usr.bin/sed/main.c

   1 /*-
   2  * SPDX-License-Identifier: BSD-3-Clause
   3  *
   4  * Copyright (c) 2013 Johann 'Myrkraverk' Oskarsson.
   5  * Copyright (c) 1992 Diomidis Spinellis.
   6  * Copyright (c) 1992, 1993
   7  *      The Regents of the University of California.  All rights reserved.
   8  *
   9  * This code is derived from software contributed to Berkeley by
  10  * Diomidis Spinellis of Imperial College, University of London.
  11  *
  12  * Redistribution and use in source and binary forms, with or without
  13  * modification, are permitted provided that the following conditions
  14  * are met:
  15  * 1. Redistributions of source code must retain the above copyright
  16  *    notice, this list of conditions and the following disclaimer.
  17  * 2. Redistributions in binary form must reproduce the above copyright
  18  *    notice, this list of conditions and the following disclaimer in the
  19  *    documentation and/or other materials provided with the distribution.
  20  * 3. 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  * @(#) Copyright (c) 1992, 1993 The Regents of the University of California.  All rights reserved.
  37  * @(#)main.c   8.2 (Berkeley) 1/3/94
  38  * $FreeBSD: head/usr.bin/sed/main.c 362017 2020-06-10 19:23:58Z 0mp $
  39  */
  40
  41 #include <sys/cdefs.h>
  42
  43 #include <sys/types.h>
  44 #include <sys/mman.h>
  45 #include <sys/param.h>
  46 #include <sys/stat.h>
  47
  48 #include <err.h>
  49 #include <errno.h>
  50 #include <fcntl.h>
  51 #include <libgen.h>
  52 #include <limits.h>
  53 #include <locale.h>
  54 #include <regex.h>
  55 #include <stddef.h>
  56 #include <stdio.h>
  57 #include <stdlib.h>
  58 #include <string.h>
  59 #include <unistd.h>
  60
  61 #include "defs.h"
  62 #include "extern.h"
  63
  64 /*
  65  * Linked list of units (strings and files) to be compiled
  66  */
  67 struct s_compunit {
  68         struct s_compunit *next;
  69         enum e_cut {CU_FILE, CU_STRING} type;
  70         char *s;                        /* Pointer to string or fname */
  71 };
  72
  73 /*
  74  * Linked list pointer to compilation units and pointer to current
  75  * next pointer.
  76  */
  77 static struct s_compunit *script, **cu_nextp = &script;
  78
  79 /*
  80  * Linked list of files to be processed
  81  */
  82 struct s_flist {
  83         char *fname;
  84         struct s_flist *next;
  85 };
  86
  87 /*
  88  * Linked list pointer to files and pointer to current
  89  * next pointer.
  90  */
  91 static struct s_flist *files, **fl_nextp = &files;
  92
  93 FILE *infile;                   /* Current input file */
  94 FILE *outfile;                  /* Current output file */
  95
  96 int aflag, eflag, nflag;
  97 int rflags = 0;
  98 int quit = 0;
  99 static int rval;                /* Exit status */
 100
 101 static int ispan;               /* Whether inplace editing spans across files */
 102
 103 /*
 104  * Current file and line number; line numbers restart across compilation
 105  * units, but span across input files.  The latter is optional if editing
 106  * in place.
 107  */
 108 const char *fname;              /* File name. */
 109 const char *outfname;           /* Output file name */
 110 static char oldfname[PATH_MAX]; /* Old file name (for in-place editing) */
 111 static char tmpfname[PATH_MAX]; /* Temporary file name (for in-place editing) */
 112 const char *inplace;            /* Inplace edit file extension. */
 113 u_long linenum;
 114
 115 static void add_compunit(enum e_cut, char *);
 116 static void add_file(char *);
 117 static void usage(void);
 118
 119 int
 120 main(int argc, char *argv[])
 121 {
 122         int c, fflag, fflagstdin;
 123         char *temp_arg;
 124
 125         (void) setlocale(LC_ALL, "");
 126
 127         fflag = 0;
 128         fflagstdin = 0;
 129         inplace = NULL;
 130
 131         while ((c = getopt(argc, argv, "EI:ae:f:i:lnru")) != -1)
 132                 switch (c) {
 133                 case 'r':               /* GNU sed compat */
 134                 case 'E':
 135                         rflags = REG_EXTENDED;
 136                         break;
 137                 case 'I':
 138                         inplace = optarg;
 139                         ispan = 1;      /* span across input files */
 140                         break;
 141                 case 'a':
 142                         aflag = 1;
 143                         break;
 144                 case 'e':
 145                         eflag = 1;
 146                         if ((temp_arg = malloc(strlen(optarg) + 2)) == NULL)
 147                                 err(1, "malloc");
 148                         strcpy(temp_arg, optarg);
 149                         strcat(temp_arg, "\n");
 150                         add_compunit(CU_STRING, temp_arg);
 151                         break;
 152                 case 'f':
 153                         fflag = 1;
 154                         if (strcmp(optarg, "-") == 0)
 155                                 fflagstdin = 1;
 156                         add_compunit(CU_FILE, optarg);
 157                         break;
 158                 case 'i':
 159                         inplace = optarg;
 160                         ispan = 0;      /* don't span across input files */
 161                         break;
 162                 case 'l':
 163                         if(setvbuf(stdout, NULL, _IOLBF, 0) != 0)
 164                                 warnx("setting line buffered output failed");
 165                         break;
 166                 case 'n':
 167                         nflag = 1;
 168                         break;
 169                 case 'u':
 170                         if(setvbuf(stdout, NULL, _IONBF, 0) != 0)
 171                                 warnx("setting unbuffered output failed");
 172                         break;
 173                 default:
 174                 case '?':
 175                         usage();
 176                 }
 177         argc -= optind;
 178         argv += optind;
 179
 180         /* First usage case; script is the first arg */
 181         if (!eflag && !fflag && *argv) {
 182                 add_compunit(CU_STRING, *argv);
 183                 argv++;
 184         }
 185
 186         compile();
 187
 188         /* Continue with first and start second usage */
 189         if (*argv)
 190                 for (; *argv; argv++)
 191                         add_file(*argv);
 192         else if (fflagstdin)
 193                 exit(rval);
 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 (strcmp(script->s, "-") == 0) {
 238                                 f = stdin;
 239                                 fname = "stdin";
 240                         } else {
 241                                 if ((f = fopen(script->s, "r")) == NULL)
 242                                         err(1, "%s", script->s);
 243                                 fname = script->s;
 244                         }
 245                         state = ST_FILE;
 246                         goto again;
 247                 case CU_STRING:
 248                         if (((size_t)snprintf(string_ident,
 249                             sizeof(string_ident), "\"%s\"", script->s)) >=
 250                             sizeof(string_ident) - 1)
 251                                 (void)strcpy(string_ident +
 252                                     sizeof(string_ident) - 6, " ...\"");
 253                         fname = string_ident;
 254                         s = script->s;
 255                         state = ST_STRING;
 256                         goto again;
 257                 default:
 258                         errx(1, "illegal script type: %d", script->type);
 259                 }
 260         case ST_FILE:
 261                 if ((p = fgets(buf, n, f)) != NULL) {
 262                         linenum++;
 263                         if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
 264                                 nflag = 1;
 265                         if (more != NULL)
 266                                 *more = !feof(f);
 267                         return (p);
 268                 }
 269                 script = script->next;
 270                 (void)fclose(f);
 271                 state = ST_EOF;
 272                 goto again;
 273         case ST_STRING:
 274                 if (linenum == 0 && s[0] == '#' && s[1] == 'n')
 275                         nflag = 1;
 276                 p = buf;
 277                 for (;;) {
 278                         if (n-- <= 1) {
 279                                 *p = '\0';
 280                                 linenum++;
 281                                 if (more != NULL)
 282                                         *more = 1;
 283                                 return (buf);
 284                         }
 285                         switch (*s) {
 286                         case '\0':
 287                                 state = ST_EOF;
 288                                 if (s == script->s) {
 289                                         script = script->next;
 290                                         goto again;
 291                                 } else {
 292                                         script = script->next;
 293                                         *p = '\0';
 294                                         linenum++;
 295                                         if (more != NULL)
 296                                                 *more = 0;
 297                                         return (buf);
 298                                 }
 299                         case '\n':
 300                                 *p++ = '\n';
 301                                 *p = '\0';
 302                                 s++;
 303                                 linenum++;
 304                                 if (more != NULL)
 305                                         *more = 0;
 306                                 return (buf);
 307                         default:
 308                                 *p++ = *s++;
 309                         }
 310                 }
 311         }
 312         /* NOTREACHED */
 313         return (NULL);
 314 }
 315
 316 /*
 317  * Like fgets, but go through the list of files chaining them together.
 318  * Set len to the length of the line.
 319  */
 320 int
 321 mf_fgets(SPACE *sp, enum e_spflag spflag)
 322 {
 323         struct stat sb;
 324         ssize_t len;
 325         char *dirbuf, *basebuf;
 326         static char *p = NULL;
 327         static size_t plen = 0;
 328         int c;
 329         static int firstfile;
 330
 331         if (infile == NULL) {
 332                 /* stdin? */
 333                 if (files->fname == NULL) {
 334                         if (inplace != NULL)
 335                                 errx(1, "-I or -i may not be used with stdin");
 336                         infile = stdin;
 337                         fname = "stdin";
 338                         outfile = stdout;
 339                         outfname = "stdout";
 340                 }
 341                 firstfile = 1;
 342         }
 343
 344         for (;;) {
 345                 if (infile != NULL && (c = getc(infile)) != EOF && !quit) {
 346                         (void)ungetc(c, infile);
 347                         break;
 348                 }
 349                 /* If we are here then either eof or no files are open yet */
 350                 if (infile == stdin) {
 351                         sp->len = 0;
 352                         return (0);
 353                 }
 354                 if (infile != NULL) {
 355                         fclose(infile);
 356                         if (*oldfname != '\0') {
 357                                 /* if there was a backup file, remove it */
 358                                 unlink(oldfname);
 359                                 /*
 360                                  * Backup the original.  Note that hard links
 361                                  * are not supported on all filesystems.
 362                                  */
 363                                 if ((link(fname, oldfname) != 0) &&
 364                                    (rename(fname, oldfname) != 0)) {
 365                                         warn("rename()");
 366                                         if (*tmpfname)
 367                                                 unlink(tmpfname);
 368                                         exit(1);
 369                                 }
 370                                 *oldfname = '\0';
 371                         }
 372                         if (*tmpfname != '\0') {
 373                                 if (outfile != NULL && outfile != stdout)
 374                                         if (fclose(outfile) != 0) {
 375                                                 warn("fclose()");
 376                                                 unlink(tmpfname);
 377                                                 exit(1);
 378                                         }
 379                                 outfile = NULL;
 380                                 if (rename(tmpfname, fname) != 0) {
 381                                         /* this should not happen really! */
 382                                         warn("rename()");
 383                                         unlink(tmpfname);
 384                                         exit(1);
 385                                 }
 386                                 *tmpfname = '\0';
 387                         }
 388                         outfname = NULL;
 389                 }
 390                 if (firstfile == 0)
 391                         files = files->next;
 392                 else
 393                         firstfile = 0;
 394                 if (files == NULL) {
 395                         sp->len = 0;
 396                         return (0);
 397                 }
 398                 fname = files->fname;
 399                 if (inplace != NULL) {
 400                         if (lstat(fname, &sb) != 0)
 401                                 err(1, "%s", fname);
 402                         if (!S_ISREG(sb.st_mode))
 403                                 errx(1, "%s: %s %s", fname,
 404                                     "in-place editing only",
 405                                     "works for regular files");
 406                         if (*inplace != '\0') {
 407                                 strlcpy(oldfname, fname,
 408                                     sizeof(oldfname));
 409                                 len = strlcat(oldfname, inplace,
 410                                     sizeof(oldfname));
 411                                 if (len > (ssize_t)sizeof(oldfname))
 412                                         errx(1, "%s: name too long", fname);
 413                         }
 414                         if ((dirbuf = strdup(fname)) == NULL ||
 415                             (basebuf = strdup(fname)) == NULL)
 416                                 err(1, "strdup");
 417                         len = snprintf(tmpfname, sizeof(tmpfname),
 418                             "%s/.!%ld!%s", dirname(dirbuf), (long)getpid(),
 419                             basename(basebuf));
 420                         free(dirbuf);
 421                         free(basebuf);
 422                         if (len >= (ssize_t)sizeof(tmpfname))
 423                                 errx(1, "%s: name too long", fname);
 424                         unlink(tmpfname);
 425                         if (outfile != NULL && outfile != stdout)
 426                                 fclose(outfile);
 427                         if ((outfile = fopen(tmpfname, "w")) == NULL)
 428                                 err(1, "%s", fname);
 429                         fchown(fileno(outfile), sb.st_uid, sb.st_gid);
 430                         fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
 431                         outfname = tmpfname;
 432                         if (!ispan) {
 433                                 linenum = 0;
 434                                 resetstate();
 435                         }
 436                 } else {
 437                         outfile = stdout;
 438                         outfname = "stdout";
 439                 }
 440                 if ((infile = fopen(fname, "r")) == NULL) {
 441                         warn("%s", fname);
 442                         rval = 1;
 443                         continue;
 444                 }
 445         }
 446         /*
 447          * We are here only when infile is open and we still have something
 448          * to read from it.
 449          *
 450          * Use getline() so that we can handle essentially infinite input
 451          * data.  The p and plen are static so each invocation gives
 452          * getline() the same buffer which is expanded as needed.
 453          */
 454         len = getline(&p, &plen, infile);
 455         if (len == -1)
 456                 err(1, "%s", fname);
 457         if (len != 0 && p[len - 1] == '\n') {
 458                 sp->append_newline = 1;
 459                 len--;
 460         } else if (!lastline()) {
 461                 sp->append_newline = 1;
 462         } else {
 463                 sp->append_newline = 0;
 464         }
 465         cspace(sp, p, len, spflag);
 466
 467         linenum++;
 468
 469         return (1);
 470 }
 471
 472 /*
 473  * Add a compilation unit to the linked list
 474  */
 475 static void
 476 add_compunit(enum e_cut type, char *s)
 477 {
 478         struct s_compunit *cu;
 479
 480         if ((cu = malloc(sizeof(struct s_compunit))) == NULL)
 481                 err(1, "malloc");
 482         cu->type = type;
 483         cu->s = s;
 484         cu->next = NULL;
 485         *cu_nextp = cu;
 486         cu_nextp = &cu->next;
 487 }
 488
 489 /*
 490  * Add a file to the linked list
 491  */
 492 static void
 493 add_file(char *s)
 494 {
 495         struct s_flist *fp;
 496
 497         if ((fp = malloc(sizeof(struct s_flist))) == NULL)
 498                 err(1, "malloc");
 499         fp->next = NULL;
 500         *fl_nextp = fp;
 501         fp->fname = s;
 502         fl_nextp = &fp->next;
 503 }
 504
 505 static int
 506 next_files_have_lines(void)
 507 {
 508         struct s_flist *file;
 509         FILE *file_fd;
 510         int ch;
 511
 512         file = files;
 513         while ((file = file->next) != NULL) {
 514                 if ((file_fd = fopen(file->fname, "r")) == NULL)
 515                         continue;
 516
 517                 if ((ch = getc(file_fd)) != EOF) {
 518                         /*
 519                          * This next file has content, therefore current
 520                          * file doesn't contains the last line.
 521                          */
 522                         ungetc(ch, file_fd);
 523                         fclose(file_fd);
 524                         return (1);
 525                 }
 526
 527                 fclose(file_fd);
 528         }
 529
 530         return (0);
 531 }
 532
 533 int
 534 lastline(void)
 535 {
 536         int ch;
 537
 538         if (feof(infile)) {
 539                 return !(
 540                     (inplace == NULL || ispan) &&
 541                     next_files_have_lines());
 542         }
 543         if ((ch = getc(infile)) == EOF) {
 544                 return !(
 545                     (inplace == NULL || ispan) &&
 546                     next_files_have_lines());
 547         }
 548         ungetc(ch, infile);
 549         return (0);
 550 }