usr.bin/sed/process.c

   1 /*-
   2  * Copyright (c) 1992 Diomidis Spinellis.
   3  * Copyright (c) 1992, 1993, 1994
   4  *      The Regents of the University of California.  All rights reserved.
   5  *
   6  * This code is derived from software contributed to Berkeley by
   7  * Diomidis Spinellis of Imperial College, University of London.
   8  *
   9  * Redistribution and use in source and binary forms, with or without
  10  * modification, are permitted provided that the following conditions
  11  * are met:
  12  * 1. Redistributions of source code must retain the above copyright
  13  *    notice, this list of conditions and the following disclaimer.
  14  * 2. Redistributions in binary form must reproduce the above copyright
  15  *    notice, this list of conditions and the following disclaimer in the
  16  *    documentation and/or other materials provided with the distribution.
  17  * 3. All advertising materials mentioning features or use of this software
  18  *    must display the following acknowledgement:
  19  *      This product includes software developed by the University of
  20  *      California, Berkeley and its contributors.
  21  * 4. 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  * @(#)process.c        8.6 (Berkeley) 4/20/94
  38  * $FreeBSD: src/usr.bin/sed/process.c,v 1.10.2.10 2003/06/13 07:32:08 fanf Exp $
  39  * $DragonFly: src/usr.bin/sed/process.c,v 1.5 2003/11/04 15:48:16 drhodus Exp $
  40  */
  41
  42 #include <sys/types.h>
  43 #include <sys/stat.h>
  44 #include <sys/ioctl.h>
  45 #include <sys/uio.h>
  46
  47 #include <ctype.h>
  48 #include <err.h>
  49 #include <errno.h>
  50 #include <fcntl.h>
  51 #include <limits.h>
  52 #include <regex.h>
  53 #include <stdio.h>
  54 #include <stdlib.h>
  55 #include <string.h>
  56 #include <unistd.h>
  57
  58 #include "defs.h"
  59 #include "extern.h"
  60
  61 static SPACE HS, PS, SS;
  62 #define pd              PS.deleted
  63 #define ps              PS.space
  64 #define psl             PS.len
  65 #define hs              HS.space
  66 #define hsl             HS.len
  67
  68 static __inline int      applies(struct s_command *);
  69 static void              flush_appends(void);
  70 static void              lputs(char *);
  71 static __inline int      regexec_e(regex_t *, const char *, int, int, size_t);
  72 static void              regsub(SPACE *, char *, char *);
  73 static int               substitute(struct s_command *);
  74
  75 struct s_appends *appends;      /* Array of pointers to strings to append. */
  76 static int appendx;             /* Index into appends array. */
  77 int appendnum;                  /* Size of appends array. */
  78
  79 static int lastaddr;            /* Set by applies if last address of a range. */
  80 static int sdone;               /* If any substitutes since last line input. */
  81                                 /* Iov structure for 'w' commands. */
  82 static regex_t *defpreg;
  83 size_t maxnsub;
  84 regmatch_t *match;
  85
  86 #define OUT(s) { fwrite(s, sizeof(u_char), psl, stdout); putchar('\n'); }
  87
  88 void
  89 process(void)
  90 {
  91         struct s_command *cp;
  92         SPACE tspace;
  93         size_t len, oldpsl = 0;
  94         char *p;
  95
  96         p = NULL;
  97
  98         for (linenum = 0; mf_fgets(&PS, REPLACE);) {
  99                 pd = 0;
 100 top:
 101                 cp = prog;
 102 redirect:
 103                 while (cp != NULL) {
 104                         if (!applies(cp)) {
 105                                 cp = cp->next;
 106                                 continue;
 107                         }
 108                         switch (cp->code) {
 109                         case '{':
 110                                 cp = cp->u.c;
 111                                 goto redirect;
 112                         case 'a':
 113                                 if (appendx >= appendnum)
 114                                         if ((appends = realloc(appends,
 115                                             sizeof(struct s_appends) *
 116                                             (appendnum *= 2))) == NULL)
 117                                                 err(1, "realloc");
 118                                 appends[appendx].type = AP_STRING;
 119                                 appends[appendx].s = cp->t;
 120                                 appends[appendx].len = strlen(cp->t);
 121                                 appendx++;
 122                                 break;
 123                         case 'b':
 124                                 cp = cp->u.c;
 125                                 goto redirect;
 126                         case 'c':
 127                                 pd = 1;
 128                                 psl = 0;
 129                                 if (cp->a2 == NULL || lastaddr)
 130                                         (void)printf("%s", cp->t);
 131                                 break;
 132                         case 'd':
 133                                 pd = 1;
 134                                 goto new;
 135                         case 'D':
 136                                 if (pd)
 137                                         goto new;
 138                                 if (psl == 0 ||
 139                                     (p = memchr(ps, '\n', psl)) == NULL) {
 140                                         pd = 1;
 141                                         goto new;
 142                                 } else {
 143                                         psl -= (p + 1) - ps;
 144                                         memmove(ps, p + 1, psl);
 145                                         goto top;
 146                                 }
 147                         case 'g':
 148                                 cspace(&PS, hs, hsl, REPLACE);
 149                                 break;
 150                         case 'G':
 151                                 cspace(&PS, "\n", 1, 0);
 152                                 cspace(&PS, hs, hsl, 0);
 153                                 break;
 154                         case 'h':
 155                                 cspace(&HS, ps, psl, REPLACE);
 156                                 break;
 157                         case 'H':
 158                                 cspace(&HS, "\n", 1, 0);
 159                                 cspace(&HS, ps, psl, 0);
 160                                 break;
 161                         case 'i':
 162                                 (void)printf("%s", cp->t);
 163                                 break;
 164                         case 'l':
 165                                 lputs(ps);
 166                                 break;
 167                         case 'n':
 168                                 if (!nflag && !pd)
 169                                         OUT(ps)
 170                                 flush_appends();
 171                                 if (!mf_fgets(&PS, REPLACE))
 172                                         exit(0);
 173                                 pd = 0;
 174                                 break;
 175                         case 'N':
 176                                 flush_appends();
 177                                 cspace(&PS, "\n", 1, 0);
 178                                 if (!mf_fgets(&PS, 0))
 179                                         exit(0);
 180                                 break;
 181                         case 'p':
 182                                 if (pd)
 183                                         break;
 184                                 OUT(ps)
 185                                 break;
 186                         case 'P':
 187                                 if (pd)
 188                                         break;
 189                                 if (psl != 0 &&
 190                                     (p = memchr(ps, '\n', psl)) != NULL) {
 191                                         oldpsl = psl;
 192                                         psl = p - ps;
 193                                 }
 194                                 OUT(ps)
 195                                 if (p != NULL)
 196                                         psl = oldpsl;
 197                                 break;
 198                         case 'q':
 199                                 if (!nflag && !pd)
 200                                         OUT(ps)
 201                                 flush_appends();
 202                                 exit(0);
 203                         case 'r':
 204                                 if (appendx >= appendnum)
 205                                         if ((appends = realloc(appends,
 206                                             sizeof(struct s_appends) *
 207                                             (appendnum *= 2))) == NULL)
 208                                                 err(1, "realloc");
 209                                 appends[appendx].type = AP_FILE;
 210                                 appends[appendx].s = cp->t;
 211                                 appends[appendx].len = strlen(cp->t);
 212                                 appendx++;
 213                                 break;
 214                         case 's':
 215                                 sdone |= substitute(cp);
 216                                 break;
 217                         case 't':
 218                                 if (sdone) {
 219                                         sdone = 0;
 220                                         cp = cp->u.c;
 221                                         goto redirect;
 222                                 }
 223                                 break;
 224                         case 'w':
 225                                 if (pd)
 226                                         break;
 227                                 if (cp->u.fd == -1 && (cp->u.fd = open(cp->t,
 228                                     O_WRONLY|O_APPEND|O_CREAT|O_TRUNC,
 229                                     DEFFILEMODE)) == -1)
 230                                         err(1, "%s", cp->t);
 231                                 if (write(cp->u.fd, ps, psl) != psl ||
 232                                     write(cp->u.fd, "\n", 1) != 1)
 233                                         err(1, "%s", cp->t);
 234                                 break;
 235                         case 'x':
 236                                 if (hs == NULL)
 237                                         cspace(&HS, "", 0, REPLACE);
 238                                 tspace = PS;
 239                                 PS = HS;
 240                                 HS = tspace;
 241                                 break;
 242                         case 'y':
 243                                 if (pd || psl == 0)
 244                                         break;
 245                                 for (p = ps, len = psl; len--; ++p)
 246                                         *p = cp->u.y[(unsigned char)*p];
 247                                 break;
 248                         case ':':
 249                         case '}':
 250                                 break;
 251                         case '=':
 252                                 (void)printf("%lu\n", linenum);
 253                         }
 254                         cp = cp->next;
 255                 } /* for all cp */
 256
 257 new:            if (!nflag && !pd)
 258                         OUT(ps)
 259                 flush_appends();
 260         } /* for all lines */
 261 }
 262
 263 /*
 264  * TRUE if the address passed matches the current program state
 265  * (lastline, linenumber, ps).
 266  */
 267 #define MATCH(a)                                                \
 268         (a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, psl) :       \
 269             (a)->type == AT_LINE ? linenum == (a)->u.l : lastline()
 270
 271 /*
 272  * Return TRUE if the command applies to the current line.  Sets the inrange
 273  * flag to process ranges.  Interprets the non-select (``!'') flag.
 274  */
 275 static __inline int
 276 applies(struct s_command *cp)
 277 {
 278         int r;
 279
 280         lastaddr = 0;
 281         if (cp->a1 == NULL && cp->a2 == NULL)
 282                 r = 1;
 283         else if (cp->a2)
 284                 if (cp->inrange) {
 285                         if (MATCH(cp->a2)) {
 286                                 cp->inrange = 0;
 287                                 lastaddr = 1;
 288                         }
 289                         r = 1;
 290                 } else if (MATCH(cp->a1)) {
 291                         /*
 292                          * If the second address is a number less than or
 293                          * equal to the line number first selected, only
 294                          * one line shall be selected.
 295                          *      -- POSIX 1003.2
 296                          */
 297                         if (cp->a2->type == AT_LINE &&
 298                             linenum >= cp->a2->u.l)
 299                                 lastaddr = 1;
 300                         else
 301                                 cp->inrange = 1;
 302                         r = 1;
 303                 } else
 304                         r = 0;
 305         else
 306                 r = MATCH(cp->a1);
 307         return (cp->nonsel ? ! r : r);
 308 }
 309
 310 /*
 311  * substitute --
 312  *      Do substitutions in the pattern space.  Currently, we build a
 313  *      copy of the new pattern space in the substitute space structure
 314  *      and then swap them.
 315  */
 316 static int
 317 substitute(struct s_command *cp)
 318 {
 319         SPACE tspace;
 320         regex_t *re;
 321         regoff_t re_off, slen;
 322         int lastempty, n;
 323         char *s;
 324
 325         s = ps;
 326         re = cp->u.s->re;
 327         if (re == NULL) {
 328                 if (defpreg != NULL && cp->u.s->maxbref > defpreg->re_nsub) {
 329                         linenum = cp->u.s->linenum;
 330                         errx(1, "%lu: %s: \\%d not defined in the RE",
 331                                         linenum, fname, cp->u.s->maxbref);
 332                 }
 333         }
 334         if (!regexec_e(re, s, 0, 0, psl))
 335                 return (0);
 336
 337         SS.len = 0;                             /* Clean substitute space. */
 338         slen = psl;
 339         n = cp->u.s->n;
 340         lastempty = 1;
 341
 342         switch (n) {
 343         case 0:                                 /* Global */
 344                 do {
 345                         if (lastempty || match[0].rm_so != match[0].rm_eo) {
 346                                 /* Locate start of replaced string. */
 347                                 re_off = match[0].rm_so;
 348                                 /* Copy leading retained string. */
 349                                 cspace(&SS, s, re_off, APPEND);
 350                                 /* Add in regular expression. */
 351                                 regsub(&SS, s, cp->u.s->new);
 352                         }
 353
 354                         /* Move past this match. */
 355                         if (match[0].rm_so != match[0].rm_eo) {
 356                                 s += match[0].rm_eo;
 357                                 slen -= match[0].rm_eo;
 358                                 lastempty = 0;
 359                         } else {
 360                                 if (match[0].rm_so < slen)
 361                                         cspace(&SS, s + match[0].rm_so, 1,
 362                                             APPEND);
 363                                 s += match[0].rm_so + 1;
 364                                 slen -= match[0].rm_so + 1;
 365                                 lastempty = 1;
 366                         }
 367                 } while (slen >= 0 && regexec_e(re, s, REG_NOTBOL, 0, slen));
 368                 /* Copy trailing retained string. */
 369                 if (slen > 0)
 370                         cspace(&SS, s, slen, APPEND);
 371                 break;
 372         default:                                /* Nth occurrence */
 373                 while (--n) {
 374                         if (match[0].rm_eo == match[0].rm_so)
 375                                 match[0].rm_eo = match[0].rm_so + 1;
 376                         s += match[0].rm_eo;
 377                         slen -= match[0].rm_eo;
 378                         if (slen < 0)
 379                                 return (0);
 380                         if (!regexec_e(re, s, REG_NOTBOL, 0, slen))
 381                                 return (0);
 382                 }
 383                 /* FALLTHROUGH */
 384         case 1:                                 /* 1st occurrence */
 385                 /* Locate start of replaced string. */
 386                 re_off = match[0].rm_so + (s - ps);
 387                 /* Copy leading retained string. */
 388                 cspace(&SS, ps, re_off, APPEND);
 389                 /* Add in regular expression. */
 390                 regsub(&SS, s, cp->u.s->new);
 391                 /* Copy trailing retained string. */
 392                 s += match[0].rm_eo;
 393                 slen -= match[0].rm_eo;
 394                 cspace(&SS, s, slen, APPEND);
 395                 break;
 396         }
 397
 398         /*
 399          * Swap the substitute space and the pattern space, and make sure
 400          * that any leftover pointers into stdio memory get lost.
 401          */
 402         tspace = PS;
 403         PS = SS;
 404         SS = tspace;
 405         SS.space = SS.back;
 406
 407         /* Handle the 'p' flag. */
 408         if (cp->u.s->p)
 409                 OUT(ps)
 410
 411         /* Handle the 'w' flag. */
 412         if (cp->u.s->wfile && !pd) {
 413                 if (cp->u.s->wfd == -1 && (cp->u.s->wfd = open(cp->u.s->wfile,
 414                     O_WRONLY|O_APPEND|O_CREAT|O_TRUNC, DEFFILEMODE)) == -1)
 415                         err(1, "%s", cp->u.s->wfile);
 416                 if (write(cp->u.s->wfd, ps, psl) != psl ||
 417                     write(cp->u.s->wfd, "\n", 1) != 1)
 418                         err(1, "%s", cp->u.s->wfile);
 419         }
 420         return (1);
 421 }
 422
 423 /*
 424  * Flush append requests.  Always called before reading a line,
 425  * therefore it also resets the substitution done (sdone) flag.
 426  */
 427 static void
 428 flush_appends(void)
 429 {
 430         FILE *f;
 431         int count, i;
 432         char buf[8 * 1024];
 433
 434         for (i = 0; i < appendx; i++)
 435                 switch (appends[i].type) {
 436                 case AP_STRING:
 437                         fwrite(appends[i].s, sizeof(char), appends[i].len,
 438                             stdout);
 439                         break;
 440                 case AP_FILE:
 441                         /*
 442                          * Read files probably shouldn't be cached.  Since
 443                          * it's not an error to read a non-existent file,
 444                          * it's possible that another program is interacting
 445                          * with the sed script through the filesystem.  It
 446                          * would be truly bizarre, but possible.  It's probably
 447                          * not that big a performance win, anyhow.
 448                          */
 449                         if ((f = fopen(appends[i].s, "r")) == NULL)
 450                                 break;
 451                         while ((count = fread(buf, sizeof(char), sizeof(buf), f)))
 452                                 (void)fwrite(buf, sizeof(char), count, stdout);
 453                         (void)fclose(f);
 454                         break;
 455                 }
 456         if (ferror(stdout))
 457                 errx(1, "stdout: %s", strerror(errno ? errno : EIO));
 458         appendx = sdone = 0;
 459 }
 460
 461 static void
 462 lputs(char *s)
 463 {
 464         int count;
 465         const char *escapes;
 466         char *p;
 467         struct winsize win;
 468         static int termwidth = -1;
 469
 470         if (termwidth == -1) {
 471                 if ((p = getenv("COLUMNS")) && *p != '\0')
 472                         termwidth = atoi(p);
 473                 else if (ioctl(STDOUT_FILENO, TIOCGWINSZ, &win) == 0 &&
 474                     win.ws_col > 0)
 475                         termwidth = win.ws_col;
 476                 else
 477                         termwidth = 60;
 478         }
 479
 480         for (count = 0; *s; ++s) {
 481                 if (count + 5 >= termwidth) {
 482                         (void)printf("\\\n");
 483                         count = 0;
 484                 }
 485                 if (isprint((unsigned char)*s) && *s != '\\') {
 486                         (void)putchar(*s);
 487                         count++;
 488                 } else if (*s == '\n') {
 489                         (void)putchar('$');
 490                         (void)putchar('\n');
 491                         count = 0;
 492                 } else {
 493                         escapes = "\\\a\b\f\r\t\v";
 494                         (void)putchar('\\');
 495                         if ((p = strchr(escapes, *s))) {
 496                                 (void)putchar("\\abfrtv"[p - escapes]);
 497                                 count += 2;
 498                         } else {
 499                                 (void)printf("%03o", *(u_char *)s);
 500                                 count += 4;
 501                         }
 502                 }
 503         }
 504         (void)putchar('$');
 505         (void)putchar('\n');
 506         if (ferror(stdout))
 507                 errx(1, "stdout: %s", strerror(errno ? errno : EIO));
 508 }
 509
 510 static __inline int
 511 regexec_e(regex_t *preg, const char *string, int eflags, int nomatch,
 512           size_t slen)
 513 {
 514         int eval;
 515
 516         if (preg == NULL) {
 517                 if (defpreg == NULL)
 518                         errx(1, "first RE may not be empty");
 519         } else
 520                 defpreg = preg;
 521
 522         /* Set anchors */
 523         match[0].rm_so = 0;
 524         match[0].rm_eo = slen;
 525
 526         eval = regexec(defpreg, string,
 527             nomatch ? 0 : maxnsub + 1, match, eflags | REG_STARTEND);
 528         switch(eval) {
 529         case 0:
 530                 return (1);
 531         case REG_NOMATCH:
 532                 return (0);
 533         }
 534         errx(1, "RE error: %s", strregerror(eval, defpreg));
 535         /* NOTREACHED */
 536 }
 537
 538 /*
 539  * regsub - perform substitutions after a regexp match
 540  * Based on a routine by Henry Spencer
 541  */
 542 static void
 543 regsub(SPACE *sp, char *string, char *src)
 544 {
 545         int len, no;
 546         char c, *dst;
 547
 548 #define NEEDSP(reqlen)                                                  \
 549         /* XXX What is the +1 for? */                                   \
 550         if (sp->len + (reqlen) + 1 >= sp->blen) {                       \
 551                 sp->blen += (reqlen) + 1024;                            \
 552                 if ((sp->space = sp->back = realloc(sp->back, sp->blen)) \
 553                     == NULL)                                            \
 554                         err(1, "realloc");                              \
 555                 dst = sp->space + sp->len;                              \
 556         }
 557
 558         dst = sp->space + sp->len;
 559         while ((c = *src++) != '\0') {
 560                 if (c == '&')
 561                         no = 0;
 562                 else if (c == '\\' && isdigit((unsigned char)*src))
 563                         no = *src++ - '0';
 564                 else
 565                         no = -1;
 566                 if (no < 0) {           /* Ordinary character. */
 567                         if (c == '\\' && (*src == '\\' || *src == '&'))
 568                                 c = *src++;
 569                         NEEDSP(1);
 570                         *dst++ = c;
 571                         ++sp->len;
 572                 } else if (match[no].rm_so != -1 && match[no].rm_eo != -1) {
 573                         len = match[no].rm_eo - match[no].rm_so;
 574                         NEEDSP(len);
 575                         memmove(dst, string + match[no].rm_so, len);
 576                         dst += len;
 577                         sp->len += len;
 578                 }
 579         }
 580         NEEDSP(1);
 581         *dst = '\0';
 582 }
 583
 584 /*
 585  * aspace --
 586  *      Append the source space to the destination space, allocating new
 587  *      space as necessary.
 588  */
 589 void
 590 cspace(SPACE *sp, const char *p, size_t len, enum e_spflag spflag)
 591 {
 592         size_t tlen;
 593
 594         /* Make sure SPACE has enough memory and ramp up quickly. */
 595         tlen = sp->len + len + 1;
 596         if (tlen > sp->blen) {
 597                 sp->blen = tlen + 1024;
 598                 if ((sp->space = sp->back = realloc(sp->back, sp->blen)) ==
 599                     NULL)
 600                         err(1, "realloc");
 601         }
 602
 603         if (spflag == REPLACE)
 604                 sp->len = 0;
 605
 606         memmove(sp->space + sp->len, p, len);
 607
 608         sp->space[sp->len += len] = '\0';
 609 }
 610
 611 /*
 612  * Close all cached opened files and report any errors
 613  */
 614 void
 615 cfclose(struct s_command *cp, struct s_command *end)
 616 {
 617
 618         for (; cp != end; cp = cp->next)
 619                 switch(cp->code) {
 620                 case 's':
 621                         if (cp->u.s->wfd != -1 && close(cp->u.s->wfd))
 622                                 err(1, "%s", cp->u.s->wfile);
 623                         cp->u.s->wfd = -1;
 624                         break;
 625                 case 'w':
 626                         if (cp->u.fd != -1 && close(cp->u.fd))
 627                                 err(1, "%s", cp->t);
 628                         cp->u.fd = -1;
 629                         break;
 630                 case '{':
 631                         cfclose(cp->u.c, cp->next);
 632                         break;
 633                 }
 634 }