp);
 528         /*
 529          * Now free the list itself.
 530          */
 531         free(tmp2);
 532
 533         /*
 534          * Free the input line buffer, if we have one.
 535          */
 536         if (inpline != NULL) {
 537                 free(inpline);
 538                 inpline = NULL;
 539         }
 540 }
 541
 542 static void
 543 run(char **argv)
 544 {
 545         pid_t pid;
 546         int fd;
 547         char **avec;
 548
 549         /*
 550          * If the user wants to be notified of each command before it is
 551          * executed, notify them.  If they want the notification to be
 552          * followed by a prompt, then prompt them.
 553          */
 554         if (tflag || pflag) {
 555                 (void)fprintf(stderr, "%s", *argv);
 556                 for (avec = argv + 1; *avec != NULL; ++avec)
 557                         (void)fprintf(stderr, " %s", *avec);
 558                 /*
 559                  * If the user has asked to be prompted, do so.
 560                  */
 561                 if (pflag)
 562                         /*
 563                          * If they asked not to exec, return without execution
 564                          * but if they asked to, go to the execution.  If we
 565                          * could not open their tty, break the switch and drop
 566                          * back to -t behaviour.
 567                          */
 568                         switch (prompt()) {
 569                         case 0:
 570                                 return;
 571                         case 1:
 572                                 goto exec;
 573                         case 2:
 574                                 break;
 575                         }
 576                 (void)fprintf(stderr, "\n");
 577                 (void)fflush(stderr);
 578         }
 579 exec:
 580         childerr = 0;
 581         switch(pid = vfork()) {
 582         case -1:
 583                 err(1, "vfork");
 584         case 0:
 585                 if (oflag) {
 586                         if ((fd = open(_PATH_TTY, O_RDONLY)) == -1)
 587                                 err(1, "can't open /dev/tty");
 588                 } else {
 589                         fd = open(_PATH_DEVNULL, O_RDONLY);
 590                 }
 591                 if (fd > STDIN_FILENO) {
 592                         if (dup2(fd, STDIN_FILENO) != 0)
 593                                 err(1, "can't dup2 to stdin");
 594                         close(fd);
 595                 }
 596                 execvp(argv[0], argv);
 597                 childerr = errno;
 598                 _exit(1);
 599         }
 600         curprocs++;
 601         waitchildren(*argv, 0);
 602 }
 603
 604 static void
 605 waitchildren(const char *name, int waitall)
 606 {
 607         pid_t pid;
 608         int status;
 609
 610         while ((pid = waitpid(-1, &status, !waitall && curprocs < maxprocs ?
 611             WNOHANG : 0)) > 0) {
 612                 curprocs--;
 613                 /* If we couldn't invoke the utility, exit. */
 614                 if (childerr != 0) {
 615                         errno = childerr;
 616                         err(errno == ENOENT ? 127 : 126, "%s", name);
 617                 }
 618                 /*
 619                  * If utility signaled or exited with a value of 255,
 620                  * exit 1-125.
 621                  */
 622                 if (WIFSIGNALED(status) || WEXITSTATUS(status) == 255)
 623                         exit(1);
 624                 if (WEXITSTATUS(status))
 625                         rval = 1;
 626         }
 627         if (pid == -1 && errno != ECHILD)
 628                 err(1, "wait3");
 629 }
 630
 631 /*
 632  * Prompt the user about running a command.
 633  */
 634 static int
 635 prompt(void)
 636 {
 637         regex_t cre;
 638         size_t rsize;
 639         int match;
 640         char *response;
 641         FILE *ttyfp;
 642
 643         if ((ttyfp = fopen(_PATH_TTY, "r")) == NULL)
 644                 return (2);     /* Indicate that the TTY failed to open. */
 645         (void)fprintf(stderr, "?...");
 646         (void)fflush(stderr);
 647         if ((response = fgetln(ttyfp, &rsize)) == NULL ||
 648             regcomp(&cre, nl_langinfo(YESEXPR), REG_BASIC) != 0) {
 649                 (void)fclose(ttyfp);
 650                 return (0);
 651         }
 652         match = regexec(&cre, response, 0, NULL, 0);
 653         (void)fclose(ttyfp);
 654         regfree(&cre);
 655         return (match == 0);
 656 }
 657
 658 static void
 659 usage(void)
 660 {
 661         fprintf(stderr,
 662 "usage: xargs [-0opt] [-E eofstr] [-I replstr [-R replacements]] [-J replstr]\n"
 663 "             [-L number] [-n number [-x]] [-P maxprocs] [-s size]\n"
 664 "             [utility [argument ...]]\n");
 665         exit(1);
 666 }