usr/src/cmd/sed/main.c

   1 /*
   2  * Copyright (c) 2011 Gary Mills
   3  * Copyright 2011 Nexenta Systems, Inc. All rights reserved.
   4  * Copyright (c) 1992 Diomidis Spinellis.
   5  * Copyright (c) 1992, 1993
   6  *      The Regents of the University of California.  All rights reserved.
   7  *
   8  * This code is derived from software contributed to Berkeley by
   9  * Diomidis Spinellis of Imperial College, University of London.
  10  *
  11  * Redistribution and use in source and binary forms, with or without
  12  * modification, are permitted provided that the following conditions
  13  * are met:
  14  * 1. Redistributions of source code must retain the above copyright
  15  *    notice, this list of conditions and the following disclaimer.
  16  * 2. Redistributions in binary form must reproduce the above copyright
  17  *    notice, this list of conditions and the following disclaimer in the
  18  *    documentation and/or other materials provided with the distribution.
  19  * 4. Neither the name of the University nor the names of its contributors
  20  *    may be used to endorse or promote products derived from this software
  21  *    without specific prior written permission.
  22  *
  23  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  33  * SUCH DAMAGE.
  34  */
  35
  36 #include <sys/types.h>
  37 #include <sys/mman.h>
  38 #include <sys/param.h>
  39 #include <sys/stat.h>
  40
  41 #include <err.h>
  42 #include <errno.h>
  43 #include <fcntl.h>
  44 #include <getopt.h>
  45 #include <libgen.h>
  46 #include <libintl.h>
  47 #include <limits.h>
  48 #include <locale.h>
  49 #include <regex.h>
  50 #include <stddef.h>
  51 #include <stdio.h>
  52 #include <stdlib.h>
  53 #include <string.h>
  54 #include <unistd.h>
  55
  56 #include "defs.h"
  57 #include "extern.h"
  58
  59 /*
  60  * Linked list of units (strings and files) to be compiled
  61  */
  62 struct s_compunit {
  63         struct s_compunit *next;
  64         enum e_cut {CU_FILE, CU_STRING} type;
  65         char *s;                        /* Pointer to string or fname */
  66 };
  67
  68 /*
  69  * Linked list pointer to compilation units and pointer to current
  70  * next pointer.
  71  */
  72 static struct s_compunit *script, **cu_nextp = &script;
  73
  74 /*
  75  * Linked list of files to be processed
  76  */
  77 struct s_flist {
  78         char *fname;
  79         struct s_flist *next;
  80 };
  81
  82 /*
  83  * Linked list pointer to files and pointer to current
  84  * next pointer.
  85  */
  86 static struct s_flist *files, **fl_nextp = &files;
  87
  88 FILE *infile;                   /* Current input file */
  89 FILE *outfile;                  /* Current output file */
  90
  91 int aflag, eflag, nflag;
  92 int rflags = 0;
  93 static int rval;                /* Exit status */
  94
  95 static int ispan;               /* Whether inplace editing spans across files */
  96
  97 /*
  98  * Current file and line number; line numbers restart across compilation
  99  * units, but span across input files.  The latter is optional if editing
 100  * in place.
 101  */
 102 const char *fname;              /* File name. */
 103 const char *outfname;           /* Output file name */
 104 static char oldfname[PATH_MAX]; /* Old file name (for in-place editing) */
 105 static char tmpfname[PATH_MAX]; /* Temporary file name (for in-place editing) */
 106 static const char *inplace;     /* Inplace edit file extension. */
 107 ulong_t linenum;
 108
 109 static const struct option lopts[] = {
 110         {"in-place",    optional_argument,      NULL,   'i'},
 111         {NULL,          0,                      NULL,   0}
 112 };
 113
 114 static void add_compunit(enum e_cut, char *);
 115 static void add_file(char *);
 116 static void usage(void);
 117 static char *getln(FILE *, size_t *);
 118
 119
 120 int
 121 main(int argc, char *argv[])
 122 {
 123         int c, fflag;
 124         char *temp_arg;
 125
 126         (void) setlocale(LC_ALL, "");
 127
 128 #ifndef TEXT_DOMAIN
 129 #define TEXT_DOMAIN     "SYS_TEST"
 130 #endif
 131         (void) textdomain(TEXT_DOMAIN);
 132
 133         fflag = 0;
 134         inplace = NULL;
 135
 136         while ((c = getopt_long(argc, argv, "EI::ae:f:i::lnr", lopts, NULL)) !=
 137             -1)
 138                 switch (c) {
 139                 case 'r':               /* Gnu sed compat */
 140                 case 'E':
 141                         rflags = REG_EXTENDED;
 142                         break;
 143                 case 'I':
 144                         if (optarg != NULL)
 145                                 inplace = optarg;
 146                         else
 147                                 inplace = "";
 148                         ispan = 1;      /* span across input files */
 149                         break;
 150                 case 'a':
 151                         aflag = 1;
 152                         break;
 153                 case 'e':
 154                         eflag = 1;
 155                         if (asprintf(&temp_arg, "%s\n", optarg) < 1)
 156                                 err(1, "asprintf");
 157                         add_compunit(CU_STRING, temp_arg);
 158                         break;
 159                 case 'f':
 160                         fflag = 1;
 161                         add_compunit(CU_FILE, optarg);
 162                         break;
 163                 case 'i':
 164                         if (optarg != NULL)
 165                                 inplace = optarg;
 166                         else
 167                                 inplace = "";
 168                         ispan = 0;      /* don't span across input files */
 169                         break;
 170                 case 'l':
 171                         /* On SunOS, setlinebuf "returns no useful value */
 172                         (void) setlinebuf(stdout);
 173                         break;
 174                 case 'n':
 175                         nflag = 1;
 176                         break;
 177                 default:
 178                 case '?':
 179                         usage();
 180                 }
 181         argc -= optind;
 182         argv += optind;
 183
 184         /* First usage case; script is the first arg */
 185         if (!eflag && !fflag && *argv) {
 186                 add_compunit(CU_STRING, *argv);
 187                 argv++;
 188         }
 189
 190         compile();
 191
 192         /* Continue with first and start second usage */
 193         if (*argv)
 194                 for (; *argv; argv++)
 195                         add_file(*argv);
 196         else
 197                 add_file(NULL);
 198         process();
 199         cfclose(prog, NULL);
 200         if (fclose(stdout))
 201                 err(1, "stdout");
 202         return (rval);
 203 }
 204
 205 static void
 206 usage(void)
 207 {
 208         (void) fputs(_("usage: sed script [-Ealn] [-i[extension]] [file...]\n"
 209             "       sed [-Ealn] [-i[extension]] [-e script]... "
 210             "[-f script_file]... [file...]\n"),
 211             stderr);
 212         exit(1);
 213 }
 214
 215 /*
 216  * Like fgets, but go through the chain of compilation units chaining them
 217  * together.  Empty strings and files are ignored.
 218  */
 219 char *
 220 cu_fgets(char *buf, int n, int *more)
 221 {
 222         static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
 223         static FILE *f;         /* Current open file */
 224         static char *s;         /* Current pointer inside string */
 225         static char string_ident[30];
 226         char *p;
 227
 228 again:
 229         switch (state) {
 230         case ST_EOF:
 231                 if (script == NULL) {
 232                         if (more != NULL)
 233                                 *more = 0;
 234                         return (NULL);
 235                 }
 236                 linenum = 0;
 237                 switch (script->type) {
 238                 case CU_FILE:
 239                         if ((f = fopen(script->s, "r")) == NULL)
 240                                 err(1, "%s", script->s);
 241                         fname = script->s;
 242                         state = ST_FILE;
 243                         goto again;
 244                 case CU_STRING:
 245                         if (((size_t)snprintf(string_ident,
 246                             sizeof (string_ident), "\"%s\"", script->s)) >=
 247                             sizeof (string_ident) - 1)
 248                                 (void) strcpy(string_ident +
 249                                     sizeof (string_ident) - 6, " ...\"");
 250                         fname = string_ident;
 251                         s = script->s;
 252                         state = ST_STRING;
 253                         goto again;
 254                 }
 255                 /*NOTREACHED*/
 256
 257         case ST_FILE:
 258                 if ((p = fgets(buf, n, f)) != NULL) {
 259                         linenum++;
 260                         if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
 261                                 nflag = 1;
 262                         if (more != NULL)
 263                                 *more = !feof(f);
 264                         return (p);
 265                 }
 266                 script = script->next;
 267                 (void) fclose(f);
 268                 state = ST_EOF;
 269                 goto again;
 270         case ST_STRING:
 271                 if (linenum == 0 && s[0] == '#' && s[1] == 'n')
 272                         nflag = 1;
 273                 p = buf;
 274                 for (;;) {
 275                         if (n-- <= 1) {
 276                                 *p = '\0';
 277                                 linenum++;
 278                                 if (more != NULL)
 279                                         *more = 1;
 280                                 return (buf);
 281                         }
 282                         switch (*s) {
 283                         case '\0':
 284                                 state = ST_EOF;
 285                                 if (s == script->s) {
 286                                         script = script->next;
 287                                         goto again;
 288                                 } else {
 289                                         script = script->next;
 290                                         *p = '\0';
 291                                         linenum++;
 292                                         if (more != NULL)
 293                                                 *more = 0;
 294                                         return (buf);
 295                                 }
 296                         case '\n':
 297                                 *p++ = '\n';
 298                                 *p = '\0';
 299                                 s++;
 300                                 linenum++;
 301                                 if (more != NULL)
 302                                         *more = 0;
 303                                 return (buf);
 304                         default:
 305                                 *p++ = *s++;
 306                         }
 307                 }
 308         }
 309         /* NOTREACHED */
 310         return (NULL);
 311 }
 312
 313 /*
 314  * Like fgets, but go through the list of files chaining them together.
 315  * Set len to the length of the line.
 316  */
 317 int
 318 mf_fgets(SPACE *sp, enum e_spflag spflag)
 319 {
 320         struct stat sb, nsb;
 321         size_t len;
 322         char *p;
 323         int c;
 324         static int firstfile;
 325
 326         if (infile == NULL) {
 327                 /* stdin? */
 328                 if (files->fname == NULL) {
 329                         if (inplace != NULL)
 330                                 errx(1,
 331                                     _("-I or -i may not be used with stdin"));
 332                         infile = stdin;
 333                         fname = "stdin";
 334                         outfile = stdout;
 335                         outfname = "stdout";
 336                 }
 337                 firstfile = 1;
 338         }
 339
 340         for (;;) {
 341                 if (infile != NULL && (c = getc(infile)) != EOF) {
 342                         (void) ungetc(c, infile);
 343                         break;
 344                 }
 345                 /* If we are here then either eof or no files are open yet */
 346                 if (infile == stdin) {
 347                         sp->len = 0;
 348                         return (0);
 349                 }
 350                 if (infile != NULL) {
 351                         (void) fclose(infile);
 352                         if (*oldfname != '\0') {
 353                                 /* if there was a backup file, remove it */
 354                                 (void) unlink(oldfname);
 355                                 /*
 356                                  * Backup the original.  Note that hard links
 357                                  * are not supported on all filesystems.
 358                                  */
 359                                 if ((link(fname, oldfname) != 0) &&
 360                                     (rename(fname, oldfname) != 0)) {
 361                                         warn("rename()");
 362                                         if (*tmpfname)
 363                                                 (void) unlink(tmpfname);
 364                                         exit(1);
 365                                 }
 366                                 *oldfname = '\0';
 367                         }
 368                         if (*tmpfname != '\0') {
 369                                 if (outfile != NULL && outfile != stdout)
 370                                         if (fclose(outfile) != 0) {
 371                                                 warn("fclose()");
 372                                                 (void) unlink(tmpfname);
 373                                                 exit(1);
 374                                         }
 375                                 outfile = NULL;
 376                                 if (rename(tmpfname, fname) != 0) {
 377                                         /* this should not happen really! */
 378                                         warn("rename()");
 379                                         (void) unlink(tmpfname);
 380                                         exit(1);
 381                                 }
 382                                 *tmpfname = '\0';
 383                         }
 384                         outfname = NULL;
 385                 }
 386                 if (firstfile == 0)
 387                         files = files->next;
 388                 else
 389                         firstfile = 0;
 390                 if (files == NULL) {
 391                         sp->len = 0;
 392                         return (0);
 393                 }
 394                 fname = files->fname;
 395                 if (inplace != NULL) {
 396                         char bn[PATH_MAX];
 397                         char dn[PATH_MAX];
 398                         (void) strlcpy(bn, fname, sizeof (bn));
 399                         (void) strlcpy(dn, fname, sizeof (dn));
 400                         if (lstat(fname, &sb) != 0)
 401                                 err(1, "%s", fname);
 402                         if (!(sb.st_mode & S_IFREG))
 403                                 fatal(_("in-place editing only "
 404                                     "works for regular files"));
 405                         if (*inplace != '\0') {
 406                                 (void) strlcpy(oldfname, fname,
 407                                     sizeof (oldfname));
 408                                 len = strlcat(oldfname, inplace,
 409                                     sizeof (oldfname));
 410                                 if (len > sizeof (oldfname))
 411                                         fatal(_("name too long"));
 412                         }
 413                         len = snprintf(tmpfname, sizeof (tmpfname),
 414                             "%s/.!%ld!%s", dirname(dn), (long)getpid(),
 415                             basename(bn));
 416                         if (len >= sizeof (tmpfname))
 417                                 fatal(_("name too long"));
 418                         (void) unlink(tmpfname);
 419                         if ((outfile = fopen(tmpfname, "w")) == NULL)
 420                                 err(1, "%s", fname);
 421                         /*
 422                          * Some file systems don't support chown or
 423                          * chmod fully.  On those, the owner/group and
 424                          * permissions will already be set to what
 425                          * they need to be.
 426                          */
 427                         if (fstat(fileno(outfile), &nsb) != 0) {
 428                                 warn("fstat()");
 429                         }
 430                         if (((sb.st_uid != nsb.st_uid) ||
 431                             (sb.st_gid != nsb.st_gid)) &&
 432                             (fchown(fileno(outfile), sb.st_uid, sb.st_gid)
 433                             != 0))
 434                                 warn("fchown()");
 435                         if ((sb.st_mode != nsb.st_mode) &&
 436                             (fchmod(fileno(outfile), sb.st_mode & 07777) != 0))
 437                                 warn("fchmod()");
 438                         outfname = tmpfname;
 439                         if (!ispan) {
 440                                 linenum = 0;
 441                                 resetstate();
 442                         }
 443                 } else {
 444                         outfile = stdout;
 445                         outfname = "stdout";
 446                 }
 447                 if ((infile = fopen(fname, "r")) == NULL) {
 448                         warn("%s", fname);
 449                         rval = 1;
 450                         continue;
 451                 }
 452         }
 453         /*
 454          * We are here only when infile is open and we still have something
 455          * to read from it.
 456          *
 457          * Use fgetln so that we can handle essentially infinite input data.
 458          * Can't use the pointer into the stdio buffer as the process space
 459          * because the ungetc() can cause it to move.
 460          */
 461         p = getln(infile, &len);
 462         if (ferror(infile))
 463                 errx(1, "%s: %s", fname, strerror(errno ? errno : EIO));
 464         if (len != 0 && p[len - 1] == '\n')
 465                 len--;
 466         cspace(sp, p, len, spflag);
 467
 468         linenum++;
 469
 470         return (1);
 471 }
 472
 473 /*
 474  * Add a compilation unit to the linked list
 475  */
 476 static void
 477 add_compunit(enum e_cut type, char *s)
 478 {
 479         struct s_compunit *cu;
 480
 481         if ((cu = malloc(sizeof (struct s_compunit))) == NULL)
 482                 err(1, "malloc");
 483         cu->type = type;
 484         cu->s = s;
 485         cu->next = NULL;
 486         *cu_nextp = cu;
 487         cu_nextp = &cu->next;
 488 }
 489
 490 /*
 491  * Add a file to the linked list
 492  */
 493 static void
 494 add_file(char *s)
 495 {
 496         struct s_flist *fp;
 497
 498         if ((fp = malloc(sizeof (struct s_flist))) == NULL)
 499                 err(1, "malloc");
 500         fp->next = NULL;
 501         *fl_nextp = fp;
 502         fp->fname = s;
 503         fl_nextp = &fp->next;
 504 }
 505
 506 int
 507 lastline(void)
 508 {
 509         int ch;
 510
 511         if (files->next != NULL && (inplace == NULL || ispan))
 512                 return (0);
 513         if ((ch = getc(infile)) == EOF)
 514                 return (1);
 515         (void) ungetc(ch, infile);
 516         return (0);
 517 }
 518
 519 char *
 520 getln(FILE *in, size_t *lenp)
 521 {
 522         static char     *buffer = NULL;
 523         static size_t   sz = 0;
 524
 525         size_t          len = 0;
 526
 527         for (;;) {
 528                 if (sz <= (len + 1)) {
 529                         char *nb;
 530                         if ((nb = realloc(buffer, sz + LINE_MAX)) == NULL) {
 531                                 err(1, "realloc");
 532                         }
 533                         buffer = nb;
 534                         sz += LINE_MAX;
 535                 }
 536
 537                 buffer[len] = 0;
 538
 539                 if (fgets(buffer + len, sz - len, in) == NULL) {
 540                         /* END OF FILE */
 541                         *lenp = len;
 542                         break;
 543                 }
 544
 545                 len += strlen(buffer + len);
 546
 547                 if (buffer[len - 1] == '\n') {
 548                         /* got the new line */
 549                         *lenp = len;
 550                         break;
 551                 }
 552         }
 553
 554         return (buffer);
 555 }