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  * 3. 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  * @(#) Copyright (c) 1992, 1993 The Regents of the University of California.  All rights reserved.
  35  * @(#)main.c   8.2 (Berkeley) 1/3/94
  36  * $FreeBSD: head/usr.bin/sed/main.c 303662 2016-08-02 15:35:53Z pfg $
  37  */
  38
  39 #include <sys/types.h>
  40 #include <sys/mman.h>
  41 #include <sys/param.h>
  42 #include <sys/stat.h>
  43
  44 #include <err.h>
  45 #include <errno.h>
  46 #include <fcntl.h>
  47 #include <libgen.h>
  48 #include <limits.h>
  49 #include <locale.h>
  50 #include <stddef.h>
  51 #include <stdio.h>
  52 #include <stdlib.h>
  53 #include <string.h>
  54 #include <unistd.h>
  55 #include <regex.h>
  56
  57 #include "defs.h"
  58 #include "extern.h"
  59
  60 /*
  61  * Linked list of units (strings and files) to be compiled
  62  */
  63 struct s_compunit {
  64         struct s_compunit *next;
  65         enum e_cut {CU_FILE, CU_STRING} type;
  66         char *s;                        /* Pointer to string or fname */
  67 };
  68
  69 /*
  70  * Linked list pointer to compilation units and pointer to current
  71  * next pointer.
  72  */
  73 static struct s_compunit *script, **cu_nextp = &script;
  74
  75 /*
  76  * Linked list of files to be processed
  77  */
  78 struct s_flist {
  79         char *fname;
  80         struct s_flist *next;
  81 };
  82
  83 /*
  84  * Linked list pointer to files and pointer to current
  85  * next pointer.
  86  */
  87 static struct s_flist *files, **fl_nextp = &files;
  88
  89 FILE *infile;                   /* Current input file */
  90 FILE *outfile;                  /* Current output file */
  91
  92 int aflag, eflag, nflag;
  93 int rflags = 0;
  94 static int rval;                /* Exit status */
  95
  96 static int ispan;               /* Whether inplace editing spans across files */
  97
  98 /*
  99  * Current file and line number; line numbers restart across compilation
 100  * units, but span across input files.  The latter is optional if editing
 101  * in place.
 102  */
 103 const char *fname;              /* File name. */
 104 const char *outfname;           /* Output file name */
 105 static char oldfname[PATH_MAX]; /* Old file name (for in-place editing) */
 106 static char tmpfname[PATH_MAX]; /* Temporary file name (for in-place editing) */
 107 static const char *inplace;     /* Inplace edit file extension. */
 108 u_long linenum;
 109
 110 static void add_compunit(enum e_cut, char *);
 111 static void add_file(char *);
 112 static void usage(void);
 113
 114 int
 115 main(int argc, char *argv[])
 116 {
 117         int c, fflag;
 118         char *temp_arg;
 119
 120         (void) setlocale(LC_ALL, "");
 121
 122         fflag = 0;
 123         inplace = NULL;
 124
 125         while ((c = getopt(argc, argv, "EI:ae:f:i:lnru")) != -1)
 126                 switch (c) {
 127                 case 'r':               /* GNU sed compat */
 128                 case 'E':
 129                         rflags = REG_EXTENDED;
 130                         break;
 131                 case 'I':
 132                         inplace = optarg;
 133                         ispan = 1;      /* span across input files */
 134                         break;
 135                 case 'a':
 136                         aflag = 1;
 137                         break;
 138                 case 'e':
 139                         eflag = 1;
 140                         if ((temp_arg = malloc(strlen(optarg) + 2)) == NULL)
 141                                 err(1, "malloc");
 142                         strcpy(temp_arg, optarg);
 143                         strcat(temp_arg, "\n");
 144                         add_compunit(CU_STRING, temp_arg);
 145                         break;
 146                 case 'f':
 147                         fflag = 1;
 148                         add_compunit(CU_FILE, optarg);
 149                         break;
 150                 case 'i':
 151                         inplace = optarg;
 152                         ispan = 0;      /* don't span across input files */
 153                         break;
 154                 case 'l':
 155                         if(setvbuf(stdout, NULL, _IOLBF, 0) != 0)
 156                                 warnx("setting line buffered output failed");
 157                         break;
 158                 case 'n':
 159                         nflag = 1;
 160                         break;
 161                 case 'u':
 162                         if(setvbuf(stdout, NULL, _IONBF, 0) != 0)
 163                                 warnx("setting unbuffered output failed");
 164                         break;
 165                 default:
 166                 case '?':
 167                         usage();
 168                 }
 169         argc -= optind;
 170         argv += optind;
 171
 172         /* First usage case; script is the first arg */
 173         if (!eflag && !fflag && *argv) {
 174                 add_compunit(CU_STRING, *argv);
 175                 argv++;
 176         }
 177
 178         compile();
 179
 180         /* Continue with first and start second usage */
 181         if (*argv)
 182                 for (; *argv; argv++)
 183                         add_file(*argv);
 184         else
 185                 add_file(NULL);
 186         process();
 187         cfclose(prog, NULL);
 188         if (fclose(stdout))
 189                 err(1, "stdout");
 190         exit(rval);
 191 }
 192
 193 static void
 194 usage(void)
 195 {
 196         (void)fprintf(stderr,
 197             "usage: %s script [-Ealnru] [-i extension] [file ...]\n"
 198             "\t%s [-Ealnu] [-i extension] [-e script] ... [-f script_file]"
 199             " ... [file ...]\n", getprogname(), getprogname());
 200         exit(1);
 201 }
 202
 203 /*
 204  * Like fgets, but go through the chain of compilation units chaining them
 205  * together.  Empty strings and files are ignored.
 206  */
 207 char *
 208 cu_fgets(char *buf, int n, int *more)
 209 {
 210         static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
 211         static FILE *f;         /* Current open file */
 212         static char *s;         /* Current pointer inside string */
 213         static char string_ident[30];
 214         char *p;
 215
 216 again:
 217         switch (state) {
 218         case ST_EOF:
 219                 if (script == NULL) {
 220                         if (more != NULL)
 221                                 *more = 0;
 222                         return (NULL);
 223                 }
 224                 linenum = 0;
 225                 switch (script->type) {
 226                 case CU_FILE:
 227                         if ((f = fopen(script->s, "r")) == NULL)
 228                                 err(1, "%s", script->s);
 229                         fname = script->s;
 230                         state = ST_FILE;
 231                         goto again;
 232                 case CU_STRING:
 233                         if (((size_t)snprintf(string_ident,
 234                             sizeof(string_ident), "\"%s\"", script->s)) >=
 235                             sizeof(string_ident) - 1)
 236                                 (void)strcpy(string_ident +
 237                                     sizeof(string_ident) - 6, " ...\"");
 238                         fname = string_ident;
 239                         s = script->s;
 240                         state = ST_STRING;
 241                         goto again;
 242                 default:
 243                         errx(1, "illegal script type: %d", script->type);
 244                 }
 245         case ST_FILE:
 246                 if ((p = fgets(buf, n, f)) != NULL) {
 247                         linenum++;
 248                         if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
 249                                 nflag = 1;
 250                         if (more != NULL)
 251                                 *more = !feof(f);
 252                         return (p);
 253                 }
 254                 script = script->next;
 255                 (void)fclose(f);
 256                 state = ST_EOF;
 257                 goto again;
 258         case ST_STRING:
 259                 if (linenum == 0 && s[0] == '#' && s[1] == 'n')
 260                         nflag = 1;
 261                 p = buf;
 262                 for (;;) {
 263                         if (n-- <= 1) {
 264                                 *p = '\0';
 265                                 linenum++;
 266                                 if (more != NULL)
 267                                         *more = 1;
 268                                 return (buf);
 269                         }
 270                         switch (*s) {
 271                         case '\0':
 272                                 state = ST_EOF;
 273                                 if (s == script->s) {
 274                                         script = script->next;
 275                                         goto again;
 276                                 } else {
 277                                         script = script->next;
 278                                         *p = '\0';
 279                                         linenum++;
 280                                         if (more != NULL)
 281                                                 *more = 0;
 282                                         return (buf);
 283                                 }
 284                         case '\n':
 285                                 *p++ = '\n';
 286                                 *p = '\0';
 287                                 s++;
 288                                 linenum++;
 289                                 if (more != NULL)
 290                                         *more = 0;
 291                                 return (buf);
 292                         default:
 293                                 *p++ = *s++;
 294                         }
 295                 }
 296         }
 297         /* NOTREACHED */
 298         return (NULL);
 299 }
 300
 301 /*
 302  * Like fgets, but go through the list of files chaining them together.
 303  * Set len to the length of the line.
 304  */
 305 int
 306 mf_fgets(SPACE *sp, enum e_spflag spflag)
 307 {
 308         struct stat sb;
 309         ssize_t len;
 310         char *dirbuf, *basebuf;
 311         static char *p = NULL;
 312         static size_t plen = 0;
 313         int c;
 314         static int firstfile;
 315
 316         if (infile == NULL) {
 317                 /* stdin? */
 318                 if (files->fname == NULL) {
 319                         if (inplace != NULL)
 320                                 errx(1, "-I or -i may not be used with stdin");
 321                         infile = stdin;
 322                         fname = "stdin";
 323                         outfile = stdout;
 324                         outfname = "stdout";
 325                 }
 326                 firstfile = 1;
 327         }
 328
 329         for (;;) {
 330                 if (infile != NULL && (c = getc(infile)) != EOF) {
 331                         (void)ungetc(c, infile);
 332                         break;
 333                 }
 334                 /* If we are here then either eof or no files are open yet */
 335                 if (infile == stdin) {
 336                         sp->len = 0;
 337                         return (0);
 338                 }
 339                 if (infile != NULL) {
 340                         fclose(infile);
 341                         if (*oldfname != '\0') {
 342                                 /* if there was a backup file, remove it */
 343                                 unlink(oldfname);
 344                                 /*
 345                                  * Backup the original.  Note that hard links
 346                                  * are not supported on all filesystems.
 347                                  */
 348                                 if ((link(fname, oldfname) != 0) &&
 349                                    (rename(fname, oldfname) != 0)) {
 350                                         warn("rename()");
 351                                         if (*tmpfname)
 352                                                 unlink(tmpfname);
 353                                         exit(1);
 354                                 }
 355                                 *oldfname = '\0';
 356                         }
 357                         if (*tmpfname != '\0') {
 358                                 if (outfile != NULL && outfile != stdout)
 359                                         if (fclose(outfile) != 0) {
 360                                                 warn("fclose()");
 361                                                 unlink(tmpfname);
 362                                                 exit(1);
 363                                         }
 364                                 outfile = NULL;
 365                                 if (rename(tmpfname, fname) != 0) {
 366                                         /* this should not happen really! */
 367                                         warn("rename()");
 368                                         unlink(tmpfname);
 369                                         exit(1);
 370                                 }
 371                                 *tmpfname = '\0';
 372                         }
 373                         outfname = NULL;
 374                 }
 375                 if (firstfile == 0)
 376                         files = files->next;
 377                 else
 378                         firstfile = 0;
 379                 if (files == NULL) {
 380                         sp->len = 0;
 381                         return (0);
 382                 }
 383                 fname = files->fname;
 384                 if (inplace != NULL) {
 385                         if (lstat(fname, &sb) != 0)
 386                                 err(1, "%s", fname);
 387                         if (!(sb.st_mode & S_IFREG))
 388                                 errx(1, "%s: %s %s", fname,
 389                                     "in-place editing only",
 390                                     "works for regular files");
 391                         if (*inplace != '\0') {
 392                                 strlcpy(oldfname, fname,
 393                                     sizeof(oldfname));
 394                                 len = strlcat(oldfname, inplace,
 395                                     sizeof(oldfname));
 396                                 if (len > (ssize_t)sizeof(oldfname))
 397                                         errx(1, "%s: name too long", fname);
 398                         }
 399                         if ((dirbuf = strdup(fname)) == NULL ||
 400                             (basebuf = strdup(fname)) == NULL)
 401                                 err(1, "strdup");
 402                         len = snprintf(tmpfname, sizeof(tmpfname),
 403                             "%s/.!%ld!%s", dirname(dirbuf), (long)getpid(),
 404                             basename(basebuf));
 405                         free(dirbuf);
 406                         free(basebuf);
 407                         if (len >= (ssize_t)sizeof(tmpfname))
 408                                 errx(1, "%s: name too long", fname);
 409                         unlink(tmpfname);
 410                         if (outfile != NULL && outfile != stdout)
 411                                 fclose(outfile);
 412                         if ((outfile = fopen(tmpfname, "w")) == NULL)
 413                                 err(1, "%s", fname);
 414                         fchown(fileno(outfile), sb.st_uid, sb.st_gid);
 415                         fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
 416                         outfname = tmpfname;
 417                         if (!ispan) {
 418                                 linenum = 0;
 419                                 resetstate();
 420                         }
 421                 } else {
 422                         outfile = stdout;
 423                         outfname = "stdout";
 424                 }
 425                 if ((infile = fopen(fname, "r")) == NULL) {
 426                         warn("%s", fname);
 427                         rval = 1;
 428                         continue;
 429                 }
 430         }
 431         /*
 432          * We are here only when infile is open and we still have something
 433          * to read from it.
 434          *
 435          * Use getline() so that we can handle essentially infinite input
 436          * data.  The p and plen are static so each invocation gives
 437          * getline() the same buffer which is expanded as needed.
 438          */
 439         len = getline(&p, &plen, infile);
 440         if (len == -1)
 441                 err(1, "%s", fname);
 442         if (len != 0 && p[len - 1] == '\n') {
 443                 sp->append_newline = 1;
 444                 len--;
 445         } else if (!lastline()) {
 446                 sp->append_newline = 1;
 447         } else {
 448                 sp->append_newline = 0;
 449         }
 450         cspace(sp, p, len, spflag);
 451
 452         linenum++;
 453
 454         return (1);
 455 }
 456
 457 /*
 458  * Add a compilation unit to the linked list
 459  */
 460 static void
 461 add_compunit(enum e_cut type, char *s)
 462 {
 463         struct s_compunit *cu;
 464
 465         if ((cu = malloc(sizeof(struct s_compunit))) == NULL)
 466                 err(1, "malloc");
 467         cu->type = type;
 468         cu->s = s;
 469         cu->next = NULL;
 470         *cu_nextp = cu;
 471         cu_nextp = &cu->next;
 472 }
 473
 474 /*
 475  * Add a file to the linked list
 476  */
 477 static void
 478 add_file(char *s)
 479 {
 480         struct s_flist *fp;
 481
 482         if ((fp = malloc(sizeof(struct s_flist))) == NULL)
 483                 err(1, "malloc");
 484         fp->next = NULL;
 485         *fl_nextp = fp;
 486         fp->fname = s;
 487         fl_nextp = &fp->next;
 488 }
 489
 490 static int
 491 next_files_have_lines(void)
 492 {
 493         struct s_flist *file;
 494         FILE *file_fd;
 495         int ch;
 496
 497         file = files;
 498         while ((file = file->next) != NULL) {
 499                 if ((file_fd = fopen(file->fname, "r")) == NULL)
 500                         continue;
 501
 502                 if ((ch = getc(file_fd)) != EOF) {
 503                         /*
 504                          * This next file has content, therefore current
 505                          * file doesn't contains the last line.
 506                          */
 507                         ungetc(ch, file_fd);
 508                         fclose(file_fd);
 509                         return (1);
 510                 }
 511
 512                 fclose(file_fd);
 513         }
 514
 515         return (0);
 516 }
 517
 518 int
 519 lastline(void)
 520 {
 521         int ch;
 522
 523         if (feof(infile)) {
 524                 return !(
 525                     (inplace == NULL || ispan) &&
 526                     next_files_have_lines());
 527         }
 528         if ((ch = getc(infile)) == EOF) {
 529                 return !(
 530                     (inplace == NULL || ispan) &&
 531                     next_files_have_lines());
 532         }
 533         ungetc(ch, infile);
 534         return (0);
 535 }