process.c revision 1.9
1 /*- 2 * Copyright (c) 1992 Diomidis Spinellis. 3 * Copyright (c) 1992, 1993 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 #ifndef lint 39 /* from: static char sccsid[] = "@(#)process.c 8.1 (Berkeley) 6/6/93"; */ 40 static char *rcsid = "$Id: process.c,v 1.9 1994/02/03 23:44:57 cgd Exp $"; 41 #endif /* not lint */ 42 43 #include <sys/types.h> 44 #include <sys/stat.h> 45 #include <sys/ioctl.h> 46 #include <sys/uio.h> 47 48 #include <ctype.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 __P((struct s_command *)); 69 static void flush_appends __P((void)); 70 static void lputs __P((char *)); 71 static inline int regexec_e __P((regex_t *, const char *, int, int, size_t)); 72 static void regsub __P((SPACE *, char *, char *)); 73 static int substitute __P((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); } 87 88 void 89 process() 90 { 91 struct s_command *cp; 92 SPACE tspace; 93 size_t len; 94 int r; 95 char oldc, *p; 96 97 for (linenum = 0; mf_fgets(&PS, REPLACE);) { 98 pd = 0; 99 cp = prog; 100 redirect: 101 while (cp != NULL) { 102 if (!applies(cp)) { 103 cp = cp->next; 104 continue; 105 } 106 switch (cp->code) { 107 case '{': 108 cp = cp->u.c; 109 goto redirect; 110 case 'a': 111 if (appendx >= appendnum) 112 appends = xrealloc(appends, 113 sizeof(struct s_appends) * 114 (appendnum *= 2)); 115 appends[appendx].type = AP_STRING; 116 appends[appendx].s = cp->t; 117 appends[appendx].len = strlen(cp->t); 118 appendx++; 119 break; 120 case 'b': 121 cp = cp->u.c; 122 goto redirect; 123 case 'c': 124 pd = 1; 125 psl = 0; 126 if (cp->a2 == NULL || lastaddr) 127 (void)printf("%s", cp->t); 128 break; 129 case 'd': 130 pd = 1; 131 goto new; 132 case 'D': 133 if (pd) 134 goto new; 135 if ((p = memchr(ps, '\n', psl)) == NULL) 136 pd = 1; 137 else { 138 psl -= (p - ps) - 1; 139 memmove(ps, p + 1, psl); 140 } 141 goto new; 142 case 'g': 143 cspace(&PS, hs, hsl, REPLACE); 144 break; 145 case 'G': 146 cspace(&PS, hs, hsl, 0); 147 break; 148 case 'h': 149 cspace(&HS, ps, psl, REPLACE); 150 break; 151 case 'H': 152 cspace(&HS, ps, psl, 0); 153 break; 154 case 'i': 155 (void)printf("%s", cp->t); 156 break; 157 case 'l': 158 lputs(ps); 159 break; 160 case 'n': 161 if (!nflag && !pd) 162 OUT(ps) 163 flush_appends(); 164 r = mf_fgets(&PS, REPLACE); 165 #ifdef HISTORIC_PRACTICE 166 if (!r) 167 exit(0); 168 #endif 169 pd = 0; 170 break; 171 case 'N': 172 flush_appends(); 173 if (!mf_fgets(&PS, 0)) { 174 if (!nflag && !pd) 175 OUT(ps) 176 exit(0); 177 } 178 break; 179 case 'p': 180 if (pd) 181 break; 182 OUT(ps) 183 break; 184 case 'P': 185 if (pd) 186 break; 187 if ((p = memchr(ps, '\n', psl)) != NULL) { 188 oldc = *p; 189 *p = '\0'; 190 } 191 OUT(ps) 192 if (p != NULL) 193 *p = oldc; 194 break; 195 case 'q': 196 if (!nflag && !pd) 197 OUT(ps) 198 flush_appends(); 199 exit(0); 200 case 'r': 201 if (appendx >= appendnum) 202 appends = xrealloc(appends, 203 sizeof(struct s_appends) * 204 (appendnum *= 2)); 205 appends[appendx].type = AP_FILE; 206 appends[appendx].s = cp->t; 207 appends[appendx].len = strlen(cp->t); 208 appendx++; 209 break; 210 case 's': 211 sdone |= substitute(cp); 212 break; 213 case 't': 214 if (sdone) { 215 sdone = 0; 216 cp = cp->u.c; 217 goto redirect; 218 } 219 break; 220 case 'w': 221 if (pd) 222 break; 223 if (cp->u.fd == -1 && (cp->u.fd = open(cp->t, 224 O_WRONLY|O_APPEND|O_CREAT|O_TRUNC, 225 DEFFILEMODE)) == -1) 226 err(FATAL, "%s: %s\n", 227 cp->t, strerror(errno)); 228 if (write(cp->u.fd, ps, psl) != psl) 229 err(FATAL, "%s: %s\n", 230 cp->t, strerror(errno)); 231 break; 232 case 'x': 233 if (hs == NULL) 234 cspace(&HS, "", 0, REPLACE); 235 tspace = PS; 236 PS = HS; 237 HS = tspace; 238 break; 239 case 'y': 240 if (pd) 241 break; 242 for (p = ps, len = psl; --len; ++p) 243 *p = cp->u.y[*p]; 244 break; 245 case ':': 246 case '}': 247 break; 248 case '=': 249 (void)printf("%lu\n", linenum); 250 } 251 cp = cp->next; 252 } /* for all cp */ 253 254 new: if (!nflag && !pd) 255 OUT(ps) 256 flush_appends(); 257 } /* for all lines */ 258 } 259 260 /* 261 * TRUE if the address passed matches the current program state 262 * (lastline, linenumber, ps). 263 */ 264 #define MATCH(a) \ 265 (a)->type == AT_RE ? regexec_e((a)->u.r, ps, 0, 1, psl) : \ 266 (a)->type == AT_LINE ? linenum == (a)->u.l : lastline 267 268 /* 269 * Return TRUE if the command applies to the current line. Sets the inrange 270 * flag to process ranges. Interprets the non-select (``!'') flag. 271 */ 272 static inline int 273 applies(cp) 274 struct s_command *cp; 275 { 276 int r; 277 278 lastaddr = 0; 279 if (cp->a1 == NULL && cp->a2 == NULL) 280 r = 1; 281 else if (cp->a2) 282 if (cp->inrange) { 283 if (MATCH(cp->a2)) { 284 cp->inrange = 0; 285 lastaddr = 1; 286 } 287 r = 1; 288 } else if (MATCH(cp->a1)) { 289 /* 290 * If the second address is a number less than or 291 * equal to the line number first selected, only 292 * one line shall be selected. 293 * -- POSIX 1003.2 294 */ 295 if (cp->a2->type == AT_LINE && 296 linenum >= cp->a2->u.l) 297 lastaddr = 1; 298 else 299 cp->inrange = 1; 300 r = 1; 301 } else 302 r = 0; 303 else 304 r = MATCH(cp->a1); 305 return (cp->nonsel ? ! r : r); 306 } 307 308 /* 309 * substitute -- 310 * Do substitutions in the pattern space. Currently, we build a 311 * copy of the new pattern space in the substitute space structure 312 * and then swap them. 313 */ 314 static int 315 substitute(cp) 316 struct s_command *cp; 317 { 318 SPACE tspace; 319 regex_t *re; 320 size_t re_off, slen; 321 int n; 322 char *s; 323 324 s = ps; 325 re = cp->u.s->re; 326 if (re == NULL) { 327 if (defpreg != NULL && cp->u.s->maxbref > defpreg->re_nsub) { 328 linenum = cp->u.s->linenum; 329 err(COMPILE, "\\%d not defined in the RE", 330 cp->u.s->maxbref); 331 } 332 } 333 if (!regexec_e(re, s, 0, 0, psl)) 334 return (0); 335 336 SS.len = 0; /* Clean substitute space. */ 337 slen = psl; 338 n = cp->u.s->n; 339 switch (n) { 340 case 0: /* Global */ 341 do { 342 /* Locate start of replaced string. */ 343 re_off = match[0].rm_so; 344 /* Copy leading retained string. */ 345 cspace(&SS, s, re_off, APPEND); 346 /* Add in regular expression. */ 347 regsub(&SS, s, cp->u.s->new); 348 /* Move past this match. */ 349 s += match[0].rm_eo; 350 slen -= match[0].rm_eo; 351 } while(regexec_e(re, s, REG_NOTBOL, 0, slen)); 352 /* Copy trailing retained string. */ 353 cspace(&SS, s, slen, APPEND); 354 break; 355 default: /* Nth occurrence */ 356 while (--n) { 357 s += match[0].rm_eo; 358 slen -= match[0].rm_eo; 359 if (!regexec_e(re, s, REG_NOTBOL, 0, slen)) 360 return (0); 361 } 362 /* FALLTHROUGH */ 363 case 1: /* 1st occurrence */ 364 /* Locate start of replaced string. */ 365 re_off = match[0].rm_so + (s - ps); 366 /* Copy leading retained string. */ 367 cspace(&SS, ps, re_off, APPEND); 368 /* Add in regular expression. */ 369 regsub(&SS, s, cp->u.s->new); 370 /* Copy trailing retained string. */ 371 s += match[0].rm_eo; 372 slen -= match[0].rm_eo; 373 cspace(&SS, s, slen, APPEND); 374 break; 375 } 376 377 /* 378 * Swap the substitute space and the pattern space, and make sure 379 * that any leftover pointers into stdio memory get lost. 380 */ 381 tspace = PS; 382 PS = SS; 383 SS = tspace; 384 SS.space = SS.back; 385 386 /* Handle the 'p' flag. */ 387 if (cp->u.s->p) 388 OUT(ps) 389 390 /* Handle the 'w' flag. */ 391 if (cp->u.s->wfile && !pd) { 392 if (cp->u.s->wfd == -1 && (cp->u.s->wfd = open(cp->u.s->wfile, 393 O_WRONLY|O_APPEND|O_CREAT|O_TRUNC, DEFFILEMODE)) == -1) 394 err(FATAL, "%s: %s\n", cp->u.s->wfile, strerror(errno)); 395 if (write(cp->u.s->wfd, ps, psl) != psl) 396 err(FATAL, "%s: %s\n", cp->u.s->wfile, strerror(errno)); 397 } 398 return (1); 399 } 400 401 /* 402 * Flush append requests. Always called before reading a line, 403 * therefore it also resets the substitution done (sdone) flag. 404 */ 405 static void 406 flush_appends() 407 { 408 FILE *f; 409 int count, i; 410 char buf[8 * 1024]; 411 412 for (i = 0; i < appendx; i++) 413 switch (appends[i].type) { 414 case AP_STRING: 415 fwrite(appends[i].s, sizeof(char), appends[i].len, 416 stdout); 417 break; 418 case AP_FILE: 419 /* 420 * Read files probably shouldn't be cached. Since 421 * it's not an error to read a non-existent file, 422 * it's possible that another program is interacting 423 * with the sed script through the file system. It 424 * would be truly bizarre, but possible. It's probably 425 * not that big a performance win, anyhow. 426 */ 427 if ((f = fopen(appends[i].s, "r")) == NULL) 428 break; 429 while (count = fread(buf, sizeof(char), sizeof(buf), f)) 430 (void)fwrite(buf, sizeof(char), count, stdout); 431 (void)fclose(f); 432 break; 433 } 434 if (ferror(stdout)) 435 err(FATAL, "stdout: %s", strerror(errno ? errno : EIO)); 436 appendx = sdone = 0; 437 } 438 439 static void 440 lputs(s) 441 register char *s; 442 { 443 register int count; 444 register char *escapes, *p; 445 struct winsize win; 446 static int termwidth = -1; 447 448 if (termwidth == -1) 449 if (p = getenv("COLUMNS")) 450 termwidth = atoi(p); 451 else if (ioctl(STDOUT_FILENO, TIOCGWINSZ, &win) == 0 && 452 win.ws_col > 0) 453 termwidth = win.ws_col; 454 else 455 termwidth = 60; 456 457 for (count = 0; *s; ++s) { 458 if (count >= termwidth) { 459 (void)printf("\\\n"); 460 count = 0; 461 } 462 if (isascii(*s) && isprint(*s) && *s != '\\') { 463 (void)putchar(*s); 464 count++; 465 } else { 466 escapes = "\\\a\b\f\n\r\t\v"; 467 (void)putchar('\\'); 468 if (p = strchr(escapes, *s)) { 469 (void)putchar("\\abfnrtv"[p - escapes]); 470 count += 2; 471 } else { 472 (void)printf("%03o", *(u_char *)s); 473 count += 4; 474 } 475 } 476 } 477 (void)putchar('$'); 478 (void)putchar('\n'); 479 if (ferror(stdout)) 480 err(FATAL, "stdout: %s", strerror(errno ? errno : EIO)); 481 } 482 483 static inline int 484 regexec_e(preg, string, eflags, nomatch, slen) 485 regex_t *preg; 486 const char *string; 487 int eflags, nomatch; 488 size_t slen; 489 { 490 int eval; 491 492 if (preg == NULL) { 493 if (defpreg == NULL) 494 err(FATAL, "first RE may not be empty"); 495 } else 496 defpreg = preg; 497 498 /* Set anchors, discounting trailing newline (if any). */ 499 if (slen > 0 && string[slen - 1] == '\n') 500 slen--; 501 match[0].rm_so = 0; 502 match[0].rm_eo = slen; 503 504 eval = regexec(defpreg, string, 505 nomatch ? 0 : maxnsub + 1, match, eflags | REG_STARTEND); 506 switch(eval) { 507 case 0: 508 return (1); 509 case REG_NOMATCH: 510 return (0); 511 } 512 err(FATAL, "RE error: %s", strregerror(eval, defpreg)); 513 /* NOTREACHED */ 514 } 515 516 /* 517 * regsub - perform substitutions after a regexp match 518 * Based on a routine by Henry Spencer 519 */ 520 static void 521 regsub(sp, string, src) 522 SPACE *sp; 523 char *string, *src; 524 { 525 register int len, no; 526 register char c, *dst; 527 528 #define NEEDSP(reqlen) \ 529 if (sp->len >= sp->blen - (reqlen) - 1) { \ 530 sp->blen += (reqlen) + 1024; \ 531 sp->space = sp->back = xrealloc(sp->back, sp->blen); \ 532 dst = sp->space + sp->len; \ 533 } 534 535 dst = sp->space + sp->len; 536 while ((c = *src++) != '\0') { 537 if (c == '&') 538 no = 0; 539 else if (c == '\\' && isdigit(*src)) 540 no = *src++ - '0'; 541 else 542 no = -1; 543 if (no < 0) { /* Ordinary character. */ 544 if (c == '\\' && (*src == '\\' || *src == '&')) 545 c = *src++; 546 NEEDSP(1); 547 *dst++ = c; 548 ++sp->len; 549 } else if (match[no].rm_so != -1 && match[no].rm_eo != -1) { 550 len = match[no].rm_eo - match[no].rm_so; 551 NEEDSP(len); 552 memmove(dst, string + match[no].rm_so, len); 553 dst += len; 554 sp->len += len; 555 } 556 } 557 NEEDSP(1); 558 *dst = '\0'; 559 } 560 561 /* 562 * aspace -- 563 * Append the source space to the destination space, allocating new 564 * space as necessary. 565 */ 566 void 567 cspace(sp, p, len, spflag) 568 SPACE *sp; 569 char *p; 570 size_t len; 571 enum e_spflag spflag; 572 { 573 size_t tlen; 574 575 /* Make sure SPACE has enough memory and ramp up quickly. */ 576 tlen = sp->len + len + 1; 577 if (tlen > sp->blen) { 578 sp->blen = tlen + 1024; 579 sp->space = sp->back = xrealloc(sp->back, sp->blen); 580 } 581 582 if (spflag == REPLACE) 583 sp->len = 0; 584 585 memmove(sp->space + sp->len, p, len); 586 587 sp->space[sp->len += len] = '\0'; 588 } 589 590 /* 591 * Close all cached opened files and report any errors 592 */ 593 void 594 cfclose(cp, end) 595 register struct s_command *cp, *end; 596 { 597 598 for (; cp != end; cp = cp->next) 599 switch(cp->code) { 600 case 's': 601 if (cp->u.s->wfd != -1 && close(cp->u.s->wfd)) 602 err(FATAL, 603 "%s: %s", cp->u.s->wfile, strerror(errno)); 604 cp->u.s->wfd = -1; 605 break; 606 case 'w': 607 if (cp->u.fd != -1 && close(cp->u.fd)) 608 err(FATAL, "%s: %s", cp->t, strerror(errno)); 609 cp->u.fd = -1; 610 break; 611 case '{': 612 cfclose(cp->u.c, cp->next); 613 break; 614 } 615 } 616
Indexes created Thu Oct 02 01:09:59 GMT 2025