process.c revision 97801
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 38#include <sys/cdefs.h> 39__FBSDID("$FreeBSD: head/usr.bin/sed/process.c 97801 2002-06-04 10:00:08Z tjr $"); 40 41#ifndef lint 42static const char sccsid[] = "@(#)process.c 8.6 (Berkeley) 4/20/94"; 43#endif 44 45#include <sys/types.h> 46#include <sys/stat.h> 47#include <sys/ioctl.h> 48#include <sys/uio.h> 49 50#include <ctype.h> 51#include <err.h> 52#include <errno.h> 53#include <fcntl.h> 54#include <limits.h> 55#include <regex.h> 56#include <stdio.h> 57#include <stdlib.h> 58#include <string.h> 59#include <unistd.h> 60 61#include "defs.h" 62#include "extern.h" 63 64static SPACE HS, PS, SS; 65#define pd PS.deleted 66#define ps PS.space 67#define psl PS.len 68#define hs HS.space 69#define hsl HS.len 70 71static inline int applies(struct s_command *); 72static void flush_appends(void); 73static void lputs(char *); 74static inline int regexec_e(regex_t *, const char *, int, int, size_t); 75static void regsub(SPACE *, char *, char *); 76static int substitute(struct s_command *); 77 78struct s_appends *appends; /* Array of pointers to strings to append. */ 79static int appendx; /* Index into appends array. */ 80int appendnum; /* Size of appends array. */ 81 82static int lastaddr; /* Set by applies if last address of a range. */ 83static int sdone; /* If any substitutes since last line input. */ 84 /* Iov structure for 'w' commands. */ 85static regex_t *defpreg; 86size_t maxnsub; 87regmatch_t *match; 88 89#define OUT(s) { fwrite(s, sizeof(u_char), psl, stdout); } 90 91void 92process() 93{ 94 struct s_command *cp; 95 SPACE tspace; 96 size_t len, oldpsl = 0; 97 char *p; 98 99 for (linenum = 0; mf_fgets(&PS, REPLACE);) { 100 pd = 0; 101top: 102 cp = prog; 103redirect: 104 while (cp != NULL) { 105 if (!applies(cp)) { 106 cp = cp->next; 107 continue; 108 } 109 switch (cp->code) { 110 case '{': 111 cp = cp->u.c; 112 goto redirect; 113 case 'a': 114 if (appendx >= appendnum) 115 if ((appends = realloc(appends, 116 sizeof(struct s_appends) * 117 (appendnum *= 2))) == NULL) 118 err(1, "realloc"); 119 appends[appendx].type = AP_STRING; 120 appends[appendx].s = cp->t; 121 appends[appendx].len = strlen(cp->t); 122 appendx++; 123 break; 124 case 'b': 125 cp = cp->u.c; 126 goto redirect; 127 case 'c': 128 pd = 1; 129 psl = 0; 130 if (cp->a2 == NULL || lastaddr) 131 (void)printf("%s", cp->t); 132 break; 133 case 'd': 134 pd = 1; 135 goto new; 136 case 'D': 137 if (pd) 138 goto new; 139 if (psl == 0 || 140 (p = memchr(ps, '\n', psl - 1)) == NULL) { 141 pd = 1; 142 goto new; 143 } else { 144 psl -= (p + 1) - ps; 145 memmove(ps, p + 1, psl); 146 goto top; 147 } 148 case 'g': 149 cspace(&PS, hs, hsl, REPLACE); 150 break; 151 case 'G': 152 if (hs == NULL) 153 cspace(&HS, "\n", 1, REPLACE); 154 cspace(&PS, hs, hsl, 0); 155 break; 156 case 'h': 157 cspace(&HS, ps, psl, REPLACE); 158 break; 159 case 'H': 160 cspace(&HS, ps, psl, 0); 161 break; 162 case 'i': 163 (void)printf("%s", cp->t); 164 break; 165 case 'l': 166 lputs(ps); 167 break; 168 case 'n': 169 if (!nflag && !pd) 170 OUT(ps) 171 flush_appends(); 172 if (!mf_fgets(&PS, REPLACE)) 173 exit(0); 174 pd = 0; 175 break; 176 case 'N': 177 flush_appends(); 178 if (!mf_fgets(&PS, 0)) { 179 if (!nflag && !pd) 180 OUT(ps) 181 exit(0); 182 } 183 break; 184 case 'p': 185 if (pd) 186 break; 187 OUT(ps) 188 break; 189 case 'P': 190 if (pd) 191 break; 192 if (psl != 0 && 193 (p = memchr(ps, '\n', psl - 1)) != NULL) { 194 oldpsl = psl; 195 psl = (p + 1) - ps; 196 } 197 OUT(ps) 198 if (p != NULL) 199 psl = oldpsl; 200 break; 201 case 'q': 202 if (!nflag && !pd) 203 OUT(ps) 204 flush_appends(); 205 exit(0); 206 case 'r': 207 if (appendx >= appendnum) 208 if ((appends = realloc(appends, 209 sizeof(struct s_appends) * 210 (appendnum *= 2))) == NULL) 211 err(1, "realloc"); 212 appends[appendx].type = AP_FILE; 213 appends[appendx].s = cp->t; 214 appends[appendx].len = strlen(cp->t); 215 appendx++; 216 break; 217 case 's': 218 sdone |= substitute(cp); 219 break; 220 case 't': 221 if (sdone) { 222 sdone = 0; 223 cp = cp->u.c; 224 goto redirect; 225 } 226 break; 227 case 'w': 228 if (pd) 229 break; 230 if (cp->u.fd == -1 && (cp->u.fd = open(cp->t, 231 O_WRONLY|O_APPEND|O_CREAT|O_TRUNC, 232 DEFFILEMODE)) == -1) 233 err(1, "%s", cp->t); 234 if (write(cp->u.fd, ps, psl) != psl) 235 err(1, "%s", cp->t); 236 break; 237 case 'x': 238 if (hs == NULL) 239 cspace(&HS, "\n", 1, REPLACE); 240 tspace = PS; 241 PS = HS; 242 HS = tspace; 243 break; 244 case 'y': 245 if (pd || psl == 0) 246 break; 247 for (p = ps, len = psl; --len; ++p) 248 *p = cp->u.y[(unsigned char)*p]; 249 break; 250 case ':': 251 case '}': 252 break; 253 case '=': 254 (void)printf("%lu\n", linenum); 255 } 256 cp = cp->next; 257 } /* for all cp */ 258 259new: if (!nflag && !pd) 260 OUT(ps) 261 flush_appends(); 262 } /* for all lines */ 263} 264 265/* 266 * TRUE if the address passed matches the current program state 267 * (lastline, linenumber, ps). 268 */ 269#define MATCH(a) \ 270 (a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, psl) : \ 271 (a)->type == AT_LINE ? linenum == (a)->u.l : lastline 272 273/* 274 * Return TRUE if the command applies to the current line. Sets the inrange 275 * flag to process ranges. Interprets the non-select (``!'') flag. 276 */ 277static inline int 278applies(cp) 279 struct s_command *cp; 280{ 281 int r; 282 283 lastaddr = 0; 284 if (cp->a1 == NULL && cp->a2 == NULL) 285 r = 1; 286 else if (cp->a2) 287 if (cp->inrange) { 288 if (MATCH(cp->a2)) { 289 cp->inrange = 0; 290 lastaddr = 1; 291 } 292 r = 1; 293 } else if (MATCH(cp->a1)) { 294 /* 295 * If the second address is a number less than or 296 * equal to the line number first selected, only 297 * one line shall be selected. 298 * -- POSIX 1003.2 299 */ 300 if (cp->a2->type == AT_LINE && 301 linenum >= cp->a2->u.l) 302 lastaddr = 1; 303 else 304 cp->inrange = 1; 305 r = 1; 306 } else 307 r = 0; 308 else 309 r = MATCH(cp->a1); 310 return (cp->nonsel ? ! r : r); 311} 312 313/* 314 * substitute -- 315 * Do substitutions in the pattern space. Currently, we build a 316 * copy of the new pattern space in the substitute space structure 317 * and then swap them. 318 */ 319static int 320substitute(cp) 321 struct s_command *cp; 322{ 323 SPACE tspace; 324 regex_t *re; 325 size_t re_off, slen; 326 int lastempty, n; 327 char *s; 328 329 s = ps; 330 re = cp->u.s->re; 331 if (re == NULL) { 332 if (defpreg != NULL && cp->u.s->maxbref > defpreg->re_nsub) { 333 linenum = cp->u.s->linenum; 334 errx(1, "%lu: %s: \\%d not defined in the RE", 335 linenum, fname, cp->u.s->maxbref); 336 } 337 } 338 if (!regexec_e(re, s, 0, 0, psl)) 339 return (0); 340 341 SS.len = 0; /* Clean substitute space. */ 342 slen = psl; 343 n = cp->u.s->n; 344 lastempty = 1; 345 346 switch (n) { 347 case 0: /* Global */ 348 do { 349 if (lastempty || match[0].rm_so != match[0].rm_eo) { 350 /* Locate start of replaced string. */ 351 re_off = match[0].rm_so; 352 /* Copy leading retained string. */ 353 cspace(&SS, s, re_off, APPEND); 354 /* Add in regular expression. */ 355 regsub(&SS, s, cp->u.s->new); 356 } 357 358 /* Move past this match. */ 359 if (match[0].rm_so != match[0].rm_eo) { 360 s += match[0].rm_eo; 361 slen -= match[0].rm_eo; 362 lastempty = 0; 363 } else { 364 if (match[0].rm_so == 0) 365 cspace(&SS, s, match[0].rm_so + 1, 366 APPEND); 367 else 368 cspace(&SS, s + match[0].rm_so, 1, 369 APPEND); 370 s += match[0].rm_so + 1; 371 slen -= match[0].rm_so + 1; 372 lastempty = 1; 373 } 374 } while (slen > 0 && regexec_e(re, s, REG_NOTBOL, 0, slen)); 375 /* Copy trailing retained string. */ 376 if (slen > 0) 377 cspace(&SS, s, slen, APPEND); 378 break; 379 default: /* Nth occurrence */ 380 while (--n) { 381 s += match[0].rm_eo; 382 slen -= match[0].rm_eo; 383 if (!regexec_e(re, s, REG_NOTBOL, 0, slen)) 384 return (0); 385 } 386 /* FALLTHROUGH */ 387 case 1: /* 1st occurrence */ 388 /* Locate start of replaced string. */ 389 re_off = match[0].rm_so + (s - ps); 390 /* Copy leading retained string. */ 391 cspace(&SS, ps, re_off, APPEND); 392 /* Add in regular expression. */ 393 regsub(&SS, s, cp->u.s->new); 394 /* Copy trailing retained string. */ 395 s += match[0].rm_eo; 396 slen -= match[0].rm_eo; 397 cspace(&SS, s, slen, APPEND); 398 break; 399 } 400 401 /* 402 * Swap the substitute space and the pattern space, and make sure 403 * that any leftover pointers into stdio memory get lost. 404 */ 405 tspace = PS; 406 PS = SS; 407 SS = tspace; 408 SS.space = SS.back; 409 410 /* Handle the 'p' flag. */ 411 if (cp->u.s->p) 412 OUT(ps) 413 414 /* Handle the 'w' flag. */ 415 if (cp->u.s->wfile && !pd) { 416 if (cp->u.s->wfd == -1 && (cp->u.s->wfd = open(cp->u.s->wfile, 417 O_WRONLY|O_APPEND|O_CREAT|O_TRUNC, DEFFILEMODE)) == -1) 418 err(1, "%s", cp->u.s->wfile); 419 if (write(cp->u.s->wfd, ps, psl) != psl) 420 err(1, "%s", cp->u.s->wfile); 421 } 422 return (1); 423} 424 425/* 426 * Flush append requests. Always called before reading a line, 427 * therefore it also resets the substitution done (sdone) flag. 428 */ 429static void 430flush_appends() 431{ 432 FILE *f; 433 int count, i; 434 char buf[8 * 1024]; 435 436 for (i = 0; i < appendx; i++) 437 switch (appends[i].type) { 438 case AP_STRING: 439 fwrite(appends[i].s, sizeof(char), appends[i].len, 440 stdout); 441 break; 442 case AP_FILE: 443 /* 444 * Read files probably shouldn't be cached. Since 445 * it's not an error to read a non-existent file, 446 * it's possible that another program is interacting 447 * with the sed script through the filesystem. It 448 * would be truly bizarre, but possible. It's probably 449 * not that big a performance win, anyhow. 450 */ 451 if ((f = fopen(appends[i].s, "r")) == NULL) 452 break; 453 while ((count = fread(buf, sizeof(char), sizeof(buf), f))) 454 (void)fwrite(buf, sizeof(char), count, stdout); 455 (void)fclose(f); 456 break; 457 } 458 if (ferror(stdout)) 459 errx(1, "stdout: %s", strerror(errno ? errno : EIO)); 460 appendx = sdone = 0; 461} 462 463static void 464lputs(s) 465 char *s; 466{ 467 int count; 468 char *escapes, *p; 469 struct winsize win; 470 static int termwidth = -1; 471 472 if (termwidth == -1) { 473 if ((p = getenv("COLUMNS")) && *p != '\0') 474 termwidth = atoi(p); 475 else if (ioctl(STDOUT_FILENO, TIOCGWINSZ, &win) == 0 && 476 win.ws_col > 0) 477 termwidth = win.ws_col; 478 else 479 termwidth = 60; 480 } 481 482 for (count = 0; *s; ++s) { 483 if (count >= termwidth) { 484 (void)printf("\\\n"); 485 count = 0; 486 } 487 if (isprint((unsigned char)*s) && *s != '\\') { 488 (void)putchar(*s); 489 count++; 490 } else { 491 escapes = "\\\a\b\f\n\r\t\v"; 492 (void)putchar('\\'); 493 if ((p = strchr(escapes, *s))) { 494 (void)putchar("\\abfnrtv"[p - escapes]); 495 count += 2; 496 } else { 497 (void)printf("%03o", *(u_char *)s); 498 count += 4; 499 } 500 } 501 } 502 (void)putchar('$'); 503 (void)putchar('\n'); 504 if (ferror(stdout)) 505 errx(1, "stdout: %s", strerror(errno ? errno : EIO)); 506} 507 508static inline int 509regexec_e(preg, string, eflags, nomatch, slen) 510 regex_t *preg; 511 const char *string; 512 int eflags, nomatch; 513 size_t slen; 514{ 515 int eval; 516 517 if (preg == NULL) { 518 if (defpreg == NULL) 519 errx(1, "first RE may not be empty"); 520 } else 521 defpreg = preg; 522 523 /* Set anchors, discounting trailing newline (if any). */ 524 if (slen > 0 && string[slen - 1] == '\n') 525 slen--; 526 match[0].rm_so = 0; 527 match[0].rm_eo = slen; 528 529 eval = regexec(defpreg, string, 530 nomatch ? 0 : maxnsub + 1, match, eflags | REG_STARTEND); 531 switch(eval) { 532 case 0: 533 return (1); 534 case REG_NOMATCH: 535 return (0); 536 } 537 errx(1, "RE error: %s", strregerror(eval, defpreg)); 538 /* NOTREACHED */ 539} 540 541/* 542 * regsub - perform substitutions after a regexp match 543 * Based on a routine by Henry Spencer 544 */ 545static void 546regsub(sp, string, src) 547 SPACE *sp; 548 char *string, *src; 549{ 550 int len, no; 551 char c, *dst; 552 553#define NEEDSP(reqlen) \ 554 if (sp->len >= sp->blen - (reqlen) - 1) { \ 555 sp->blen += (reqlen) + 1024; \ 556 if ((sp->space = sp->back = realloc(sp->back, sp->blen)) \ 557 == NULL) \ 558 err(1, "realloc"); \ 559 dst = sp->space + sp->len; \ 560 } 561 562 dst = sp->space + sp->len; 563 while ((c = *src++) != '\0') { 564 if (c == '&') 565 no = 0; 566 else if (c == '\\' && isdigit((unsigned char)*src)) 567 no = *src++ - '0'; 568 else 569 no = -1; 570 if (no < 0) { /* Ordinary character. */ 571 if (c == '\\' && (*src == '\\' || *src == '&')) 572 c = *src++; 573 NEEDSP(1); 574 *dst++ = c; 575 ++sp->len; 576 } else if (match[no].rm_so != -1 && match[no].rm_eo != -1) { 577 len = match[no].rm_eo - match[no].rm_so; 578 NEEDSP(len); 579 memmove(dst, string + match[no].rm_so, len); 580 dst += len; 581 sp->len += len; 582 } 583 } 584 NEEDSP(1); 585 *dst = '\0'; 586} 587 588/* 589 * aspace -- 590 * Append the source space to the destination space, allocating new 591 * space as necessary. 592 */ 593void 594cspace(sp, p, len, spflag) 595 SPACE *sp; 596 char *p; 597 size_t len; 598 enum e_spflag spflag; 599{ 600 size_t tlen; 601 602 /* Make sure SPACE has enough memory and ramp up quickly. */ 603 tlen = sp->len + len + 1; 604 if (tlen > sp->blen) { 605 sp->blen = tlen + 1024; 606 if ((sp->space = sp->back = realloc(sp->back, sp->blen)) == 607 NULL) 608 err(1, "realloc"); 609 } 610 611 if (spflag == REPLACE) 612 sp->len = 0; 613 614 memmove(sp->space + sp->len, p, len); 615 616 sp->space[sp->len += len] = '\0'; 617} 618 619/* 620 * Close all cached opened files and report any errors 621 */ 622void 623cfclose(cp, end) 624 struct s_command *cp, *end; 625{ 626 627 for (; cp != end; cp = cp->next) 628 switch(cp->code) { 629 case 's': 630 if (cp->u.s->wfd != -1 && close(cp->u.s->wfd)) 631 err(1, "%s", cp->u.s->wfile); 632 cp->u.s->wfd = -1; 633 break; 634 case 'w': 635 if (cp->u.fd != -1 && close(cp->u.fd)) 636 err(1, "%s", cp->t); 637 cp->u.fd = -1; 638 break; 639 case '{': 640 cfclose(cp->u.c, cp->next); 641 break; 642 } 643} 644