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