package/toolbox/src/sed/main.c

   1 /*      $OpenBSD: main.c,v 1.35 2017/08/01 18:05:53 martijn Exp $       */
   2
   3 /*-
   4  * Copyright (c) 2016, 2017
   5  *      mirabilos <m@mirbsd.org>
   6  * Copyright (c) 1992 Diomidis Spinellis.
   7  * Copyright (c) 1992, 1993
   8  *      The Regents of the University of California.  All rights reserved.
   9  *
  10  * This code is derived from software contributed to Berkeley by
  11  * Diomidis Spinellis of Imperial College, University of London.
  12  *
  13  * Redistribution and use in source and binary forms, with or without
  14  * modification, are permitted provided that the following conditions
  15  * are met:
  16  * 1. Redistributions of source code must retain the above copyright
  17  *    notice, this list of conditions and the following disclaimer.
  18  * 2. Redistributions in binary form must reproduce the above copyright
  19  *    notice, this list of conditions and the following disclaimer in the
  20  *    documentation and/or other materials provided with the distribution.
  21  * 3. Neither the name of the University nor the names of its contributors
  22  *    may be used to endorse or promote products derived from this software
  23  *    without specific prior written permission.
  24  *
  25  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  26  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  27  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  28  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  29  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  30  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  31  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  32  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  33  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  34  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  35  * SUCH DAMAGE.
  36  */
  37
  38 #include <sys/types.h>
  39 #include <sys/ioctl.h>
  40 #include <sys/stat.h>
  41
  42 #include <ctype.h>
  43 #include <errno.h>
  44 #include <fcntl.h>
  45 #include <limits.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 #include <libgen.h>
  53
  54 #include "defs.h"
  55 #include "extern.h"
  56
  57 __RCSID("$MirOS: src/usr.bin/sed/main.c,v 1.3 2017/11/20 01:23:57 tg Exp $");
  58
  59 /*
  60  * Linked list of units (strings and files) to be compiled
  61  */
  62 struct s_compunit {
  63         struct s_compunit *next;
  64         enum e_cut {CU_FILE, CU_STRING} type;
  65         char *s;                        /* Pointer to string or fname */
  66 };
  67
  68 /*
  69  * Linked list pointer to compilation units and pointer to current
  70  * next pointer.
  71  */
  72 static struct s_compunit *script, **cu_nextp = &script;
  73
  74 /*
  75  * Linked list of files to be processed
  76  */
  77 struct s_flist {
  78         char *fname;
  79         struct s_flist *next;
  80 };
  81
  82 /*
  83  * Linked list pointer to files and pointer to current
  84  * next pointer.
  85  */
  86 static struct s_flist *files, **fl_nextp = &files;
  87
  88 FILE *infile;                   /* Current input file */
  89 FILE *outfile;                  /* Current output file */
  90
  91 int Eflag, aflag, eflag, nflag;
  92 static int rval;        /* Exit status */
  93
  94 /*
  95  * Current file and line number; line numbers restart across compilation
  96  * units, but span across input files.  The latter is optional if editing
  97  * in place.
  98  */
  99 const char *fname;              /* File name. */
 100 const char *outfname;           /* Output file name */
 101 static char oldfname[PATH_MAX]; /* Old file name (for in-place editing) */
 102 static char tmpfname[PATH_MAX]; /* Temporary file name (for in-place editing) */
 103 const char *inplace;            /* Inplace edit file extension */
 104 u_long linenum;
 105
 106 static void add_compunit(enum e_cut, char *);
 107 static void add_file(char *);
 108 static int next_files_have_lines(void);
 109
 110 int termwidth;
 111
 112 int pledge_wpath, pledge_rpath;
 113
 114 int
 115 main(int argc, char *argv[])
 116 {
 117         int c, fflag;
 118
 119         fflag = 0;
 120         inplace = NULL;
 121         while ((c = getopt(argc, argv, "Eae:f:i::nru")) != -1)
 122                 switch (c) {
 123                 case 'E':
 124                 case 'r':
 125                         Eflag = 1;
 126                         break;
 127                 case 'a':
 128                         aflag = 1;
 129                         break;
 130                 case 'e':
 131                         eflag = 1;
 132                         add_compunit(CU_STRING, optarg);
 133                         break;
 134                 case 'f':
 135                         fflag = 1;
 136                         add_compunit(CU_FILE, optarg);
 137                         break;
 138                 case 'i':
 139                         inplace = optarg ? optarg : "";
 140                         break;
 141                 case 'n':
 142                         nflag = 1;
 143                         break;
 144                 case 'u':
 145                         setvbuf(stdout, NULL, _IOLBF, 0);
 146                         break;
 147                 default:
 148                 case '?':
 149                         (void)fprintf(stderr,
 150                             "usage: sed [-aEnru] [-i[extension]] command [file ...]\n"
 151                             "       sed [-aEnru] [-e command] [-f command_file] [-i[extension]] [file ...]\n");
 152                         exit(1);
 153                 }
 154         argc -= optind;
 155         argv += optind;
 156
 157         termwidth = 60;
 158
 159 #if defined(__OpenBSD__) && !defined(__MirBSD__)
 160         if (inplace != NULL) {
 161                 if (pledge("stdio rpath wpath cpath fattr chown", NULL) == -1)
 162                         error(FATAL, "pledge: %s", strerror(errno));
 163         } else {
 164                 if (pledge("stdio rpath wpath cpath", NULL) == -1)
 165                         error(FATAL, "pledge: %s", strerror(errno));
 166         }
 167 #endif
 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 #if defined(__OpenBSD__) && !defined(__MirBSD__)
 180                 if (!pledge_wpath && inplace == NULL) {
 181                         if (pledge("stdio rpath", NULL) == -1)
 182                                 error(FATAL, "pledge: %s", strerror(errno));
 183                 }
 184 #endif
 185                 for (; *argv; argv++)
 186                         add_file(*argv);
 187         } else {
 188 #if defined(__OpenBSD__) && !defined(__MirBSD__)
 189                 if (!pledge_wpath && !pledge_rpath) {
 190                         if (pledge("stdio", NULL) == -1)
 191                                 error(FATAL, "pledge: %s", strerror(errno));
 192                 } else if (pledge_rpath) {
 193                         if (pledge("stdio rpath", NULL) == -1)
 194                                 error(FATAL, "pledge: %s", strerror(errno));
 195                 } else if (pledge_wpath) {
 196                         if (pledge("stdio wpath cpath", NULL) == -1)
 197                                 error(FATAL, "pledge: %s", strerror(errno));
 198                 }
 199 #endif
 200                 add_file(NULL);
 201         }
 202         process();
 203         cfclose(prog, NULL);
 204         if (fclose(stdout))
 205                 error(FATAL, "stdout: %s", strerror(errno));
 206         exit (rval);
 207 }
 208
 209 /*
 210  * Like fgets, but go through the chain of compilation units chaining them
 211  * together.  Empty strings and files are ignored.
 212  */
 213 char *
 214 cu_fgets(char **outbuf, size_t *outsize)
 215 {
 216         static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
 217         static FILE *f;         /* Current open file */
 218         static char *s;         /* Current pointer inside string */
 219         static char string_ident[30];
 220         size_t len;
 221         char *p;
 222
 223         if (*outbuf == NULL)
 224                 *outsize = 0;
 225
 226 again:
 227         switch (state) {
 228         case ST_EOF:
 229                 if (script == NULL)
 230                         goto cu_fgets_nilreturn;
 231                 linenum = 0;
 232                 switch (script->type) {
 233                 case CU_FILE:
 234                         if ((f = fopen(script->s, "r")) == NULL)
 235                                 error(FATAL,
 236                                     "%s: %s", script->s, strerror(errno));
 237                         fname = script->s;
 238                         state = ST_FILE;
 239                         goto again;
 240                 case CU_STRING:
 241                         if (((size_t)snprintf(string_ident,
 242                             sizeof(string_ident), "\"%s\"", script->s)) >=
 243                             sizeof(string_ident))
 244                                 strlcpy(string_ident +
 245                                     sizeof(string_ident) - 6, " ...\"", 5);
 246                         fname = string_ident;
 247                         s = script->s;
 248                         state = ST_STRING;
 249                         goto again;
 250                 }
 251         case ST_FILE:
 252                 if ((p = fgetln(f, &len)) != NULL) {
 253                         linenum++;
 254                         if (len >= *outsize) {
 255                                 free(*outbuf);
 256                                 *outsize = ROUNDLEN(len + 1);
 257                                 *outbuf = xmalloc(*outsize);
 258                         }
 259                         memcpy(*outbuf, p, len);
 260                         (*outbuf)[len] = '\0';
 261                         if (linenum == 1 && p[0] == '#' && p[1] == 'n')
 262                                 nflag = 1;
 263                         return (*outbuf);
 264                 }
 265                 script = script->next;
 266                 (void)fclose(f);
 267                 state = ST_EOF;
 268                 goto again;
 269         case ST_STRING:
 270                 if (linenum == 0 && s[0] == '#' && s[1] == 'n')
 271                         nflag = 1;
 272                 p = *outbuf;
 273                 len = *outsize;
 274                 for (;;) {
 275                         if (len <= 1) {
 276                                 *outbuf = xrealloc(*outbuf,
 277                                     *outsize + _POSIX2_LINE_MAX);
 278                                 p = *outbuf + *outsize - len;
 279                                 len += _POSIX2_LINE_MAX;
 280                                 *outsize += _POSIX2_LINE_MAX;
 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                                         return (*outbuf);
 293                                 }
 294                         case '\n':
 295                                 *p++ = '\n';
 296                                 *p = '\0';
 297                                 s++;
 298                                 linenum++;
 299                                 return (*outbuf);
 300                         default:
 301                                 *p++ = *s++;
 302                                 len--;
 303                         }
 304                 }
 305         }
 306         /* NOTREACHED */
 307         /* but GCC doesn't care, so: */
 308  cu_fgets_nilreturn:
 309         return (NULL);
 310 }
 311
 312 /*
 313  * Like fgets, but go through the list of files chaining them together.
 314  * Set len to the length of the line.
 315  */
 316 int
 317 mf_fgets(SPACE *sp, enum e_spflag spflag)
 318 {
 319         struct stat sb;
 320         size_t len;
 321         char *p;
 322         int c, fd;
 323         static int firstfile;
 324
 325         if (infile == NULL) {
 326                 /* stdin? */
 327                 if (files->fname == NULL) {
 328                         if (inplace != NULL)
 329                                 error(FATAL, "-i may not be used with stdin");
 330                         infile = stdin;
 331                         fname = "stdin";
 332                         outfile = stdout;
 333                         outfname = "stdout";
 334                 }
 335
 336                 firstfile = 1;
 337         }
 338
 339         for (;;) {
 340                 if (infile != NULL && (c = getc(infile)) != EOF) {
 341                         (void)ungetc(c, infile);
 342                         break;
 343                 }
 344                 /* If we are here then either eof or no files are open yet */
 345                 if (infile == stdin) {
 346                         sp->len = 0;
 347                         return (0);
 348                 }
 349                 if (infile != NULL) {
 350                         fclose(infile);
 351                         if (*oldfname != '\0') {
 352                                 if (rename(fname, oldfname) != 0) {
 353                                         warning("rename()");
 354                                         unlink(tmpfname);
 355                                         exit(1);
 356                                 }
 357                                 *oldfname = '\0';
 358                         }
 359                         if (*tmpfname != '\0') {
 360                                 if (outfile != NULL && outfile != stdout)
 361                                         fclose(outfile);
 362                                 outfile = NULL;
 363                                 rename(tmpfname, fname);
 364                                 *tmpfname = '\0';
 365                         }
 366                         outfname = NULL;
 367                 }
 368                 if (firstfile == 0)
 369                         files = files->next;
 370                 else
 371                         firstfile = 0;
 372                 if (files == NULL) {
 373                         sp->len = 0;
 374                         return (0);
 375                 }
 376                 fname = files->fname;
 377                 if (inplace != NULL) {
 378                         char *tmpdirname;
 379
 380                         if (lstat(fname, &sb) != 0)
 381                                 error(FATAL, "%s: %s", fname,
 382                                     strerror(errno ? errno : EIO));
 383                         if (!S_ISREG(sb.st_mode))
 384                                 error(FATAL, "%s: %s %s", fname,
 385                                     "in-place editing only",
 386                                     "works for regular files");
 387                         if (*inplace != '\0') {
 388                                 strlcpy(oldfname, fname,
 389                                     sizeof(oldfname));
 390                                 len = strlcat(oldfname, inplace,
 391                                     sizeof(oldfname));
 392                                 if (len > sizeof(oldfname))
 393                                         error(FATAL, "%s: name too long", fname);
 394                         }
 395                         len = snprintf(tmpfname, sizeof(tmpfname), "%s/sedXXXXXXXXXX",
 396                             dirname(tmpdirname = strdup(fname)));
 397                         free(tmpdirname);
 398                         if (len >= sizeof(tmpfname))
 399                                 error(FATAL, "%s: name too long", fname);
 400                         if ((fd = mkstemp(tmpfname)) == -1)
 401                                 error(FATAL, "%s: %s", fname, strerror(errno));
 402                         if ((outfile = fdopen(fd, "w")) == NULL) {
 403                                 unlink(tmpfname);
 404                                 error(FATAL, "%s", fname);
 405                         }
 406                         fchown(fileno(outfile), sb.st_uid, sb.st_gid);
 407                         fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
 408                         outfname = tmpfname;
 409                         linenum = 0;
 410                         resetranges();
 411                 } else {
 412                         outfile = stdout;
 413                         outfname = "stdout";
 414                 }
 415                 if ((infile = fopen(fname, "r")) == NULL) {
 416                         warning("%s", strerror(errno));
 417                         rval = 1;
 418                         continue;
 419                 }
 420         }
 421
 422         /*
 423          * We are here only when infile is open and we still have something
 424          * to read from it.
 425          *
 426          * Use fgetln so that we can handle essentially infinite input data.
 427          * Can't use the pointer into the stdio buffer as the process space
 428          * because the ungetc() can cause it to move.
 429          */
 430         p = fgetln(infile, &len);
 431         if (ferror(infile))
 432                 error(FATAL, "%s: %s", fname, strerror(errno ? errno : EIO));
 433         if (len != 0 && p[len - 1] == '\n') {
 434                 sp->append_newline = 1;
 435                 len--;
 436         } else if (!lastline()) {
 437                 sp->append_newline = 1;
 438         } else {
 439                 sp->append_newline = 0;
 440         }
 441         cspace(sp, p, len, spflag);
 442
 443         linenum++;
 444
 445         return (1);
 446 }
 447
 448 /*
 449  * Add a compilation unit to the linked list
 450  */
 451 static void
 452 add_compunit(enum e_cut type, char *s)
 453 {
 454         struct s_compunit *cu;
 455
 456         cu = xmalloc(sizeof(struct s_compunit));
 457         cu->type = type;
 458         cu->s = s;
 459         cu->next = NULL;
 460         *cu_nextp = cu;
 461         cu_nextp = &cu->next;
 462 }
 463
 464 /*
 465  * Add a file to the linked list
 466  */
 467 static void
 468 add_file(char *s)
 469 {
 470         struct s_flist *fp;
 471
 472         fp = xmalloc(sizeof(struct s_flist));
 473         fp->next = NULL;
 474         *fl_nextp = fp;
 475         fp->fname = s;
 476         fl_nextp = &fp->next;
 477 }
 478
 479
 480 static int
 481 next_files_have_lines(void)
 482 {
 483         struct s_flist *file;
 484         FILE *file_fd;
 485         int ch;
 486
 487         file = files;
 488         while ((file = file->next) != NULL) {
 489                 if ((file_fd = fopen(file->fname, "r")) == NULL)
 490                         continue;
 491
 492                 if ((ch = getc(file_fd)) != EOF) {
 493                         /*
 494                          * This next file has content, therefore current
 495                          * file doesn't contains the last line.
 496                          */
 497                         ungetc(ch, file_fd);
 498                         fclose(file_fd);
 499                         return (1);
 500                 }
 501                 fclose(file_fd);
 502         }
 503         return (0);
 504 }
 505
 506 int
 507 lastline(void)
 508 {
 509         int ch;
 510
 511         if (feof(infile)) {
 512                 return !(
 513                     (inplace == NULL) &&
 514                     next_files_have_lines());
 515         }
 516         if ((ch = getc(infile)) == EOF) {
 517                 return !(
 518                     (inplace == NULL) &&
 519                     next_files_have_lines());
 520         }
 521         ungetc(ch, infile);
 522         return (0);
 523 }