text_cmds/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
  37 #include <sys/cdefs.h>
  38 __FBSDID("$FreeBSD$");
  39
  40 #ifndef lint
  41 static const char copyright[] =
  42 "@(#) Copyright (c) 1992, 1993\n\
  43         The Regents of the University of California.  All rights reserved.\n";
  44 #endif
  45
  46 #ifndef lint
  47 static const char sccsid[] = "@(#)main.c        8.2 (Berkeley) 1/3/94";
  48 #endif
  49
  50 #include <sys/types.h>
  51 #include <sys/mman.h>
  52 #include <sys/param.h>
  53 #include <sys/stat.h>
  54
  55 #include <err.h>
  56 #include <errno.h>
  57 #include <fcntl.h>
  58 #include <libgen.h>
  59 #include <limits.h>
  60 #include <locale.h>
  61 #include <regex.h>
  62 #include <stddef.h>
  63 #include <stdio.h>
  64 #include <stdlib.h>
  65 #include <string.h>
  66 #include <unistd.h>
  67
  68 #include "defs.h"
  69 #include "extern.h"
  70
  71 /*
  72  * Linked list of units (strings and files) to be compiled
  73  */
  74 struct s_compunit {
  75         struct s_compunit *next;
  76         enum e_cut {CU_FILE, CU_STRING} type;
  77         char *s;                        /* Pointer to string or fname */
  78 };
  79
  80 /*
  81  * Linked list pointer to compilation units and pointer to current
  82  * next pointer.
  83  */
  84 static struct s_compunit *script, **cu_nextp = &script;
  85
  86 /*
  87  * Linked list of files to be processed
  88  */
  89 struct s_flist {
  90         char *fname;
  91         struct s_flist *next;
  92 };
  93
  94 /*
  95  * Linked list pointer to files and pointer to current
  96  * next pointer.
  97  */
  98 static struct s_flist *files, **fl_nextp = &files;
  99
 100 FILE *infile;                   /* Current input file */
 101 FILE *outfile;                  /* Current output file */
 102
 103 int aflag, eflag, nflag;
 104 int rflags = 0;
 105 int quit = 0;
 106 static int rval;                /* Exit status */
 107
 108 static int ispan;               /* Whether inplace editing spans across files */
 109
 110 /*
 111  * Current file and line number; line numbers restart across compilation
 112  * units, but span across input files.  The latter is optional if editing
 113  * in place.
 114  */
 115 const char *fname;              /* File name. */
 116 const char *outfname;           /* Output file name */
 117 static char oldfname[PATH_MAX]; /* Old file name (for in-place editing) */
 118 static char tmpfname[PATH_MAX]; /* Temporary file name (for in-place editing) */
 119 const char *inplace;            /* Inplace edit file extension. */
 120 u_long linenum;
 121
 122 static void add_compunit(enum e_cut, char *);
 123 static void add_file(char *);
 124 static void usage(void);
 125
 126 int
 127 main(int argc, char *argv[])
 128 {
 129         int c, fflag;
 130         char *temp_arg;
 131
 132         (void) setlocale(LC_ALL, "");
 133
 134         fflag = 0;
 135         inplace = NULL;
 136
 137         while ((c = getopt(argc, argv, "EI:ae:f:i:lnru")) != -1)
 138                 switch (c) {
 139                 case 'r':               /* Gnu sed compat */
 140                 case 'E':
 141                         rflags = REG_EXTENDED;
 142                         break;
 143                 case 'I':
 144                         inplace = optarg;
 145                         ispan = 1;      /* span across input files */
 146                         break;
 147                 case 'a':
 148                         aflag = 1;
 149                         break;
 150                 case 'e':
 151                         eflag = 1;
 152                         if ((temp_arg = malloc(strlen(optarg) + 2)) == NULL)
 153                                 err(1, "malloc");
 154                         strcpy(temp_arg, optarg);
 155                         strcat(temp_arg, "\n");
 156                         add_compunit(CU_STRING, temp_arg);
 157                         break;
 158                 case 'f':
 159                         fflag = 1;
 160                         add_compunit(CU_FILE, optarg);
 161                         break;
 162                 case 'i':
 163                         inplace = optarg;
 164                         ispan = 0;      /* don't span across input files */
 165                         break;
 166                 case 'l':
 167                         if(setvbuf(stdout, NULL, _IOLBF, 0) != 0)
 168                                 warnx("setting line buffered output failed");
 169                         break;
 170                 case 'n':
 171                         nflag = 1;
 172                         break;
 173                 case 'u':
 174                         if(setvbuf(stdout, NULL, _IONBF, 0) != 0)
 175                                 warnx("setting unbuffered output failed");
 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         exit(rval);
 203 }
 204
 205 static void
 206 usage(void)
 207 {
 208         (void)fprintf(stderr,
 209             "usage: %s script [-Ealnru] [-i extension] [file ...]\n"
 210             "\t%s [-Ealnu] [-i extension] [-e script] ... [-f script_file]"
 211             " ... [file ...]\n", getprogname(), getprogname());
 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                 default:
 255                         __builtin_unreachable();
 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;
 321         ssize_t len;
 322         char *dirbuf, *basebuf;
 323         static char *p = NULL;
 324         static size_t plen = 0;
 325         int c;
 326         static int firstfile;
 327
 328         if (infile == NULL) {
 329                 /* stdin? */
 330                 if (files->fname == NULL) {
 331                         if (inplace != NULL)
 332                                 errx(1, "-I or -i may not be used with stdin");
 333                         infile = stdin;
 334                         fname = "stdin";
 335                         outfile = stdout;
 336                         outfname = "stdout";
 337                 }
 338                 firstfile = 1;
 339         }
 340
 341         for (;;) {
 342                 if (infile != NULL && (c = getc(infile)) != EOF && !quit) {
 343                         (void)ungetc(c, infile);
 344                         break;
 345                 }
 346                 /* If we are here then either eof or no files are open yet */
 347                 if (infile == stdin) {
 348                         sp->len = 0;
 349                         return (0);
 350                 }
 351                 if (infile != NULL) {
 352                         fclose(infile);
 353                         if (*oldfname != '\0') {
 354                                 /* if there was a backup file, remove it */
 355                                 unlink(oldfname);
 356                                 /*
 357                                  * Backup the original.  Note that hard links
 358                                  * are not supported on all filesystems.
 359                                  */
 360                                 if ((link(fname, oldfname) != 0) &&
 361                                    (rename(fname, oldfname) != 0)) {
 362                                         warn("rename()");
 363                                         if (*tmpfname)
 364                                                 unlink(tmpfname);
 365                                         exit(1);
 366                                 }
 367                                 *oldfname = '\0';
 368                         }
 369                         if (*tmpfname != '\0') {
 370                                 if (outfile != NULL && outfile != stdout)
 371                                         if (fclose(outfile) != 0) {
 372                                                 warn("fclose()");
 373                                                 unlink(tmpfname);
 374                                                 exit(1);
 375                                         }
 376                                 outfile = NULL;
 377                                 if (rename(tmpfname, fname) != 0) {
 378                                         /* this should not happen really! */
 379                                         warn("rename()");
 380                                         unlink(tmpfname);
 381                                         exit(1);
 382                                 }
 383                                 *tmpfname = '\0';
 384                         }
 385                         outfname = NULL;
 386                 }
 387                 if (firstfile == 0)
 388                         files = files->next;
 389                 else
 390                         firstfile = 0;
 391                 if (files == NULL) {
 392                         sp->len = 0;
 393                         return (0);
 394                 }
 395                 fname = files->fname;
 396                 if (inplace != NULL) {
 397                         if (lstat(fname, &sb) != 0)
 398                                 err(1, "%s", fname);
 399                         if (!S_ISREG(sb.st_mode))
 400                                 errx(1, "%s: %s %s", fname,
 401                                     "in-place editing only",
 402                                     "works for regular files");
 403                         if (*inplace != '\0') {
 404                                 strlcpy(oldfname, fname,
 405                                     sizeof(oldfname));
 406                                 len = strlcat(oldfname, inplace,
 407                                     sizeof(oldfname));
 408                                 if (len > (ssize_t)sizeof(oldfname))
 409                                         errx(1, "%s: name too long", fname);
 410                         }
 411                         if ((dirbuf = strdup(fname)) == NULL ||
 412                             (basebuf = strdup(fname)) == NULL)
 413                                 err(1, "strdup");
 414                         len = snprintf(tmpfname, sizeof(tmpfname),
 415                             "%s/.!%ld!%s", dirname(dirbuf), (long)getpid(),
 416                             basename(basebuf));
 417                         free(dirbuf);
 418                         free(basebuf);
 419                         if (len >= (ssize_t)sizeof(tmpfname))
 420                                 errx(1, "%s: name too long", fname);
 421                         unlink(tmpfname);
 422                         if (outfile != NULL && outfile != stdout)
 423                                 fclose(outfile);
 424                         if ((outfile = fopen(tmpfname, "w")) == NULL)
 425                                 err(1, "%s", fname);
 426                         fchown(fileno(outfile), sb.st_uid, sb.st_gid);
 427                         fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
 428                         outfname = tmpfname;
 429                         if (!ispan) {
 430                                 linenum = 0;
 431                                 resetstate();
 432                         }
 433                 } else {
 434                         outfile = stdout;
 435                         outfname = "stdout";
 436                 }
 437                 if ((infile = fopen(fname, "r")) == NULL) {
 438                         warn("%s", fname);
 439                         rval = 1;
 440                         continue;
 441                 }
 442         }
 443         /*
 444          * We are here only when infile is open and we still have something
 445          * to read from it.
 446          *
 447          * Use getline() so that we can handle essentially infinite input
 448          * data.  The p and plen are static so each invocation gives
 449          * getline() the same buffer which is expanded as needed.
 450          */
 451         len = getline(&p, &plen, infile);
 452         if (len == -1)
 453                 err(1, "%s", fname);
 454         if (len != 0 && p[len - 1] == '\n') {
 455                 sp->append_newline = 1;
 456                 len--;
 457         } else if (!lastline()) {
 458                 sp->append_newline = 1;
 459         } else {
 460                 sp->append_newline = 0;
 461         }
 462         cspace(sp, p, len, spflag);
 463
 464         linenum++;
 465
 466         return (1);
 467 }
 468
 469 /*
 470  * Add a compilation unit to the linked list
 471  */
 472 static void
 473 add_compunit(enum e_cut type, char *s)
 474 {
 475         struct s_compunit *cu;
 476
 477         if ((cu = malloc(sizeof(struct s_compunit))) == NULL)
 478                 err(1, "malloc");
 479         cu->type = type;
 480         cu->s = s;
 481         cu->next = NULL;
 482         *cu_nextp = cu;
 483         cu_nextp = &cu->next;
 484 }
 485
 486 /*
 487  * Add a file to the linked list
 488  */
 489 static void
 490 add_file(char *s)
 491 {
 492         struct s_flist *fp;
 493
 494         if ((fp = malloc(sizeof(struct s_flist))) == NULL)
 495                 err(1, "malloc");
 496         fp->next = NULL;
 497         *fl_nextp = fp;
 498         fp->fname = s;
 499         fl_nextp = &fp->next;
 500 }
 501
 502 static int
 503 next_files_have_lines(void)
 504 {
 505         struct s_flist *file;
 506         FILE *file_fd;
 507         int ch;
 508
 509         file = files;
 510         while ((file = file->next) != NULL) {
 511                 if ((file_fd = fopen(file->fname, "r")) == NULL)
 512                         continue;
 513
 514                 if ((ch = getc(file_fd)) != EOF) {
 515                         /*
 516                          * This next file has content, therefore current
 517                          * file doesn't contains the last line.
 518                          */
 519                         ungetc(ch, file_fd);
 520                         fclose(file_fd);
 521                         return (1);
 522                 }
 523
 524                 fclose(file_fd);
 525         }
 526
 527         return (0);
 528 }
 529
 530 int
 531 lastline(void)
 532 {
 533         int ch;
 534
 535         if (feof(infile)) {
 536                 return !(
 537                     (inplace == NULL || ispan) &&
 538                     next_files_have_lines());
 539         }
 540         if ((ch = getc(infile)) == EOF) {
 541                 return !(
 542                     (inplace == NULL || ispan) &&
 543                     next_files_have_lines());
 544         }
 545         ungetc(ch, infile);
 546         return (0);
 547 }