Index: stable/9/lib/libc/regex/engine.c =================================================================== --- stable/9/lib/libc/regex/engine.c (revision 302168) +++ stable/9/lib/libc/regex/engine.c (revision 302169) @@ -1,1183 +1,1183 @@ /*- * Copyright (c) 1992, 1993, 1994 Henry Spencer. * Copyright (c) 1992, 1993, 1994 * The Regents of the University of California. All rights reserved. * * This code is derived from software contributed to Berkeley by * Henry Spencer. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * @(#)engine.c 8.5 (Berkeley) 3/20/94 */ #include __FBSDID("$FreeBSD$"); /* * The matching engine and friends. This file is #included by regexec.c * after suitable #defines of a variety of macros used herein, so that * different state representations can be used without duplicating masses * of code. */ #ifdef SNAMES #define matcher smatcher #define fast sfast #define slow sslow #define dissect sdissect #define backref sbackref #define step sstep #define print sprint #define at sat #define match smat #endif #ifdef LNAMES #define matcher lmatcher #define fast lfast #define slow lslow #define dissect ldissect #define backref lbackref #define step lstep #define print lprint #define at lat #define match lmat #endif #ifdef MNAMES #define matcher mmatcher #define fast mfast #define slow mslow #define dissect mdissect #define backref mbackref #define step mstep #define print mprint #define at mat #define match mmat #endif /* another structure passed up and down to avoid zillions of parameters */ struct match { struct re_guts *g; int eflags; regmatch_t *pmatch; /* [nsub+1] (0 element unused) */ const char *offp; /* offsets work from here */ const char *beginp; /* start of string -- virtual NUL precedes */ const char *endp; /* end of string -- virtual NUL here */ const char *coldp; /* can be no match starting before here */ const char **lastpos; /* [nplus+1] */ STATEVARS; states st; /* current states */ states fresh; /* states for a fresh start */ states tmp; /* temporary */ states empty; /* empty set of states */ mbstate_t mbs; /* multibyte conversion state */ }; /* ========= begin header generated by ./mkh ========= */ #ifdef __cplusplus extern "C" { #endif /* === engine.c === */ static int matcher(struct re_guts *g, const char *string, size_t nmatch, regmatch_t pmatch[], int eflags); static const char *dissect(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); static const char *backref(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst, sopno lev, int); static const char *fast(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); static const char *slow(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); static states step(struct re_guts *g, sopno start, sopno stop, states bef, wint_t ch, states aft); #define MAX_RECURSION 100 #define BOL (OUT-1) #define EOL (BOL-1) #define BOLEOL (BOL-2) #define NOTHING (BOL-3) #define BOW (BOL-4) #define EOW (BOL-5) #define BADCHAR (BOL-6) #define NONCHAR(c) ((c) <= OUT) #ifdef REDEBUG static void print(struct match *m, const char *caption, states st, int ch, FILE *d); #endif #ifdef REDEBUG static void at(struct match *m, const char *title, const char *start, const char *stop, sopno startst, sopno stopst); #endif #ifdef REDEBUG static const char *pchar(int ch); #endif #ifdef __cplusplus } #endif /* ========= end header generated by ./mkh ========= */ #ifdef REDEBUG #define SP(t, s, c) print(m, t, s, c, stdout) #define AT(t, p1, p2, s1, s2) at(m, t, p1, p2, s1, s2) #define NOTE(str) { if (m->eflags®_TRACE) printf("=%s\n", (str)); } #else #define SP(t, s, c) /* nothing */ #define AT(t, p1, p2, s1, s2) /* nothing */ #define NOTE(s) /* nothing */ #endif /* - matcher - the actual matching engine == static int matcher(struct re_guts *g, const char *string, \ == size_t nmatch, regmatch_t pmatch[], int eflags); */ static int /* 0 success, REG_NOMATCH failure */ matcher(struct re_guts *g, const char *string, size_t nmatch, regmatch_t pmatch[], int eflags) { const char *endp; size_t i; struct match mv; struct match *m = &mv; const char *dp; const sopno gf = g->firststate+1; /* +1 for OEND */ const sopno gl = g->laststate; const char *start; const char *stop; /* Boyer-Moore algorithms variables */ const char *pp; int cj, mj; const char *mustfirst; const char *mustlast; int *matchjump; int *charjump; /* simplify the situation where possible */ if (g->cflags®_NOSUB) nmatch = 0; if (eflags®_STARTEND) { start = string + pmatch[0].rm_so; stop = string + pmatch[0].rm_eo; } else { start = string; stop = start + strlen(start); } if (stop < start) return(REG_INVARG); /* prescreening; this does wonders for this rather slow code */ if (g->must != NULL) { if (g->charjump != NULL && g->matchjump != NULL) { mustfirst = g->must; mustlast = g->must + g->mlen - 1; charjump = g->charjump; matchjump = g->matchjump; pp = mustlast; for (dp = start+g->mlen-1; dp < stop;) { /* Fast skip non-matches */ while (dp < stop && charjump[(int)*dp]) dp += charjump[(int)*dp]; if (dp >= stop) break; /* Greedy matcher */ /* We depend on not being used for * for strings of length 1 */ while (*--dp == *--pp && pp != mustfirst); if (*dp == *pp) break; /* Jump to next possible match */ mj = matchjump[pp - mustfirst]; cj = charjump[(int)*dp]; dp += (cj < mj ? mj : cj); pp = mustlast; } if (pp != mustfirst) return(REG_NOMATCH); } else { for (dp = start; dp < stop; dp++) if (*dp == g->must[0] && stop - dp >= g->mlen && memcmp(dp, g->must, (size_t)g->mlen) == 0) break; if (dp == stop) /* we didn't find g->must */ return(REG_NOMATCH); } } /* match struct setup */ m->g = g; m->eflags = eflags; m->pmatch = NULL; m->lastpos = NULL; m->offp = string; m->beginp = start; m->endp = stop; STATESETUP(m, 4); SETUP(m->st); SETUP(m->fresh); SETUP(m->tmp); SETUP(m->empty); CLEAR(m->empty); ZAPSTATE(&m->mbs); /* Adjust start according to moffset, to speed things up */ if (g->moffset > -1) start = ((dp - g->moffset) < start) ? start : dp - g->moffset; SP("mloop", m->st, *start); /* this loop does only one repetition except for backrefs */ for (;;) { endp = fast(m, start, stop, gf, gl); if (endp == NULL) { /* a miss */ if (m->pmatch != NULL) free((char *)m->pmatch); if (m->lastpos != NULL) free((char *)m->lastpos); STATETEARDOWN(m); return(REG_NOMATCH); } if (nmatch == 0 && !g->backrefs) break; /* no further info needed */ /* where? */ assert(m->coldp != NULL); for (;;) { NOTE("finding start"); endp = slow(m, m->coldp, stop, gf, gl); if (endp != NULL) break; assert(m->coldp < m->endp); m->coldp += XMBRTOWC(NULL, m->coldp, m->endp - m->coldp, &m->mbs, 0); } if (nmatch == 1 && !g->backrefs) break; /* no further info needed */ /* oh my, he wants the subexpressions... */ if (m->pmatch == NULL) m->pmatch = (regmatch_t *)malloc((m->g->nsub + 1) * sizeof(regmatch_t)); if (m->pmatch == NULL) { STATETEARDOWN(m); return(REG_ESPACE); } for (i = 1; i <= m->g->nsub; i++) m->pmatch[i].rm_so = m->pmatch[i].rm_eo = -1; if (!g->backrefs && !(m->eflags®_BACKR)) { NOTE("dissecting"); dp = dissect(m, m->coldp, endp, gf, gl); } else { if (g->nplus > 0 && m->lastpos == NULL) m->lastpos = malloc((g->nplus+1) * sizeof(const char *)); if (g->nplus > 0 && m->lastpos == NULL) { free(m->pmatch); STATETEARDOWN(m); return(REG_ESPACE); } NOTE("backref dissect"); dp = backref(m, m->coldp, endp, gf, gl, (sopno)0, 0); } if (dp != NULL) break; /* uh-oh... we couldn't find a subexpression-level match */ assert(g->backrefs); /* must be back references doing it */ assert(g->nplus == 0 || m->lastpos != NULL); for (;;) { if (dp != NULL || endp <= m->coldp) break; /* defeat */ NOTE("backoff"); endp = slow(m, m->coldp, endp-1, gf, gl); if (endp == NULL) break; /* defeat */ /* try it on a shorter possibility */ #ifndef NDEBUG for (i = 1; i <= m->g->nsub; i++) { assert(m->pmatch[i].rm_so == -1); assert(m->pmatch[i].rm_eo == -1); } #endif NOTE("backoff dissect"); dp = backref(m, m->coldp, endp, gf, gl, (sopno)0, 0); } assert(dp == NULL || dp == endp); if (dp != NULL) /* found a shorter one */ break; /* despite initial appearances, there is no match here */ NOTE("false alarm"); /* recycle starting later */ start = m->coldp + XMBRTOWC(NULL, m->coldp, stop - m->coldp, &m->mbs, 0); assert(start <= stop); } /* fill in the details if requested */ if (nmatch > 0) { pmatch[0].rm_so = m->coldp - m->offp; pmatch[0].rm_eo = endp - m->offp; } if (nmatch > 1) { assert(m->pmatch != NULL); for (i = 1; i < nmatch; i++) if (i <= m->g->nsub) pmatch[i] = m->pmatch[i]; else { pmatch[i].rm_so = -1; pmatch[i].rm_eo = -1; } } if (m->pmatch != NULL) free((char *)m->pmatch); if (m->lastpos != NULL) free((char *)m->lastpos); STATETEARDOWN(m); return(0); } /* - dissect - figure out what matched what, no back references == static const char *dissect(struct match *m, const char *start, \ == const char *stop, sopno startst, sopno stopst); */ static const char * /* == stop (success) always */ dissect(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst) { int i; sopno ss; /* start sop of current subRE */ sopno es; /* end sop of current subRE */ const char *sp; /* start of string matched by it */ const char *stp; /* string matched by it cannot pass here */ const char *rest; /* start of rest of string */ const char *tail; /* string unmatched by rest of RE */ sopno ssub; /* start sop of subsubRE */ sopno esub; /* end sop of subsubRE */ const char *ssp; /* start of string matched by subsubRE */ const char *sep; /* end of string matched by subsubRE */ const char *oldssp; /* previous ssp */ const char *dp; AT("diss", start, stop, startst, stopst); sp = start; for (ss = startst; ss < stopst; ss = es) { /* identify end of subRE */ es = ss; switch (OP(m->g->strip[es])) { case OPLUS_: case OQUEST_: es += OPND(m->g->strip[es]); break; case OCH_: while (OP(m->g->strip[es]) != O_CH) es += OPND(m->g->strip[es]); break; } es++; /* figure out what it matched */ switch (OP(m->g->strip[ss])) { case OEND: assert(nope); break; case OCHAR: sp += XMBRTOWC(NULL, sp, stop - start, &m->mbs, 0); break; case OBOL: case OEOL: case OBOW: case OEOW: break; case OANY: case OANYOF: sp += XMBRTOWC(NULL, sp, stop - start, &m->mbs, 0); break; case OBACK_: case O_BACK: assert(nope); break; /* cases where length of match is hard to find */ case OQUEST_: stp = stop; for (;;) { /* how long could this one be? */ rest = slow(m, sp, stp, ss, es); assert(rest != NULL); /* it did match */ /* could the rest match the rest? */ tail = slow(m, rest, stop, es, stopst); if (tail == stop) break; /* yes! */ /* no -- try a shorter match for this one */ stp = rest - 1; assert(stp >= sp); /* it did work */ } ssub = ss + 1; esub = es - 1; /* did innards match? */ if (slow(m, sp, rest, ssub, esub) != NULL) { dp = dissect(m, sp, rest, ssub, esub); assert(dp == rest); } else /* no */ assert(sp == rest); sp = rest; break; case OPLUS_: stp = stop; for (;;) { /* how long could this one be? */ rest = slow(m, sp, stp, ss, es); assert(rest != NULL); /* it did match */ /* could the rest match the rest? */ tail = slow(m, rest, stop, es, stopst); if (tail == stop) break; /* yes! */ /* no -- try a shorter match for this one */ stp = rest - 1; assert(stp >= sp); /* it did work */ } ssub = ss + 1; esub = es - 1; ssp = sp; oldssp = ssp; for (;;) { /* find last match of innards */ sep = slow(m, ssp, rest, ssub, esub); if (sep == NULL || sep == ssp) break; /* failed or matched null */ oldssp = ssp; /* on to next try */ ssp = sep; } if (sep == NULL) { /* last successful match */ sep = ssp; ssp = oldssp; } assert(sep == rest); /* must exhaust substring */ assert(slow(m, ssp, sep, ssub, esub) == rest); dp = dissect(m, ssp, sep, ssub, esub); assert(dp == sep); sp = rest; break; case OCH_: stp = stop; for (;;) { /* how long could this one be? */ rest = slow(m, sp, stp, ss, es); assert(rest != NULL); /* it did match */ /* could the rest match the rest? */ tail = slow(m, rest, stop, es, stopst); if (tail == stop) break; /* yes! */ /* no -- try a shorter match for this one */ stp = rest - 1; assert(stp >= sp); /* it did work */ } ssub = ss + 1; esub = ss + OPND(m->g->strip[ss]) - 1; assert(OP(m->g->strip[esub]) == OOR1); for (;;) { /* find first matching branch */ if (slow(m, sp, rest, ssub, esub) == rest) break; /* it matched all of it */ /* that one missed, try next one */ assert(OP(m->g->strip[esub]) == OOR1); esub++; assert(OP(m->g->strip[esub]) == OOR2); ssub = esub + 1; esub += OPND(m->g->strip[esub]); if (OP(m->g->strip[esub]) == OOR2) esub--; else assert(OP(m->g->strip[esub]) == O_CH); } dp = dissect(m, sp, rest, ssub, esub); assert(dp == rest); sp = rest; break; case O_PLUS: case O_QUEST: case OOR1: case OOR2: case O_CH: assert(nope); break; case OLPAREN: i = OPND(m->g->strip[ss]); assert(0 < i && i <= m->g->nsub); m->pmatch[i].rm_so = sp - m->offp; break; case ORPAREN: i = OPND(m->g->strip[ss]); assert(0 < i && i <= m->g->nsub); m->pmatch[i].rm_eo = sp - m->offp; break; default: /* uh oh */ assert(nope); break; } } assert(sp == stop); return(sp); } /* - backref - figure out what matched what, figuring in back references == static const char *backref(struct match *m, const char *start, \ == const char *stop, sopno startst, sopno stopst, sopno lev); */ static const char * /* == stop (success) or NULL (failure) */ backref(struct match *m, const char *start, const char *stop, sopno startst, sopno stopst, sopno lev, /* PLUS nesting level */ int rec) { int i; sopno ss; /* start sop of current subRE */ const char *sp; /* start of string matched by it */ sopno ssub; /* start sop of subsubRE */ sopno esub; /* end sop of subsubRE */ const char *ssp; /* start of string matched by subsubRE */ const char *dp; size_t len; int hard; sop s; regoff_t offsave; cset *cs; wint_t wc; AT("back", start, stop, startst, stopst); sp = start; /* get as far as we can with easy stuff */ hard = 0; for (ss = startst; !hard && ss < stopst; ss++) switch (OP(s = m->g->strip[ss])) { case OCHAR: if (sp == stop) return(NULL); sp += XMBRTOWC(&wc, sp, stop - sp, &m->mbs, BADCHAR); if (wc != OPND(s)) return(NULL); break; case OANY: if (sp == stop) return(NULL); sp += XMBRTOWC(&wc, sp, stop - sp, &m->mbs, BADCHAR); if (wc == BADCHAR) return (NULL); break; case OANYOF: if (sp == stop) return (NULL); cs = &m->g->sets[OPND(s)]; sp += XMBRTOWC(&wc, sp, stop - sp, &m->mbs, BADCHAR); if (wc == BADCHAR || !CHIN(cs, wc)) return(NULL); break; case OBOL: if ((sp == m->beginp && !(m->eflags®_NOTBOL)) || (sp > m->offp && sp < m->endp && *(sp-1) == '\n' && (m->g->cflags®_NEWLINE))) { /* yes */ } else return(NULL); break; case OEOL: if ( (sp == m->endp && !(m->eflags®_NOTEOL)) || (sp < m->endp && *sp == '\n' && (m->g->cflags®_NEWLINE)) ) { /* yes */ } else return(NULL); break; case OBOW: if (sp < m->endp && ISWORD(*sp) && ((sp == m->beginp && !(m->eflags®_NOTBOL)) || (sp > m->offp && !ISWORD(*(sp-1))))) { /* yes */ } else return(NULL); break; case OEOW: if (( (sp == m->endp && !(m->eflags®_NOTEOL)) || (sp < m->endp && *sp == '\n' && (m->g->cflags®_NEWLINE)) || (sp < m->endp && !ISWORD(*sp)) ) && (sp > m->beginp && ISWORD(*(sp-1))) ) { /* yes */ } else return(NULL); break; case O_QUEST: break; case OOR1: /* matches null but needs to skip */ ss++; s = m->g->strip[ss]; do { assert(OP(s) == OOR2); ss += OPND(s); } while (OP(s = m->g->strip[ss]) != O_CH); /* note that the ss++ gets us past the O_CH */ break; default: /* have to make a choice */ hard = 1; break; } if (!hard) { /* that was it! */ if (sp != stop) return(NULL); return(sp); } ss--; /* adjust for the for's final increment */ /* the hard stuff */ AT("hard", sp, stop, ss, stopst); s = m->g->strip[ss]; switch (OP(s)) { case OBACK_: /* the vilest depths */ i = OPND(s); assert(0 < i && i <= m->g->nsub); if (m->pmatch[i].rm_eo == -1) return(NULL); assert(m->pmatch[i].rm_so != -1); len = m->pmatch[i].rm_eo - m->pmatch[i].rm_so; if (len == 0 && rec++ > MAX_RECURSION) return(NULL); assert(stop - m->beginp >= len); if (sp > stop - len) return(NULL); /* not enough left to match */ ssp = m->offp + m->pmatch[i].rm_so; if (memcmp(sp, ssp, len) != 0) return(NULL); while (m->g->strip[ss] != SOP(O_BACK, i)) ss++; return(backref(m, sp+len, stop, ss+1, stopst, lev, rec)); case OQUEST_: /* to null or not */ dp = backref(m, sp, stop, ss+1, stopst, lev, rec); if (dp != NULL) return(dp); /* not */ return(backref(m, sp, stop, ss+OPND(s)+1, stopst, lev, rec)); case OPLUS_: assert(m->lastpos != NULL); assert(lev+1 <= m->g->nplus); m->lastpos[lev+1] = sp; return(backref(m, sp, stop, ss+1, stopst, lev+1, rec)); case O_PLUS: if (sp == m->lastpos[lev]) /* last pass matched null */ return(backref(m, sp, stop, ss+1, stopst, lev-1, rec)); /* try another pass */ m->lastpos[lev] = sp; dp = backref(m, sp, stop, ss-OPND(s)+1, stopst, lev, rec); if (dp == NULL) return(backref(m, sp, stop, ss+1, stopst, lev-1, rec)); else return(dp); case OCH_: /* find the right one, if any */ ssub = ss + 1; esub = ss + OPND(s) - 1; assert(OP(m->g->strip[esub]) == OOR1); for (;;) { /* find first matching branch */ dp = backref(m, sp, stop, ssub, esub, lev, rec); if (dp != NULL) return(dp); /* that one missed, try next one */ if (OP(m->g->strip[esub]) == O_CH) return(NULL); /* there is none */ esub++; assert(OP(m->g->strip[esub]) == OOR2); ssub = esub + 1; esub += OPND(m->g->strip[esub]); if (OP(m->g->strip[esub]) == OOR2) esub--; else assert(OP(m->g->strip[esub]) == O_CH); } /* NOTREACHED */ break; case OLPAREN: /* must undo assignment if rest fails */ i = OPND(s); assert(0 < i && i <= m->g->nsub); offsave = m->pmatch[i].rm_so; m->pmatch[i].rm_so = sp - m->offp; dp = backref(m, sp, stop, ss+1, stopst, lev, rec); if (dp != NULL) return(dp); m->pmatch[i].rm_so = offsave; return(NULL); case ORPAREN: /* must undo assignment if rest fails */ i = OPND(s); assert(0 < i && i <= m->g->nsub); offsave = m->pmatch[i].rm_eo; m->pmatch[i].rm_eo = sp - m->offp; dp = backref(m, sp, stop, ss+1, stopst, lev, rec); if (dp != NULL) return(dp); m->pmatch[i].rm_eo = offsave; return(NULL); default: /* uh oh */ assert(nope); break; } /* "can't happen" */ assert(nope); /* NOTREACHED */ return "shut up gcc"; } /* - fast - step through the string at top speed == static const char *fast(struct match *m, const char *start, \ == const char *stop, sopno startst, sopno stopst); */ static const char * /* where tentative match ended, or NULL */ fast( struct match *m, const char *start, const char *stop, sopno startst, sopno stopst) { states st = m->st; states fresh = m->fresh; states tmp = m->tmp; const char *p = start; wint_t c; wint_t lastc; /* previous c */ wint_t flagch; int i; const char *coldp; /* last p after which no match was underway */ size_t clen; CLEAR(st); SET1(st, startst); SP("fast", st, *p); st = step(m->g, startst, stopst, st, NOTHING, st); ASSIGN(fresh, st); SP("start", st, *p); coldp = NULL; - if (start == m->beginp) + if (start == m->offp || (start == m->beginp && !(m->eflags®_NOTBOL))) c = OUT; else { /* * XXX Wrong if the previous character was multi-byte. * Newline never is (in encodings supported by FreeBSD), * so this only breaks the ISWORD tests below. */ c = (uch)*(start - 1); } for (;;) { /* next character */ lastc = c; if (p == m->endp) { clen = 0; c = OUT; } else clen = XMBRTOWC(&c, p, m->endp - p, &m->mbs, BADCHAR); if (EQ(st, fresh)) coldp = p; /* is there an EOL and/or BOL between lastc and c? */ flagch = '\0'; i = 0; if ( (lastc == '\n' && m->g->cflags®_NEWLINE) || (lastc == OUT && !(m->eflags®_NOTBOL)) ) { flagch = BOL; i = m->g->nbol; } if ( (c == '\n' && m->g->cflags®_NEWLINE) || (c == OUT && !(m->eflags®_NOTEOL)) ) { flagch = (flagch == BOL) ? BOLEOL : EOL; i += m->g->neol; } if (i != 0) { for (; i > 0; i--) st = step(m->g, startst, stopst, st, flagch, st); SP("boleol", st, c); } /* how about a word boundary? */ if ( (flagch == BOL || (lastc != OUT && !ISWORD(lastc))) && (c != OUT && ISWORD(c)) ) { flagch = BOW; } if ( (lastc != OUT && ISWORD(lastc)) && (flagch == EOL || (c != OUT && !ISWORD(c))) ) { flagch = EOW; } if (flagch == BOW || flagch == EOW) { st = step(m->g, startst, stopst, st, flagch, st); SP("boweow", st, c); } /* are we done? */ if (ISSET(st, stopst) || p == stop || clen > stop - p) break; /* NOTE BREAK OUT */ /* no, we must deal with this character */ ASSIGN(tmp, st); ASSIGN(st, fresh); assert(c != OUT); st = step(m->g, startst, stopst, tmp, c, st); SP("aft", st, c); assert(EQ(step(m->g, startst, stopst, st, NOTHING, st), st)); p += clen; } assert(coldp != NULL); m->coldp = coldp; if (ISSET(st, stopst)) return(p+XMBRTOWC(NULL, p, stop - p, &m->mbs, 0)); else return(NULL); } /* - slow - step through the string more deliberately == static const char *slow(struct match *m, const char *start, \ == const char *stop, sopno startst, sopno stopst); */ static const char * /* where it ended */ slow( struct match *m, const char *start, const char *stop, sopno startst, sopno stopst) { states st = m->st; states empty = m->empty; states tmp = m->tmp; const char *p = start; wint_t c; wint_t lastc; /* previous c */ wint_t flagch; int i; const char *matchp; /* last p at which a match ended */ size_t clen; AT("slow", start, stop, startst, stopst); CLEAR(st); SET1(st, startst); SP("sstart", st, *p); st = step(m->g, startst, stopst, st, NOTHING, st); matchp = NULL; - if (start == m->beginp) + if (start == m->offp || (start == m->beginp && !(m->eflags®_NOTBOL))) c = OUT; else { /* * XXX Wrong if the previous character was multi-byte. * Newline never is (in encodings supported by FreeBSD), * so this only breaks the ISWORD tests below. */ c = (uch)*(start - 1); } for (;;) { /* next character */ lastc = c; if (p == m->endp) { c = OUT; clen = 0; } else clen = XMBRTOWC(&c, p, m->endp - p, &m->mbs, BADCHAR); /* is there an EOL and/or BOL between lastc and c? */ flagch = '\0'; i = 0; if ( (lastc == '\n' && m->g->cflags®_NEWLINE) || (lastc == OUT && !(m->eflags®_NOTBOL)) ) { flagch = BOL; i = m->g->nbol; } if ( (c == '\n' && m->g->cflags®_NEWLINE) || (c == OUT && !(m->eflags®_NOTEOL)) ) { flagch = (flagch == BOL) ? BOLEOL : EOL; i += m->g->neol; } if (i != 0) { for (; i > 0; i--) st = step(m->g, startst, stopst, st, flagch, st); SP("sboleol", st, c); } /* how about a word boundary? */ if ( (flagch == BOL || (lastc != OUT && !ISWORD(lastc))) && (c != OUT && ISWORD(c)) ) { flagch = BOW; } if ( (lastc != OUT && ISWORD(lastc)) && (flagch == EOL || (c != OUT && !ISWORD(c))) ) { flagch = EOW; } if (flagch == BOW || flagch == EOW) { st = step(m->g, startst, stopst, st, flagch, st); SP("sboweow", st, c); } /* are we done? */ if (ISSET(st, stopst)) matchp = p; if (EQ(st, empty) || p == stop || clen > stop - p) break; /* NOTE BREAK OUT */ /* no, we must deal with this character */ ASSIGN(tmp, st); ASSIGN(st, empty); assert(c != OUT); st = step(m->g, startst, stopst, tmp, c, st); SP("saft", st, c); assert(EQ(step(m->g, startst, stopst, st, NOTHING, st), st)); p += clen; } return(matchp); } /* - step - map set of states reachable before char to set reachable after == static states step(struct re_guts *g, sopno start, sopno stop, \ == states bef, int ch, states aft); == #define BOL (OUT-1) == #define EOL (BOL-1) == #define BOLEOL (BOL-2) == #define NOTHING (BOL-3) == #define BOW (BOL-4) == #define EOW (BOL-5) == #define BADCHAR (BOL-6) == #define NONCHAR(c) ((c) <= OUT) */ static states step(struct re_guts *g, sopno start, /* start state within strip */ sopno stop, /* state after stop state within strip */ states bef, /* states reachable before */ wint_t ch, /* character or NONCHAR code */ states aft) /* states already known reachable after */ { cset *cs; sop s; sopno pc; onestate here; /* note, macros know this name */ sopno look; int i; for (pc = start, INIT(here, pc); pc != stop; pc++, INC(here)) { s = g->strip[pc]; switch (OP(s)) { case OEND: assert(pc == stop-1); break; case OCHAR: /* only characters can match */ assert(!NONCHAR(ch) || ch != OPND(s)); if (ch == OPND(s)) FWD(aft, bef, 1); break; case OBOL: if (ch == BOL || ch == BOLEOL) FWD(aft, bef, 1); break; case OEOL: if (ch == EOL || ch == BOLEOL) FWD(aft, bef, 1); break; case OBOW: if (ch == BOW) FWD(aft, bef, 1); break; case OEOW: if (ch == EOW) FWD(aft, bef, 1); break; case OANY: if (!NONCHAR(ch)) FWD(aft, bef, 1); break; case OANYOF: cs = &g->sets[OPND(s)]; if (!NONCHAR(ch) && CHIN(cs, ch)) FWD(aft, bef, 1); break; case OBACK_: /* ignored here */ case O_BACK: FWD(aft, aft, 1); break; case OPLUS_: /* forward, this is just an empty */ FWD(aft, aft, 1); break; case O_PLUS: /* both forward and back */ FWD(aft, aft, 1); i = ISSETBACK(aft, OPND(s)); BACK(aft, aft, OPND(s)); if (!i && ISSETBACK(aft, OPND(s))) { /* oho, must reconsider loop body */ pc -= OPND(s) + 1; INIT(here, pc); } break; case OQUEST_: /* two branches, both forward */ FWD(aft, aft, 1); FWD(aft, aft, OPND(s)); break; case O_QUEST: /* just an empty */ FWD(aft, aft, 1); break; case OLPAREN: /* not significant here */ case ORPAREN: FWD(aft, aft, 1); break; case OCH_: /* mark the first two branches */ FWD(aft, aft, 1); assert(OP(g->strip[pc+OPND(s)]) == OOR2); FWD(aft, aft, OPND(s)); break; case OOR1: /* done a branch, find the O_CH */ if (ISSTATEIN(aft, here)) { for (look = 1; OP(s = g->strip[pc+look]) != O_CH; look += OPND(s)) assert(OP(s) == OOR2); FWD(aft, aft, look + 1); } break; case OOR2: /* propagate OCH_'s marking */ FWD(aft, aft, 1); if (OP(g->strip[pc+OPND(s)]) != O_CH) { assert(OP(g->strip[pc+OPND(s)]) == OOR2); FWD(aft, aft, OPND(s)); } break; case O_CH: /* just empty */ FWD(aft, aft, 1); break; default: /* ooooops... */ assert(nope); break; } } return(aft); } #ifdef REDEBUG /* - print - print a set of states == #ifdef REDEBUG == static void print(struct match *m, const char *caption, states st, \ == int ch, FILE *d); == #endif */ static void print(struct match *m, const char *caption, states st, int ch, FILE *d) { struct re_guts *g = m->g; sopno i; int first = 1; if (!(m->eflags®_TRACE)) return; fprintf(d, "%s", caption); if (ch != '\0') fprintf(d, " %s", pchar(ch)); for (i = 0; i < g->nstates; i++) if (ISSET(st, i)) { fprintf(d, "%s%d", (first) ? "\t" : ", ", i); first = 0; } fprintf(d, "\n"); } /* - at - print current situation == #ifdef REDEBUG == static void at(struct match *m, const char *title, const char *start, \ == const char *stop, sopno startst, sopno stopst); == #endif */ static void at( struct match *m, const char *title, const char *start, const char *stop, sopno startst, sopno stopst) { if (!(m->eflags®_TRACE)) return; printf("%s %s-", title, pchar(*start)); printf("%s ", pchar(*stop)); printf("%ld-%ld\n", (long)startst, (long)stopst); } #ifndef PCHARDONE #define PCHARDONE /* never again */ /* - pchar - make a character printable == #ifdef REDEBUG == static const char *pchar(int ch); == #endif * * Is this identical to regchar() over in debug.c? Well, yes. But a * duplicate here avoids having a debugging-capable regexec.o tied to * a matching debug.o, and this is convenient. It all disappears in * the non-debug compilation anyway, so it doesn't matter much. */ static const char * /* -> representation */ pchar(int ch) { static char pbuf[10]; if (isprint((uch)ch) || ch == ' ') sprintf(pbuf, "%c", ch); else sprintf(pbuf, "\\%o", ch); return(pbuf); } #endif #endif #undef matcher #undef fast #undef slow #undef dissect #undef backref #undef step #undef print #undef at #undef match Index: stable/9/lib/libc/regex/regex.3 =================================================================== --- stable/9/lib/libc/regex/regex.3 (revision 302168) +++ stable/9/lib/libc/regex/regex.3 (revision 302169) @@ -1,727 +1,753 @@ .\" Copyright (c) 1992, 1993, 1994 Henry Spencer. .\" Copyright (c) 1992, 1993, 1994 .\" The Regents of the University of California. All rights reserved. .\" .\" This code is derived from software contributed to Berkeley by .\" Henry Spencer. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. Redistributions in binary form must reproduce the above copyright .\" notice, this list of conditions and the following disclaimer in the .\" documentation and/or other materials provided with the distribution. .\" 4. Neither the name of the University nor the names of its contributors .\" may be used to endorse or promote products derived from this software .\" without specific prior written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF .\" SUCH DAMAGE. .\" .\" @(#)regex.3 8.4 (Berkeley) 3/20/94 .\" $FreeBSD$ .\" -.Dd August 17, 2005 +.Dd May 25, 2016 .Dt REGEX 3 .Os .Sh NAME .Nm regcomp , .Nm regexec , .Nm regerror , .Nm regfree .Nd regular-expression library .Sh LIBRARY .Lb libc .Sh SYNOPSIS .In regex.h .Ft int .Fo regcomp .Fa "regex_t * restrict preg" "const char * restrict pattern" "int cflags" .Fc .Ft int .Fo regexec .Fa "const regex_t * restrict preg" "const char * restrict string" .Fa "size_t nmatch" "regmatch_t pmatch[restrict]" "int eflags" .Fc .Ft size_t .Fo regerror .Fa "int errcode" "const regex_t * restrict preg" .Fa "char * restrict errbuf" "size_t errbuf_size" .Fc .Ft void .Fn regfree "regex_t *preg" .Sh DESCRIPTION These routines implement .St -p1003.2 regular expressions .Pq Do RE Dc Ns s ; see .Xr re_format 7 . The .Fn regcomp function compiles an RE written as a string into an internal form, .Fn regexec matches that internal form against a string and reports results, .Fn regerror transforms error codes from either into human-readable messages, and .Fn regfree frees any dynamically-allocated storage used by the internal form of an RE. .Pp The header .In regex.h declares two structure types, .Ft regex_t and .Ft regmatch_t , the former for compiled internal forms and the latter for match reporting. It also declares the four functions, a type .Ft regoff_t , and a number of constants with names starting with .Dq Dv REG_ . .Pp The .Fn regcomp function compiles the regular expression contained in the .Fa pattern string, subject to the flags in .Fa cflags , and places the results in the .Ft regex_t structure pointed to by .Fa preg . The .Fa cflags argument is the bitwise OR of zero or more of the following flags: .Bl -tag -width REG_EXTENDED .It Dv REG_EXTENDED Compile modern .Pq Dq extended REs, rather than the obsolete .Pq Dq basic REs that are the default. .It Dv REG_BASIC This is a synonym for 0, provided as a counterpart to .Dv REG_EXTENDED to improve readability. .It Dv REG_NOSPEC Compile with recognition of all special characters turned off. All characters are thus considered ordinary, so the .Dq RE is a literal string. This is an extension, compatible with but not specified by .St -p1003.2 , and should be used with caution in software intended to be portable to other systems. .Dv REG_EXTENDED and .Dv REG_NOSPEC may not be used in the same call to .Fn regcomp . .It Dv REG_ICASE Compile for matching that ignores upper/lower case distinctions. See .Xr re_format 7 . .It Dv REG_NOSUB Compile for matching that need only report success or failure, not what was matched. .It Dv REG_NEWLINE Compile for newline-sensitive matching. By default, newline is a completely ordinary character with no special meaning in either REs or strings. With this flag, .Ql [^ bracket expressions and .Ql .\& never match newline, a .Ql ^\& anchor matches the null string after any newline in the string in addition to its normal function, and the .Ql $\& anchor matches the null string before any newline in the string in addition to its normal function. .It Dv REG_PEND The regular expression ends, not at the first NUL, but just before the character pointed to by the .Va re_endp member of the structure pointed to by .Fa preg . The .Va re_endp member is of type .Ft "const char *" . This flag permits inclusion of NULs in the RE; they are considered ordinary characters. This is an extension, compatible with but not specified by .St -p1003.2 , and should be used with caution in software intended to be portable to other systems. .El .Pp When successful, .Fn regcomp returns 0 and fills in the structure pointed to by .Fa preg . One member of that structure (other than .Va re_endp ) is publicized: .Va re_nsub , of type .Ft size_t , contains the number of parenthesized subexpressions within the RE (except that the value of this member is undefined if the .Dv REG_NOSUB flag was used). If .Fn regcomp fails, it returns a non-zero error code; see .Sx DIAGNOSTICS . .Pp The .Fn regexec function matches the compiled RE pointed to by .Fa preg against the .Fa string , subject to the flags in .Fa eflags , and reports results using .Fa nmatch , .Fa pmatch , and the returned value. The RE must have been compiled by a previous invocation of .Fn regcomp . The compiled form is not altered during execution of .Fn regexec , so a single compiled RE can be used simultaneously by multiple threads. .Pp By default, the NUL-terminated string pointed to by .Fa string is considered to be the text of an entire line, minus any terminating newline. The .Fa eflags argument is the bitwise OR of zero or more of the following flags: .Bl -tag -width REG_STARTEND .It Dv REG_NOTBOL -The first character of -the string -is not the beginning of a line, so the -.Ql ^\& -anchor should not match before it. +The first character of the string is treated as the continuation +of a line. +This means that the anchors +.Ql ^\& , +.Ql [[:<:]] , +and +.Ql \e< +do not match before it; but see +.Dv REG_STARTEND +below. This does not affect the behavior of newlines under .Dv REG_NEWLINE . .It Dv REG_NOTEOL The NUL terminating the string does not end a line, so the .Ql $\& -anchor should not match before it. +anchor does not match before it. This does not affect the behavior of newlines under .Dv REG_NEWLINE . .It Dv REG_STARTEND The string is considered to start at -.Fa string -+ -.Fa pmatch Ns [0]. Ns Va rm_so -and to have a terminating NUL located at -.Fa string -+ -.Fa pmatch Ns [0]. Ns Va rm_eo -(there need not actually be a NUL at that location), +.Fa string No + +.Fa pmatch Ns [0]. Ns Fa rm_so +and to end before the byte located at +.Fa string No + +.Fa pmatch Ns [0]. Ns Fa rm_eo , regardless of the value of .Fa nmatch . See below for the definition of .Fa pmatch and .Fa nmatch . This is an extension, compatible with but not specified by .St -p1003.2 , and should be used with caution in software intended to be portable to other systems. -Note that a non-zero -.Va rm_so -does not imply -.Dv REG_NOTBOL ; -.Dv REG_STARTEND -affects only the location of the string, -not how it is matched. +.Pp +Without +.Dv REG_NOTBOL , +the position +.Fa rm_so +is considered the beginning of a line, such that +.Ql ^ +matches before it, and the beginning of a word if there is a word +character at this position, such that +.Ql [[:<:]] +and +.Ql \e< +match before it. +.Pp +With +.Dv REG_NOTBOL , +the character at position +.Fa rm_so +is treated as the continuation of a line, and if +.Fa rm_so +is greater than 0, the preceding character is taken into consideration. +If the preceding character is a newline and the regular expression was compiled +with +.Dv REG_NEWLINE , +.Ql ^ +matches before the string; if the preceding character is not a word character +but the string starts with a word character, +.Ql [[:<:]] +and +.Ql \e< +match before the string. .El .Pp See .Xr re_format 7 for a discussion of what is matched in situations where an RE or a portion thereof could match any of several substrings of .Fa string . .Pp Normally, .Fn regexec returns 0 for success and the non-zero code .Dv REG_NOMATCH for failure. Other non-zero error codes may be returned in exceptional situations; see .Sx DIAGNOSTICS . .Pp If .Dv REG_NOSUB was specified in the compilation of the RE, or if .Fa nmatch is 0, .Fn regexec ignores the .Fa pmatch argument (but see below for the case where .Dv REG_STARTEND is specified). Otherwise, .Fa pmatch points to an array of .Fa nmatch structures of type .Ft regmatch_t . Such a structure has at least the members .Va rm_so and .Va rm_eo , both of type .Ft regoff_t (a signed arithmetic type at least as large as an .Ft off_t and a .Ft ssize_t ) , containing respectively the offset of the first character of a substring and the offset of the first character after the end of the substring. Offsets are measured from the beginning of the .Fa string argument given to .Fn regexec . An empty substring is denoted by equal offsets, both indicating the character following the empty substring. .Pp The 0th member of the .Fa pmatch array is filled in to indicate what substring of .Fa string was matched by the entire RE. Remaining members report what substring was matched by parenthesized subexpressions within the RE; member .Va i reports subexpression .Va i , with subexpressions counted (starting at 1) by the order of their opening parentheses in the RE, left to right. Unused entries in the array (corresponding either to subexpressions that did not participate in the match at all, or to subexpressions that do not exist in the RE (that is, .Va i > .Fa preg Ns -> Ns Va re_nsub ) ) have both .Va rm_so and .Va rm_eo set to -1. If a subexpression participated in the match several times, the reported substring is the last one it matched. (Note, as an example in particular, that when the RE .Ql "(b*)+" matches .Ql bbb , the parenthesized subexpression matches each of the three .So Li b Sc Ns s and then an infinite number of empty strings following the last .Ql b , so the reported substring is one of the empties.) .Pp If .Dv REG_STARTEND is specified, .Fa pmatch must point to at least one .Ft regmatch_t (even if .Fa nmatch is 0 or .Dv REG_NOSUB was specified), to hold the input offsets for .Dv REG_STARTEND . Use for output is still entirely controlled by .Fa nmatch ; if .Fa nmatch is 0 or .Dv REG_NOSUB was specified, the value of .Fa pmatch Ns [0] will not be changed by a successful .Fn regexec . .Pp The .Fn regerror function maps a non-zero .Fa errcode from either .Fn regcomp or .Fn regexec to a human-readable, printable message. If .Fa preg is .No non\- Ns Dv NULL , the error code should have arisen from use of the .Ft regex_t pointed to by .Fa preg , and if the error code came from .Fn regcomp , it should have been the result from the most recent .Fn regcomp using that .Ft regex_t . The .Fn ( regerror may be able to supply a more detailed message using information from the .Ft regex_t . ) The .Fn regerror function places the NUL-terminated message into the buffer pointed to by .Fa errbuf , limiting the length (including the NUL) to at most .Fa errbuf_size bytes. If the whole message will not fit, as much of it as will fit before the terminating NUL is supplied. In any case, the returned value is the size of buffer needed to hold the whole message (including terminating NUL). If .Fa errbuf_size is 0, .Fa errbuf is ignored but the return value is still correct. .Pp If the .Fa errcode given to .Fn regerror is first ORed with .Dv REG_ITOA , the .Dq message that results is the printable name of the error code, e.g.\& .Dq Dv REG_NOMATCH , rather than an explanation thereof. If .Fa errcode is .Dv REG_ATOI , then .Fa preg shall be .No non\- Ns Dv NULL and the .Va re_endp member of the structure it points to must point to the printable name of an error code; in this case, the result in .Fa errbuf is the decimal digits of the numeric value of the error code (0 if the name is not recognized). .Dv REG_ITOA and .Dv REG_ATOI are intended primarily as debugging facilities; they are extensions, compatible with but not specified by .St -p1003.2 , and should be used with caution in software intended to be portable to other systems. Be warned also that they are considered experimental and changes are possible. .Pp The .Fn regfree function frees any dynamically-allocated storage associated with the compiled RE pointed to by .Fa preg . The remaining .Ft regex_t is no longer a valid compiled RE and the effect of supplying it to .Fn regexec or .Fn regerror is undefined. .Pp None of these functions references global variables except for tables of constants; all are safe for use from multiple threads if the arguments are safe. .Sh IMPLEMENTATION CHOICES There are a number of decisions that .St -p1003.2 leaves up to the implementor, either by explicitly saying .Dq undefined or by virtue of them being forbidden by the RE grammar. This implementation treats them as follows. .Pp See .Xr re_format 7 for a discussion of the definition of case-independent matching. .Pp There is no particular limit on the length of REs, except insofar as memory is limited. Memory usage is approximately linear in RE size, and largely insensitive to RE complexity, except for bounded repetitions. See .Sx BUGS for one short RE using them that will run almost any system out of memory. .Pp A backslashed character other than one specifically given a magic meaning by .St -p1003.2 (such magic meanings occur only in obsolete .Bq Dq basic REs) is taken as an ordinary character. .Pp Any unmatched .Ql [\& is a .Dv REG_EBRACK error. .Pp Equivalence classes cannot begin or end bracket-expression ranges. The endpoint of one range cannot begin another. .Pp .Dv RE_DUP_MAX , the limit on repetition counts in bounded repetitions, is 255. .Pp A repetition operator .Ql ( ?\& , .Ql *\& , .Ql +\& , or bounds) cannot follow another repetition operator. A repetition operator cannot begin an expression or subexpression or follow .Ql ^\& or .Ql |\& . .Pp .Ql |\& cannot appear first or last in a (sub)expression or after another .Ql |\& , i.e., an operand of .Ql |\& cannot be an empty subexpression. An empty parenthesized subexpression, .Ql "()" , is legal and matches an empty (sub)string. An empty string is not a legal RE. .Pp A .Ql {\& followed by a digit is considered the beginning of bounds for a bounded repetition, which must then follow the syntax for bounds. A .Ql {\& .Em not followed by a digit is considered an ordinary character. .Pp .Ql ^\& and .Ql $\& beginning and ending subexpressions in obsolete .Pq Dq basic REs are anchors, not ordinary characters. .Sh DIAGNOSTICS Non-zero error codes from .Fn regcomp and .Fn regexec include the following: .Pp .Bl -tag -width REG_ECOLLATE -compact .It Dv REG_NOMATCH The .Fn regexec function failed to match .It Dv REG_BADPAT invalid regular expression .It Dv REG_ECOLLATE invalid collating element .It Dv REG_ECTYPE invalid character class .It Dv REG_EESCAPE .Ql \e applied to unescapable character .It Dv REG_ESUBREG invalid backreference number .It Dv REG_EBRACK brackets .Ql "[ ]" not balanced .It Dv REG_EPAREN parentheses .Ql "( )" not balanced .It Dv REG_EBRACE braces .Ql "{ }" not balanced .It Dv REG_BADBR invalid repetition count(s) in .Ql "{ }" .It Dv REG_ERANGE invalid character range in .Ql "[ ]" .It Dv REG_ESPACE ran out of memory .It Dv REG_BADRPT .Ql ?\& , .Ql *\& , or .Ql +\& operand invalid .It Dv REG_EMPTY empty (sub)expression .It Dv REG_ASSERT cannot happen - you found a bug .It Dv REG_INVARG invalid argument, e.g.\& negative-length string .It Dv REG_ILLSEQ illegal byte sequence (bad multibyte character) .El .Sh SEE ALSO .Xr grep 1 , .Xr re_format 7 .Pp .St -p1003.2 , sections 2.8 (Regular Expression Notation) and B.5 (C Binding for Regular Expression Matching). .Sh HISTORY Originally written by .An Henry Spencer . Altered for inclusion in the .Bx 4.4 distribution. .Sh BUGS This is an alpha release with known defects. Please report problems. .Pp The back-reference code is subtle and doubts linger about its correctness in complex cases. .Pp The .Fn regexec function performance is poor. This will improve with later releases. The .Fa nmatch argument exceeding 0 is expensive; .Fa nmatch exceeding 1 is worse. The .Fn regexec function is largely insensitive to RE complexity .Em except that back references are massively expensive. RE length does matter; in particular, there is a strong speed bonus for keeping RE length under about 30 characters, with most special characters counting roughly double. .Pp The .Fn regcomp function implements bounded repetitions by macro expansion, which is costly in time and space if counts are large or bounded repetitions are nested. An RE like, say, .Ql "((((a{1,100}){1,100}){1,100}){1,100}){1,100}" will (eventually) run almost any existing machine out of swap space. .Pp There are suspected problems with response to obscure error conditions. Notably, certain kinds of internal overflow, produced only by truly enormous REs or by multiply nested bounded repetitions, are probably not handled well. .Pp Due to a mistake in .St -p1003.2 , things like .Ql "a)b" are legal REs because .Ql )\& is a special character only in the presence of a previous unmatched .Ql (\& . This cannot be fixed until the spec is fixed. .Pp The standard's definition of back references is vague. For example, does .Ql "a\e(\e(b\e)*\e2\e)*d" match .Ql "abbbd" ? Until the standard is clarified, behavior in such cases should not be relied on. .Pp The implementation of word-boundary matching is a bit of a kludge, and bugs may lurk in combinations of word-boundary matching and anchoring. .Pp Word-boundary matching does not work properly in multibyte locales. Index: stable/9/lib/libc =================================================================== --- stable/9/lib/libc (revision 302168) +++ stable/9/lib/libc (revision 302169) Property changes on: stable/9/lib/libc ___________________________________________________________________ Modified: svn:mergeinfo ## -0,0 +0,1 ## Merged /head/lib/libc:r300683