Annotation of src/usr.bin/sed/compile.c, Revision 1.1
1.1 ! deraadt 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[] = "@(#)compile.c 8.1 (Berkeley) 6/6/93"; */
! 40: static char *rcsid = "$Id: compile.c,v 1.14 1995/03/09 11:19:24 mycroft Exp $";
! 41: #endif /* not lint */
! 42:
! 43: #include <sys/types.h>
! 44: #include <sys/stat.h>
! 45:
! 46: #include <ctype.h>
! 47: #include <errno.h>
! 48: #include <fcntl.h>
! 49: #include <limits.h>
! 50: #include <regex.h>
! 51: #include <stdio.h>
! 52: #include <stdlib.h>
! 53: #include <string.h>
! 54:
! 55: #include "defs.h"
! 56: #include "extern.h"
! 57:
! 58: #define LHSZ 128
! 59: #define LHMASK (LHSZ - 1)
! 60: static struct labhash {
! 61: struct labhash *lh_next;
! 62: u_int lh_hash;
! 63: struct s_command *lh_cmd;
! 64: int lh_ref;
! 65: } *labels[LHSZ];
! 66:
! 67: static char *compile_addr __P((char *, struct s_addr *));
! 68: static char *compile_ccl __P((char **, char *));
! 69: static char *compile_delimited __P((char *, char *));
! 70: static char *compile_flags __P((char *, struct s_subst *));
! 71: static char *compile_re __P((char *, regex_t **));
! 72: static char *compile_subst __P((char *, struct s_subst *));
! 73: static char *compile_text __P((void));
! 74: static char *compile_tr __P((char *, char **));
! 75: static struct s_command
! 76: **compile_stream __P((struct s_command **));
! 77: static char *duptoeol __P((char *, char *));
! 78: static void enterlabel __P((struct s_command *));
! 79: static struct s_command
! 80: *findlabel __P((char *));
! 81: static void fixuplabel __P((struct s_command *, struct s_command *));
! 82: static void uselabel __P((void));
! 83:
! 84: /*
! 85: * Command specification. This is used to drive the command parser.
! 86: */
! 87: struct s_format {
! 88: char code; /* Command code */
! 89: int naddr; /* Number of address args */
! 90: enum e_args args; /* Argument type */
! 91: };
! 92:
! 93: static struct s_format cmd_fmts[] = {
! 94: {'{', 2, GROUP},
! 95: {'}', 0, ENDGROUP},
! 96: {'a', 1, TEXT},
! 97: {'b', 2, BRANCH},
! 98: {'c', 2, TEXT},
! 99: {'d', 2, EMPTY},
! 100: {'D', 2, EMPTY},
! 101: {'g', 2, EMPTY},
! 102: {'G', 2, EMPTY},
! 103: {'h', 2, EMPTY},
! 104: {'H', 2, EMPTY},
! 105: {'i', 1, TEXT},
! 106: {'l', 2, EMPTY},
! 107: {'n', 2, EMPTY},
! 108: {'N', 2, EMPTY},
! 109: {'p', 2, EMPTY},
! 110: {'P', 2, EMPTY},
! 111: {'q', 1, EMPTY},
! 112: {'r', 1, RFILE},
! 113: {'s', 2, SUBST},
! 114: {'t', 2, BRANCH},
! 115: {'w', 2, WFILE},
! 116: {'x', 2, EMPTY},
! 117: {'y', 2, TR},
! 118: {'!', 2, NONSEL},
! 119: {':', 0, LABEL},
! 120: {'#', 0, COMMENT},
! 121: {'=', 1, EMPTY},
! 122: {'\0', 0, COMMENT},
! 123: };
! 124:
! 125: /* The compiled program. */
! 126: struct s_command *prog;
! 127:
! 128: /*
! 129: * Compile the program into prog.
! 130: * Initialise appends.
! 131: */
! 132: void
! 133: compile()
! 134: {
! 135: *compile_stream(&prog) = NULL;
! 136: fixuplabel(prog, NULL);
! 137: uselabel();
! 138: appends = xmalloc(sizeof(struct s_appends) * appendnum);
! 139: match = xmalloc((maxnsub + 1) * sizeof(regmatch_t));
! 140: }
! 141:
! 142: #define EATSPACE() do { \
! 143: if (p) \
! 144: while (*p && isascii(*p) && isspace(*p)) \
! 145: p++; \
! 146: } while (0)
! 147:
! 148: static struct s_command **
! 149: compile_stream(link)
! 150: struct s_command **link;
! 151: {
! 152: register char *p;
! 153: static char lbuf[_POSIX2_LINE_MAX + 1]; /* To save stack */
! 154: struct s_command *cmd, *cmd2, *stack;
! 155: struct s_format *fp;
! 156: int naddr; /* Number of addresses */
! 157:
! 158: stack = 0;
! 159: for (;;) {
! 160: if ((p = cu_fgets(lbuf, sizeof(lbuf))) == NULL) {
! 161: if (stack != 0)
! 162: err(COMPILE, "unexpected EOF (pending }'s)");
! 163: return (link);
! 164: }
! 165:
! 166: semicolon: EATSPACE();
! 167: if (p && (*p == '#' || *p == '\0'))
! 168: continue;
! 169: *link = cmd = xmalloc(sizeof(struct s_command));
! 170: link = &cmd->next;
! 171: cmd->nonsel = cmd->inrange = 0;
! 172: /* First parse the addresses */
! 173: naddr = 0;
! 174:
! 175: /* Valid characters to start an address */
! 176: #define addrchar(c) (strchr("0123456789/\\$", (c)))
! 177: if (addrchar(*p)) {
! 178: naddr++;
! 179: cmd->a1 = xmalloc(sizeof(struct s_addr));
! 180: p = compile_addr(p, cmd->a1);
! 181: EATSPACE(); /* EXTENSION */
! 182: if (*p == ',') {
! 183: p++;
! 184: EATSPACE(); /* EXTENSION */
! 185: naddr++;
! 186: cmd->a2 = xmalloc(sizeof(struct s_addr));
! 187: p = compile_addr(p, cmd->a2);
! 188: EATSPACE();
! 189: } else
! 190: cmd->a2 = 0;
! 191: } else
! 192: cmd->a1 = cmd->a2 = 0;
! 193:
! 194: nonsel: /* Now parse the command */
! 195: if (!*p)
! 196: err(COMPILE, "command expected");
! 197: cmd->code = *p;
! 198: for (fp = cmd_fmts; fp->code; fp++)
! 199: if (fp->code == *p)
! 200: break;
! 201: if (!fp->code)
! 202: err(COMPILE, "invalid command code %c", *p);
! 203: if (naddr > fp->naddr)
! 204: err(COMPILE,
! 205: "command %c expects up to %d address(es), found %d", *p, fp->naddr, naddr);
! 206: switch (fp->args) {
! 207: case NONSEL: /* ! */
! 208: p++;
! 209: EATSPACE();
! 210: cmd->nonsel = ! cmd->nonsel;
! 211: goto nonsel;
! 212: case GROUP: /* { */
! 213: p++;
! 214: EATSPACE();
! 215: cmd->next = stack;
! 216: stack = cmd;
! 217: link = &cmd->u.c;
! 218: if (*p)
! 219: goto semicolon;
! 220: break;
! 221: case ENDGROUP:
! 222: /*
! 223: * Short-circuit command processing, since end of
! 224: * group is really just a noop.
! 225: */
! 226: cmd->nonsel = 1;
! 227: if (stack == 0)
! 228: err(COMPILE, "unexpected }");
! 229: cmd2 = stack;
! 230: stack = cmd2->next;
! 231: cmd2->next = cmd;
! 232: /*FALLTHROUGH*/
! 233: case EMPTY: /* d D g G h H l n N p P q x = \0 */
! 234: p++;
! 235: EATSPACE();
! 236: if (*p == ';') {
! 237: p++;
! 238: link = &cmd->next;
! 239: goto semicolon;
! 240: }
! 241: if (*p)
! 242: err(COMPILE,
! 243: "extra characters at the end of %c command", cmd->code);
! 244: break;
! 245: case TEXT: /* a c i */
! 246: p++;
! 247: EATSPACE();
! 248: if (*p != '\\')
! 249: err(COMPILE,
! 250: "command %c expects \\ followed by text", cmd->code);
! 251: p++;
! 252: EATSPACE();
! 253: if (*p)
! 254: err(COMPILE,
! 255: "extra characters after \\ at the end of %c command", cmd->code);
! 256: cmd->t = compile_text();
! 257: break;
! 258: case COMMENT: /* \0 # */
! 259: break;
! 260: case WFILE: /* w */
! 261: p++;
! 262: EATSPACE();
! 263: if (*p == '\0')
! 264: err(COMPILE, "filename expected");
! 265: cmd->t = duptoeol(p, "w command");
! 266: if (aflag)
! 267: cmd->u.fd = -1;
! 268: else if ((cmd->u.fd = open(p,
! 269: O_WRONLY|O_APPEND|O_CREAT|O_TRUNC,
! 270: DEFFILEMODE)) == -1)
! 271: err(FATAL, "%s: %s\n", p, strerror(errno));
! 272: break;
! 273: case RFILE: /* r */
! 274: p++;
! 275: EATSPACE();
! 276: if (*p == '\0')
! 277: err(COMPILE, "filename expected");
! 278: else
! 279: cmd->t = duptoeol(p, "read command");
! 280: break;
! 281: case BRANCH: /* b t */
! 282: p++;
! 283: EATSPACE();
! 284: if (*p == '\0')
! 285: cmd->t = NULL;
! 286: else
! 287: cmd->t = duptoeol(p, "branch");
! 288: break;
! 289: case LABEL: /* : */
! 290: p++;
! 291: EATSPACE();
! 292: cmd->t = duptoeol(p, "label");
! 293: if (strlen(p) == 0)
! 294: err(COMPILE, "empty label");
! 295: enterlabel(cmd);
! 296: break;
! 297: case SUBST: /* s */
! 298: p++;
! 299: if (*p == '\0' || *p == '\\')
! 300: err(COMPILE,
! 301: "substitute pattern can not be delimited by newline or backslash");
! 302: cmd->u.s = xmalloc(sizeof(struct s_subst));
! 303: p = compile_re(p, &cmd->u.s->re);
! 304: if (p == NULL)
! 305: err(COMPILE, "unterminated substitute pattern");
! 306: --p;
! 307: p = compile_subst(p, cmd->u.s);
! 308: p = compile_flags(p, cmd->u.s);
! 309: EATSPACE();
! 310: if (*p == ';') {
! 311: p++;
! 312: link = &cmd->next;
! 313: goto semicolon;
! 314: }
! 315: break;
! 316: case TR: /* y */
! 317: p++;
! 318: p = compile_tr(p, (char **)&cmd->u.y);
! 319: EATSPACE();
! 320: if (*p == ';') {
! 321: p++;
! 322: link = &cmd->next;
! 323: goto semicolon;
! 324: }
! 325: if (*p)
! 326: err(COMPILE,
! 327: "extra text at the end of a transform command");
! 328: break;
! 329: }
! 330: }
! 331: }
! 332:
! 333: /*
! 334: * Get a delimited string. P points to the delimeter of the string; d points
! 335: * to a buffer area. Newline and delimiter escapes are processed; other
! 336: * escapes are ignored.
! 337: *
! 338: * Returns a pointer to the first character after the final delimiter or NULL
! 339: * in the case of a non-terminated string. The character array d is filled
! 340: * with the processed string.
! 341: */
! 342: static char *
! 343: compile_delimited(p, d)
! 344: char *p, *d;
! 345: {
! 346: char c;
! 347:
! 348: c = *p++;
! 349: if (c == '\0')
! 350: return (NULL);
! 351: else if (c == '\\')
! 352: err(COMPILE, "\\ can not be used as a string delimiter");
! 353: else if (c == '\n')
! 354: err(COMPILE, "newline can not be used as a string delimiter");
! 355: while (*p) {
! 356: if (*p == '[') {
! 357: if ((d = compile_ccl(&p, d)) == NULL)
! 358: err(COMPILE, "unbalanced brackets ([])");
! 359: continue;
! 360: } else if (*p == '\\' && p[1] == '[') {
! 361: *d++ = *p++;
! 362: } else if (*p == '\\' && p[1] == c)
! 363: p++;
! 364: else if (*p == '\\' && p[1] == 'n') {
! 365: *d++ = '\n';
! 366: p += 2;
! 367: continue;
! 368: } else if (*p == '\\' && p[1] == '\\')
! 369: *d++ = *p++;
! 370: else if (*p == c) {
! 371: *d = '\0';
! 372: return (p + 1);
! 373: }
! 374: *d++ = *p++;
! 375: }
! 376: return (NULL);
! 377: }
! 378:
! 379:
! 380: /* compile_ccl: expand a POSIX character class */
! 381: static char *
! 382: compile_ccl(sp, t)
! 383: char **sp;
! 384: char *t;
! 385: {
! 386: int c, d;
! 387: char *s = *sp;
! 388:
! 389: *t++ = *s++;
! 390: if (*s == '^')
! 391: *t++ = *s++;
! 392: if (*s == ']')
! 393: *t++ = *s++;
! 394: for (; *s && (*t = *s) != ']'; s++, t++)
! 395: if (*s == '[' && ((d = *(s+1)) == '.' || d == ':' || d == '=')) {
! 396: *++t = *++s, t++, s++;
! 397: for (c = *s; (*t = *s) != ']' || c != d; s++, t++)
! 398: if ((c = *s) == '\0')
! 399: return NULL;
! 400: } else if (*s == '\\' && s[1] == 'n')
! 401: *t = '\n', s++;
! 402: return (*s == ']') ? *sp = ++s, ++t : NULL;
! 403: }
! 404:
! 405: /*
! 406: * Get a regular expression. P points to the delimiter of the regular
! 407: * expression; repp points to the address of a regexp pointer. Newline
! 408: * and delimiter escapes are processed; other escapes are ignored.
! 409: * Returns a pointer to the first character after the final delimiter
! 410: * or NULL in the case of a non terminated regular expression. The regexp
! 411: * pointer is set to the compiled regular expression.
! 412: * Cflags are passed to regcomp.
! 413: */
! 414: static char *
! 415: compile_re(p, repp)
! 416: char *p;
! 417: regex_t **repp;
! 418: {
! 419: int eval;
! 420: char re[_POSIX2_LINE_MAX + 1];
! 421:
! 422: p = compile_delimited(p, re);
! 423: if (p && strlen(re) == 0) {
! 424: *repp = NULL;
! 425: return (p);
! 426: }
! 427: *repp = xmalloc(sizeof(regex_t));
! 428: if (p && (eval = regcomp(*repp, re, 0)) != 0)
! 429: err(COMPILE, "RE error: %s", strregerror(eval, *repp));
! 430: if (maxnsub < (*repp)->re_nsub)
! 431: maxnsub = (*repp)->re_nsub;
! 432: return (p);
! 433: }
! 434:
! 435: /*
! 436: * Compile the substitution string of a regular expression and set res to
! 437: * point to a saved copy of it. Nsub is the number of parenthesized regular
! 438: * expressions.
! 439: */
! 440: static char *
! 441: compile_subst(p, s)
! 442: char *p;
! 443: struct s_subst *s;
! 444: {
! 445: static char lbuf[_POSIX2_LINE_MAX + 1];
! 446: int asize, ref, size;
! 447: char c, *text, *op, *sp;
! 448:
! 449: c = *p++; /* Terminator character */
! 450: if (c == '\0')
! 451: return (NULL);
! 452:
! 453: s->maxbref = 0;
! 454: s->linenum = linenum;
! 455: asize = 2 * _POSIX2_LINE_MAX + 1;
! 456: text = xmalloc(asize);
! 457: size = 0;
! 458: do {
! 459: op = sp = text + size;
! 460: for (; *p; p++) {
! 461: if (*p == '\\') {
! 462: p++;
! 463: if (strchr("123456789", *p) != NULL) {
! 464: *sp++ = '\\';
! 465: ref = *p - '0';
! 466: if (s->re != NULL &&
! 467: ref > s->re->re_nsub)
! 468: err(COMPILE,
! 469: "\\%c not defined in the RE", *p);
! 470: if (s->maxbref < ref)
! 471: s->maxbref = ref;
! 472: } else if (*p == '&' || *p == '\\')
! 473: *sp++ = '\\';
! 474: } else if (*p == c) {
! 475: p++;
! 476: *sp++ = '\0';
! 477: size += sp - op;
! 478: s->new = xrealloc(text, size);
! 479: return (p);
! 480: } else if (*p == '\n') {
! 481: err(COMPILE,
! 482: "unescaped newline inside substitute pattern");
! 483: /* NOTREACHED */
! 484: }
! 485: *sp++ = *p;
! 486: }
! 487: size += sp - op;
! 488: if (asize - size < _POSIX2_LINE_MAX + 1) {
! 489: asize *= 2;
! 490: text = xmalloc(asize);
! 491: }
! 492: } while (cu_fgets(p = lbuf, sizeof(lbuf)));
! 493: err(COMPILE, "unterminated substitute in regular expression");
! 494: /* NOTREACHED */
! 495: }
! 496:
! 497: /*
! 498: * Compile the flags of the s command
! 499: */
! 500: static char *
! 501: compile_flags(p, s)
! 502: char *p;
! 503: struct s_subst *s;
! 504: {
! 505: int gn; /* True if we have seen g or n */
! 506: char wfile[_POSIX2_LINE_MAX + 1], *q;
! 507:
! 508: s->n = 1; /* Default */
! 509: s->p = 0;
! 510: s->wfile = NULL;
! 511: s->wfd = -1;
! 512: for (gn = 0;;) {
! 513: EATSPACE(); /* EXTENSION */
! 514: switch (*p) {
! 515: case 'g':
! 516: if (gn)
! 517: err(COMPILE,
! 518: "more than one number or 'g' in substitute flags");
! 519: gn = 1;
! 520: s->n = 0;
! 521: break;
! 522: case '\0':
! 523: case '\n':
! 524: case ';':
! 525: return (p);
! 526: case 'p':
! 527: s->p = 1;
! 528: break;
! 529: case '1': case '2': case '3':
! 530: case '4': case '5': case '6':
! 531: case '7': case '8': case '9':
! 532: if (gn)
! 533: err(COMPILE,
! 534: "more than one number or 'g' in substitute flags");
! 535: gn = 1;
! 536: /* XXX Check for overflow */
! 537: s->n = (int)strtol(p, &p, 10);
! 538: break;
! 539: case 'w':
! 540: p++;
! 541: #ifdef HISTORIC_PRACTICE
! 542: if (*p != ' ') {
! 543: err(WARNING, "space missing before w wfile");
! 544: return (p);
! 545: }
! 546: #endif
! 547: EATSPACE();
! 548: q = wfile;
! 549: while (*p) {
! 550: if (*p == '\n')
! 551: break;
! 552: *q++ = *p++;
! 553: }
! 554: *q = '\0';
! 555: if (q == wfile)
! 556: err(COMPILE, "no wfile specified");
! 557: s->wfile = strdup(wfile);
! 558: if (!aflag && (s->wfd = open(wfile,
! 559: O_WRONLY|O_APPEND|O_CREAT|O_TRUNC,
! 560: DEFFILEMODE)) == -1)
! 561: err(FATAL, "%s: %s\n", wfile, strerror(errno));
! 562: return (p);
! 563: default:
! 564: err(COMPILE,
! 565: "bad flag in substitute command: '%c'", *p);
! 566: break;
! 567: }
! 568: p++;
! 569: }
! 570: }
! 571:
! 572: /*
! 573: * Compile a translation set of strings into a lookup table.
! 574: */
! 575: static char *
! 576: compile_tr(p, transtab)
! 577: char *p;
! 578: char **transtab;
! 579: {
! 580: int i;
! 581: char *lt, *op, *np;
! 582: char old[_POSIX2_LINE_MAX + 1];
! 583: char new[_POSIX2_LINE_MAX + 1];
! 584:
! 585: if (*p == '\0' || *p == '\\')
! 586: err(COMPILE,
! 587: "transform pattern can not be delimited by newline or backslash");
! 588: p = compile_delimited(p, old);
! 589: if (p == NULL) {
! 590: err(COMPILE, "unterminated transform source string");
! 591: return (NULL);
! 592: }
! 593: p = compile_delimited(--p, new);
! 594: if (p == NULL) {
! 595: err(COMPILE, "unterminated transform target string");
! 596: return (NULL);
! 597: }
! 598: EATSPACE();
! 599: if (strlen(new) != strlen(old)) {
! 600: err(COMPILE, "transform strings are not the same length");
! 601: return (NULL);
! 602: }
! 603: /* We assume characters are 8 bits */
! 604: lt = xmalloc(UCHAR_MAX);
! 605: for (i = 0; i <= UCHAR_MAX; i++)
! 606: lt[i] = (char)i;
! 607: for (op = old, np = new; *op; op++, np++)
! 608: lt[(u_char)*op] = *np;
! 609: *transtab = lt;
! 610: return (p);
! 611: }
! 612:
! 613: /*
! 614: * Compile the text following an a or i command.
! 615: */
! 616: static char *
! 617: compile_text()
! 618: {
! 619: int asize, size;
! 620: char *text, *p, *op, *s;
! 621: char lbuf[_POSIX2_LINE_MAX + 1];
! 622:
! 623: asize = 2 * _POSIX2_LINE_MAX + 1;
! 624: text = xmalloc(asize);
! 625: size = 0;
! 626: while (cu_fgets(lbuf, sizeof(lbuf))) {
! 627: op = s = text + size;
! 628: p = lbuf;
! 629: EATSPACE();
! 630: for (; *p; p++) {
! 631: if (*p == '\\')
! 632: p++;
! 633: *s++ = *p;
! 634: }
! 635: size += s - op;
! 636: if (p[-2] != '\\') {
! 637: *s = '\0';
! 638: break;
! 639: }
! 640: if (asize - size < _POSIX2_LINE_MAX + 1) {
! 641: asize *= 2;
! 642: text = xmalloc(asize);
! 643: }
! 644: }
! 645: return (xrealloc(text, size + 1));
! 646: }
! 647:
! 648: /*
! 649: * Get an address and return a pointer to the first character after
! 650: * it. Fill the structure pointed to according to the address.
! 651: */
! 652: static char *
! 653: compile_addr(p, a)
! 654: char *p;
! 655: struct s_addr *a;
! 656: {
! 657: char *end;
! 658:
! 659: switch (*p) {
! 660: case '\\': /* Context address */
! 661: ++p;
! 662: /* FALLTHROUGH */
! 663: case '/': /* Context address */
! 664: p = compile_re(p, &a->u.r);
! 665: if (p == NULL)
! 666: err(COMPILE, "unterminated regular expression");
! 667: a->type = AT_RE;
! 668: return (p);
! 669:
! 670: case '$': /* Last line */
! 671: a->type = AT_LAST;
! 672: return (p + 1);
! 673: /* Line number */
! 674: case '0': case '1': case '2': case '3': case '4':
! 675: case '5': case '6': case '7': case '8': case '9':
! 676: a->type = AT_LINE;
! 677: a->u.l = strtol(p, &end, 10);
! 678: return (end);
! 679: default:
! 680: err(COMPILE, "expected context address");
! 681: return (NULL);
! 682: }
! 683: }
! 684:
! 685: /*
! 686: * duptoeol --
! 687: * Return a copy of all the characters up to \n or \0.
! 688: */
! 689: static char *
! 690: duptoeol(s, ctype)
! 691: register char *s;
! 692: char *ctype;
! 693: {
! 694: size_t len;
! 695: int ws;
! 696: char *start;
! 697:
! 698: ws = 0;
! 699: for (start = s; *s != '\0' && *s != '\n'; ++s)
! 700: ws = isspace(*s);
! 701: *s = '\0';
! 702: if (ws)
! 703: err(WARNING, "whitespace after %s", ctype);
! 704: len = s - start + 1;
! 705: return (memmove(xmalloc(len), start, len));
! 706: }
! 707:
! 708: /*
! 709: * Convert goto label names to addresses, and count a and r commands, in
! 710: * the given subset of the script. Free the memory used by labels in b
! 711: * and t commands (but not by :).
! 712: *
! 713: * TODO: Remove } nodes
! 714: */
! 715: static void
! 716: fixuplabel(cp, end)
! 717: struct s_command *cp, *end;
! 718: {
! 719:
! 720: for (; cp != end; cp = cp->next)
! 721: switch (cp->code) {
! 722: case 'a':
! 723: case 'r':
! 724: appendnum++;
! 725: break;
! 726: case 'b':
! 727: case 't':
! 728: /* Resolve branch target. */
! 729: if (cp->t == NULL) {
! 730: cp->u.c = NULL;
! 731: break;
! 732: }
! 733: if ((cp->u.c = findlabel(cp->t)) == NULL)
! 734: err(COMPILE2, "undefined label '%s'", cp->t);
! 735: free(cp->t);
! 736: break;
! 737: case '{':
! 738: /* Do interior commands. */
! 739: fixuplabel(cp->u.c, cp->next);
! 740: break;
! 741: }
! 742: }
! 743:
! 744: /*
! 745: * Associate the given command label for later lookup.
! 746: */
! 747: static void
! 748: enterlabel(cp)
! 749: struct s_command *cp;
! 750: {
! 751: register struct labhash **lhp, *lh;
! 752: register u_char *p;
! 753: register u_int h, c;
! 754:
! 755: for (h = 0, p = (u_char *)cp->t; (c = *p) != 0; p++)
! 756: h = (h << 5) + h + c;
! 757: lhp = &labels[h & LHMASK];
! 758: for (lh = *lhp; lh != NULL; lh = lh->lh_next)
! 759: if (lh->lh_hash == h && strcmp(cp->t, lh->lh_cmd->t) == 0)
! 760: err(COMPILE2, "duplicate label '%s'", cp->t);
! 761: lh = xmalloc(sizeof *lh);
! 762: lh->lh_next = *lhp;
! 763: lh->lh_hash = h;
! 764: lh->lh_cmd = cp;
! 765: lh->lh_ref = 0;
! 766: *lhp = lh;
! 767: }
! 768:
! 769: /*
! 770: * Find the label contained in the command l in the command linked
! 771: * list cp. L is excluded from the search. Return NULL if not found.
! 772: */
! 773: static struct s_command *
! 774: findlabel(name)
! 775: char *name;
! 776: {
! 777: register struct labhash *lh;
! 778: register u_char *p;
! 779: register u_int h, c;
! 780:
! 781: for (h = 0, p = (u_char *)name; (c = *p) != 0; p++)
! 782: h = (h << 5) + h + c;
! 783: for (lh = labels[h & LHMASK]; lh != NULL; lh = lh->lh_next) {
! 784: if (lh->lh_hash == h && strcmp(name, lh->lh_cmd->t) == 0) {
! 785: lh->lh_ref = 1;
! 786: return (lh->lh_cmd);
! 787: }
! 788: }
! 789: return (NULL);
! 790: }
! 791:
! 792: /*
! 793: * Warn about any unused labels. As a side effect, release the label hash
! 794: * table space.
! 795: */
! 796: static void
! 797: uselabel()
! 798: {
! 799: register struct labhash *lh, *next;
! 800: register int i;
! 801:
! 802: for (i = 0; i < LHSZ; i++) {
! 803: for (lh = labels[i]; lh != NULL; lh = next) {
! 804: next = lh->lh_next;
! 805: if (!lh->lh_ref)
! 806: err(WARNING, "unused label '%s'",
! 807: lh->lh_cmd->t);
! 808: free(lh);
! 809: }
! 810: }
! 811: }