usr.bin/sed/main.c

   1 /*-
   2  * Copyright (c) 2013 Johann 'Myrkraverk' Oskarsson.
   3  * Copyright (c) 1992 Diomidis Spinellis.
   4  * Copyright (c) 1992, 1993
   5  *      The Regents of the University of California.  All rights reserved.
   6  *
   7  * This code is derived from software contributed to Berkeley by
   8  * Diomidis Spinellis of Imperial College, University of London.
   9  *
  10  * Redistribution and use in source and binary forms, with or without
  11  * modification, are permitted provided that the following conditions
  12  * are met:
  13  * 1. Redistributions of source code must retain the above copyright
  14  *    notice, this list of conditions and the following disclaimer.
  15  * 2. Redistributions in binary form must reproduce the above copyright
  16  *    notice, this list of conditions and the following disclaimer in the
  17  *    documentation and/or other materials provided with the distribution.
  18  * 4. Neither the name of the University nor the names of its contributors
  19  *    may be used to endorse or promote products derived from this software
  20  *    without specific prior written permission.
  21  *
  22  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  23  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  24  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  25  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  26  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  27  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  28  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  29  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  30  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  31  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  32  * SUCH DAMAGE.
  33  */
  34
  35 #include <sys/cdefs.h>
  36 __FBSDID("$FreeBSD$");
  37
  38 #ifndef lint
  39 static const char copyright[] =
  40 "@(#) Copyright (c) 1992, 1993\n\
  41         The Regents of the University of California.  All rights reserved.\n";
  42 #endif
  43
  44 #ifndef lint
  45 static const char sccsid[] = "@(#)main.c        8.2 (Berkeley) 1/3/94";
  46 #endif
  47
  48 #include <sys/types.h>
  49 #include <sys/mman.h>
  50 #include <sys/param.h>
  51 #include <sys/stat.h>
  52
  53 #include <err.h>
  54 #include <errno.h>
  55 #include <fcntl.h>
  56 #include <libgen.h>
  57 #include <limits.h>
  58 #include <locale.h>
  59 #include <regex.h>
  60 #include <stddef.h>
  61 #define _WITH_GETLINE
  62 #include <stdio.h>
  63 #include <stdlib.h>
  64 #include <string.h>
  65 #include <unistd.h>
  66
  67 #include "defs.h"
  68 #include "extern.h"
  69
  70 /*
  71  * Linked list of units (strings and files) to be compiled
  72  */
  73 struct s_compunit {
  74         struct s_compunit *next;
  75         enum e_cut {CU_FILE, CU_STRING} type;
  76         char *s;                        /* Pointer to string or fname */
  77 };
  78
  79 /*
  80  * Linked list pointer to compilation units and pointer to current
  81  * next pointer.
  82  */
  83 static struct s_compunit *script, **cu_nextp = &script;
  84
  85 /*
  86  * Linked list of files to be processed
  87  */
  88 struct s_flist {
  89         char *fname;
  90         struct s_flist *next;
  91 };
  92
  93 /*
  94  * Linked list pointer to files and pointer to current
  95  * next pointer.
  96  */
  97 static struct s_flist *files, **fl_nextp = &files;
  98
  99 FILE *infile;                   /* Current input file */
 100 FILE *outfile;                  /* Current output file */
 101
 102 int aflag, eflag, nflag;
 103 int rflags = 0;
 104 static int rval;                /* Exit status */
 105
 106 static int ispan;               /* Whether inplace editing spans across files */
 107
 108 /*
 109  * Current file and line number; line numbers restart across compilation
 110  * units, but span across input files.  The latter is optional if editing
 111  * in place.
 112  */
 113 const char *fname;              /* File name. */
 114 const char *outfname;           /* Output file name */
 115 static char oldfname[PATH_MAX]; /* Old file name (for in-place editing) */
 116 static char tmpfname[PATH_MAX]; /* Temporary file name (for in-place editing) */
 117 static const char *inplace;     /* Inplace edit file extension. */
 118 u_long linenum;
 119
 120 static void add_compunit(enum e_cut, char *);
 121 static void add_file(char *);
 122 static void usage(void);
 123
 124 int
 125 main(int argc, char *argv[])
 126 {
 127         int c, fflag;
 128         char *temp_arg;
 129
 130         (void) setlocale(LC_ALL, "");
 131
 132         fflag = 0;
 133         inplace = NULL;
 134
 135         while ((c = getopt(argc, argv, "EI:ae:f:i:lnru")) != -1)
 136                 switch (c) {
 137                 case 'r':               /* Gnu sed compat */
 138                 case 'E':
 139                         rflags = REG_EXTENDED;
 140                         break;
 141                 case 'I':
 142                         inplace = optarg;
 143                         ispan = 1;      /* span across input files */
 144                         break;
 145                 case 'a':
 146                         aflag = 1;
 147                         break;
 148                 case 'e':
 149                         eflag = 1;
 150                         if ((temp_arg = malloc(strlen(optarg) + 2)) == NULL)
 151                                 err(1, "malloc");
 152                         strcpy(temp_arg, optarg);
 153                         strcat(temp_arg, "\n");
 154                         add_compunit(CU_STRING, temp_arg);
 155                         break;
 156                 case 'f':
 157                         fflag = 1;
 158                         add_compunit(CU_FILE, optarg);
 159                         break;
 160                 case 'i':
 161                         inplace = optarg;
 162                         ispan = 0;      /* don't span across input files */
 163                         break;
 164                 case 'l':
 165                         if(setvbuf(stdout, NULL, _IOLBF, 0) != 0)
 166                                 warnx("setting line buffered output failed");
 167                         break;
 168                 case 'n':
 169                         nflag = 1;
 170                         break;
 171                 case 'u':
 172                         if(setvbuf(stdout, NULL, _IONBF, 0) != 0)
 173                                 warnx("setting unbuffered output failed");
 174                         break;
 175                 default:
 176                 case '?':
 177                         usage();
 178                 }
 179         argc -= optind;
 180         argv += optind;
 181
 182         /* First usage case; script is the first arg */
 183         if (!eflag && !fflag && *argv) {
 184                 add_compunit(CU_STRING, *argv);
 185                 argv++;
 186         }
 187
 188         compile();
 189
 190         /* Continue with first and start second usage */
 191         if (*argv)
 192                 for (; *argv; argv++)
 193                         add_file(*argv);
 194         else
 195                 add_file(NULL);
 196         process();
 197         cfclose(prog, NULL);
 198         if (fclose(stdout))
 199                 err(1, "stdout");
 200         exit(rval);
 201 }
 202
 203 static void
 204 usage(void)
 205 {
 206         (void)fprintf(stderr,
 207             "usage: %s script [-Ealnru] [-i extension] [file ...]\n"
 208             "\t%s [-Ealnu] [-i extension] [-e script] ... [-f script_file]"
 209             " ... [file ...]\n", getprogname(), getprogname());
 210         exit(1);
 211 }
 212
 213 /*
 214  * Like fgets, but go through the chain of compilation units chaining them
 215  * together.  Empty strings and files are ignored.
 216  */
 217 char *
 218 cu_fgets(char *buf, int n, int *more)
 219 {
 220         static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
 221         static FILE *f;         /* Current open file */
 222         static char *s;         /* Current pointer inside string */
 223         static char string_ident[30];
 224         char *p;
 225
 226 again:
 227         switch (state) {
 228         case ST_EOF:
 229                 if (script == NULL) {
 230                         if (more != NULL)
 231                                 *more = 0;
 232                         return (NULL);
 233                 }
 234                 linenum = 0;
 235                 switch (script->type) {
 236                 case CU_FILE:
 237                         if ((f = fopen(script->s, "r")) == NULL)
 238                                 err(1, "%s", script->s);
 239                         fname = script->s;
 240                         state = ST_FILE;
 241                         goto again;
 242                 case CU_STRING:
 243                         if (((size_t)snprintf(string_ident,
 244                             sizeof(string_ident), "\"%s\"", script->s)) >=
 245                             sizeof(string_ident) - 1)
 246                                 (void)strcpy(string_ident +
 247                                     sizeof(string_ident) - 6, " ...\"");
 248                         fname = string_ident;
 249                         s = script->s;
 250                         state = ST_STRING;
 251                         goto again;
 252                 }
 253         case ST_FILE:
 254                 if ((p = fgets(buf, n, f)) != NULL) {
 255                         linenum++;
 256                         if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
 257                                 nflag = 1;
 258                         if (more != NULL)
 259                                 *more = !feof(f);
 260                         return (p);
 261                 }
 262                 script = script->next;
 263                 (void)fclose(f);
 264                 state = ST_EOF;
 265                 goto again;
 266         case ST_STRING:
 267                 if (linenum == 0 && s[0] == '#' && s[1] == 'n')
 268                         nflag = 1;
 269                 p = buf;
 270                 for (;;) {
 271                         if (n-- <= 1) {
 272                                 *p = '\0';
 273                                 linenum++;
 274                                 if (more != NULL)
 275                                         *more = 1;
 276                                 return (buf);
 277                         }
 278                         switch (*s) {
 279                         case '\0':
 280                                 state = ST_EOF;
 281                                 if (s == script->s) {
 282                                         script = script->next;
 283                                         goto again;
 284                                 } else {
 285                                         script = script->next;
 286                                         *p = '\0';
 287                                         linenum++;
 288                                         if (more != NULL)
 289                                                 *more = 0;
 290                                         return (buf);
 291                                 }
 292                         case '\n':
 293                                 *p++ = '\n';
 294                                 *p = '\0';
 295                                 s++;
 296                                 linenum++;
 297                                 if (more != NULL)
 298                                         *more = 0;
 299                                 return (buf);
 300                         default:
 301                                 *p++ = *s++;
 302                         }
 303                 }
 304         }
 305         /* NOTREACHED */
 306         return (NULL);
 307 }
 308
 309 /*
 310  * Like fgets, but go through the list of files chaining them together.
 311  * Set len to the length of the line.
 312  */
 313 int
 314 mf_fgets(SPACE *sp, enum e_spflag spflag)
 315 {
 316         struct stat sb;
 317         ssize_t len;
 318         static char *p = NULL;
 319         static size_t plen = 0;
 320         int c;
 321         static int firstfile;
 322
 323         if (infile == NULL) {
 324                 /* stdin? */
 325                 if (files->fname == NULL) {
 326                         if (inplace != NULL)
 327                                 errx(1, "-I or -i may not be used with stdin");
 328                         infile = stdin;
 329                         fname = "stdin";
 330                         outfile = stdout;
 331                         outfname = "stdout";
 332                 }
 333                 firstfile = 1;
 334         }
 335
 336         for (;;) {
 337                 if (infile != NULL && (c = getc(infile)) != EOF) {
 338                         (void)ungetc(c, infile);
 339                         break;
 340                 }
 341                 /* If we are here then either eof or no files are open yet */
 342                 if (infile == stdin) {
 343                         sp->len = 0;
 344                         return (0);
 345                 }
 346                 if (infile != NULL) {
 347                         fclose(infile);
 348                         if (*oldfname != '\0') {
 349                                 /* if there was a backup file, remove it */
 350                                 unlink(oldfname);
 351                                 /*
 352                                  * Backup the original.  Note that hard links
 353                                  * are not supported on all filesystems.
 354                                  */
 355                                 if ((link(fname, oldfname) != 0) &&
 356                                    (rename(fname, oldfname) != 0)) {
 357                                         warn("rename()");
 358                                         if (*tmpfname)
 359                                                 unlink(tmpfname);
 360                                         exit(1);
 361                                 }
 362                                 *oldfname = '\0';
 363                         }
 364                         if (*tmpfname != '\0') {
 365                                 if (outfile != NULL && outfile != stdout)
 366                                         if (fclose(outfile) != 0) {
 367                                                 warn("fclose()");
 368                                                 unlink(tmpfname);
 369                                                 exit(1);
 370                                         }
 371                                 outfile = NULL;
 372                                 if (rename(tmpfname, fname) != 0) {
 373                                         /* this should not happen really! */
 374                                         warn("rename()");
 375                                         unlink(tmpfname);
 376                                         exit(1);
 377                                 }
 378                                 *tmpfname = '\0';
 379                         }
 380                         outfname = NULL;
 381                 }
 382                 if (firstfile == 0)
 383                         files = files->next;
 384                 else
 385                         firstfile = 0;
 386                 if (files == NULL) {
 387                         sp->len = 0;
 388                         return (0);
 389                 }
 390                 fname = files->fname;
 391                 if (inplace != NULL) {
 392                         if (lstat(fname, &sb) != 0)
 393                                 err(1, "%s", fname);
 394                         if (!(sb.st_mode & S_IFREG))
 395                                 errx(1, "%s: %s %s", fname,
 396                                     "in-place editing only",
 397                                     "works for regular files");
 398                         if (*inplace != '\0') {
 399                                 strlcpy(oldfname, fname,
 400                                     sizeof(oldfname));
 401                                 len = strlcat(oldfname, inplace,
 402                                     sizeof(oldfname));
 403                                 if (len > (ssize_t)sizeof(oldfname))
 404                                         errx(1, "%s: name too long", fname);
 405                         }
 406                         len = snprintf(tmpfname, sizeof(tmpfname),
 407                             "%s/.!%ld!%s", dirname(fname), (long)getpid(),
 408                             basename(fname));
 409                         if (len >= (ssize_t)sizeof(tmpfname))
 410                                 errx(1, "%s: name too long", fname);
 411                         unlink(tmpfname);
 412                         if (outfile != NULL && outfile != stdout)
 413                                 fclose(outfile);
 414                         if ((outfile = fopen(tmpfname, "w")) == NULL)
 415                                 err(1, "%s", fname);
 416                         fchown(fileno(outfile), sb.st_uid, sb.st_gid);
 417                         fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
 418                         outfname = tmpfname;
 419                         if (!ispan) {
 420                                 linenum = 0;
 421                                 resetstate();
 422                         }
 423                 } else {
 424                         outfile = stdout;
 425                         outfname = "stdout";
 426                 }
 427                 if ((infile = fopen(fname, "r")) == NULL) {
 428                         warn("%s", fname);
 429                         rval = 1;
 430                         continue;
 431                 }
 432         }
 433         /*
 434          * We are here only when infile is open and we still have something
 435          * to read from it.
 436          *
 437          * Use getline() so that we can handle essentially infinite input
 438          * data.  The p and plen are static so each invocation gives
 439          * getline() the same buffer which is expanded as needed.
 440          */
 441         len = getline(&p, &plen, infile);
 442         if (len == -1)
 443                 err(1, "%s", fname);
 444         if (len != 0 && p[len - 1] == '\n') {
 445                 sp->append_newline = 1;
 446                 len--;
 447         } else if (!lastline()) {
 448                 sp->append_newline = 1;
 449         } else {
 450                 sp->append_newline = 0;
 451         }
 452         cspace(sp, p, len, spflag);
 453
 454         linenum++;
 455
 456         return (1);
 457 }
 458
 459 /*
 460  * Add a compilation unit to the linked list
 461  */
 462 static void
 463 add_compunit(enum e_cut type, char *s)
 464 {
 465         struct s_compunit *cu;
 466
 467         if ((cu = malloc(sizeof(struct s_compunit))) == NULL)
 468                 err(1, "malloc");
 469         cu->type = type;
 470         cu->s = s;
 471         cu->next = NULL;
 472         *cu_nextp = cu;
 473         cu_nextp = &cu->next;
 474 }
 475
 476 /*
 477  * Add a file to the linked list
 478  */
 479 static void
 480 add_file(char *s)
 481 {
 482         struct s_flist *fp;
 483
 484         if ((fp = malloc(sizeof(struct s_flist))) == NULL)
 485                 err(1, "malloc");
 486         fp->next = NULL;
 487         *fl_nextp = fp;
 488         fp->fname = s;
 489         fl_nextp = &fp->next;
 490 }
 491
 492 static int
 493 next_files_have_lines(void)
 494 {
 495         struct s_flist *file;
 496         FILE *file_fd;
 497         int ch;
 498
 499         file = files;
 500         while ((file = file->next) != NULL) {
 501                 if ((file_fd = fopen(file->fname, "r")) == NULL)
 502                         continue;
 503
 504                 if ((ch = getc(file_fd)) != EOF) {
 505                         /*
 506                          * This next file has content, therefore current
 507                          * file doesn't contains the last line.
 508                          */
 509                         ungetc(ch, file_fd);
 510                         fclose(file_fd);
 511                         return (1);
 512                 }
 513
 514                 fclose(file_fd);
 515         }
 516
 517         return (0);
 518 }
 519
 520 int
 521 lastline(void)
 522 {
 523         int ch;
 524
 525         if (feof(infile)) {
 526                 return !(
 527                     (inplace == NULL || ispan) &&
 528                     next_files_have_lines());
 529         }
 530         if ((ch = getc(infile)) == EOF) {
 531                 return !(
 532                     (inplace == NULL || ispan) &&
 533                     next_files_have_lines());
 534         }
 535         ungetc(ch, infile);
 536         return (0);
 537 }