X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=regexec.c;h=002b66ac85ae93dcd4f6bfffb835261e4dcfe546;hb=02237c86f171377b770f43f3200e5b80153a9722;hp=8f5278c254bde811374fe497597d23b9c7955636;hpb=05b4157f6fee2ece5589511f927d566b229523f9;p=p5sagit%2Fp5-mst-13.2.git diff --git a/regexec.c b/regexec.c index 8f5278c..002b66a 100644 --- a/regexec.c +++ b/regexec.c @@ -221,6 +221,22 @@ typedef struct re_cc_state #define regcpblow(cp) LEAVE_SCOPE(cp) +#define TRYPAREN(paren, n, input) { \ + if (paren) { \ + if (n) { \ + PL_regstartp[paren] = HOPc(input, -1) - PL_bostr; \ + PL_regendp[paren] = input - PL_bostr; \ + } \ + else \ + PL_regendp[paren] = -1; \ + } \ + if (regmatch(next)) \ + sayYES; \ + if (paren && n) \ + PL_regendp[paren] = -1; \ +} + + /* * pregexec and friends */ @@ -599,9 +615,10 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, find_anchor: while (t < strend - prog->minlen) { if (*t == '\n') { - if (t < s - prog->check_offset_min) { + if (t < check_at - prog->check_offset_min) { if (prog->anchored_substr) { - /* We definitely contradict the found anchored + /* Since we moved from the found position, + we definitely contradict the found anchored substr. Due to the above check we do not contradict "check" substr. Thus we can arrive here only if check substr @@ -612,14 +629,19 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, PL_colors[0],PL_colors[1], (long)(strpos - i_strpos), (long)(strpos - i_strpos + prog->anchored_offset))); goto do_other_anchored; } + /* We don't contradict the found floating substring. */ + /* XXXX Why not check for STCLASS? */ s = t + 1; DEBUG_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m at offset %ld...\n", PL_colors[0],PL_colors[1], (long)(s - i_strpos))); goto set_useful; } - DEBUG_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m, restarting at offset %ld...\n", + /* Position contradicts check-string */ + /* XXXX probably better to look for check-string + than for "\n", so one should lower the limit for t? */ + DEBUG_r(PerlIO_printf(Perl_debug_log, "Found /%s^%s/m, restarting lookup for check-string at offset %ld...\n", PL_colors[0],PL_colors[1], (long)(t + 1 - i_strpos))); - strpos = s = t + 1; + other_last = strpos = s = t + 1; goto restart; } t++; @@ -628,15 +650,20 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, PL_colors[0],PL_colors[1])); goto fail_finish; } + else { + DEBUG_r(PerlIO_printf(Perl_debug_log, "Starting position does not contradict /%s^%s/m...\n", + PL_colors[0],PL_colors[1])); + } s = t; set_useful: ++BmUSEFUL(prog->check_substr); /* hooray/5 */ } else { PL_bostr = tmp; - /* The found string does not prohibit matching at beg-of-str + /* The found string does not prohibit matching at strpos, - no optimization of calling REx engine can be performed, - unless it was an MBOL and we are not after MBOL. */ + unless it was an MBOL and we are not after MBOL, + or a future STCLASS check will fail this. */ try_at_start: /* Even in this situation we may use MBOL flag if strpos is offset wrt the start of the string. */ @@ -649,8 +676,8 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, goto find_anchor; } DEBUG_r( if (ml_anch) - PerlIO_printf(Perl_debug_log, "Does not contradict /%s^%s/m...\n", - PL_colors[0],PL_colors[1]); + PerlIO_printf(Perl_debug_log, "Position at offset %ld does not contradict /%s^%s/m...\n", + (long)(strpos - i_strpos), PL_colors[0],PL_colors[1]); ); success_at_start: if (!(prog->reganch & ROPT_NAUGHTY) /* XXXX If strpos moved? */ @@ -659,6 +686,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, && prog->check_substr == prog->float_substr) { /* If flags & SOMETHING - do not do it many times on the same match */ + DEBUG_r(PerlIO_printf(Perl_debug_log, "... Disabling check substring...\n")); SvREFCNT_dec(prog->check_substr); prog->check_substr = Nullsv; /* disable */ prog->float_substr = Nullsv; /* clear */ @@ -725,7 +753,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, goto fail; } DEBUG_r( PerlIO_printf(Perl_debug_log, - "Trying %s substr starting at offset %ld...\n", + "Looking for %s substr starting at offset %ld...\n", what, (long)(s + start_shift - i_strpos)) ); goto restart; } @@ -735,7 +763,7 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, /* Recheck anchored substring, but not floating... */ s = check_at; DEBUG_r( PerlIO_printf(Perl_debug_log, - "Trying anchored substr starting at offset %ld...\n", + "Looking for anchored substr starting at offset %ld...\n", (long)(other_last - i_strpos)) ); goto do_other_anchored; } @@ -744,8 +772,8 @@ Perl_re_intuit_start(pTHX_ regexp *prog, SV *sv, char *strpos, if (ml_anch) { s = t = t + 1; DEBUG_r( PerlIO_printf(Perl_debug_log, - "Trying /^/m starting at offset %ld...\n", - (long)(t - i_strpos)) ); + "Looking for /%s^%s/m starting at offset %ld...\n", + PL_colors[0],PL_colors[1], (long)(t - i_strpos)) ); goto try_at_offset; } if (!prog->float_substr) /* Could have been deleted */ @@ -1434,9 +1462,14 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * /* we have /x+whatever/ */ /* it must be a one character string (XXXX Except UTF?) */ char ch = SvPVX(prog->anchored_substr)[0]; +#ifdef DEBUGGING + int did_match = 0; +#endif + if (UTF) { while (s < strend) { if (*s == ch) { + DEBUG_r( did_match = 1 ); if (regtry(prog, s)) goto got_it; s += UTF8SKIP(s); while (s < strend && *s == ch) @@ -1448,6 +1481,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * else { while (s < strend) { if (*s == ch) { + DEBUG_r( did_match = 1 ); if (regtry(prog, s)) goto got_it; s++; while (s < strend && *s == ch) @@ -1456,6 +1490,9 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * s++; } } + DEBUG_r(did_match || + PerlIO_printf(Perl_debug_log, + "Did not find anchored character...\n")); } /*SUPPRESS 560*/ else if (prog->anchored_substr != Nullsv @@ -1471,6 +1508,9 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * -(I32)(CHR_SVLEN(must) - (SvTAIL(must) != 0) + back_min)); char *last1; /* Last position checked before */ +#ifdef DEBUGGING + int did_match = 0; +#endif if (s > PL_bostr) last1 = HOPc(s, -1); @@ -1489,6 +1529,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * : (s = fbm_instr((unsigned char*)HOP(s, back_min), (unsigned char*)strend, must, PL_multiline ? FBMrf_MULTILINE : 0))) ) { + DEBUG_r( did_match = 1 ); if (HOPc(s, -back_max) > last1) { last1 = HOPc(s, -back_min); s = HOPc(s, -back_max); @@ -1514,6 +1555,14 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * } } } + DEBUG_r(did_match || + PerlIO_printf(Perl_debug_log, "Did not find %s substr `%s%.*s%s'%s...\n", + ((must == prog->anchored_substr) + ? "anchored" : "floating"), + PL_colors[0], + (int)(SvCUR(must) - (SvTAIL(must)!=0)), + SvPVX(must), + PL_colors[1], (SvTAIL(must) ? "$" : ""))); goto phooey; } else if ((c = prog->regstclass)) { @@ -1522,6 +1571,7 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * strend = HOPc(strend, -(minlen - 1)); if (find_byclass(prog, c, s, strend, startpos, 0)) goto got_it; + DEBUG_r(PerlIO_printf(Perl_debug_log, "Contradicts stclass...\n")); } else { dontbother = 0; @@ -1554,7 +1604,12 @@ Perl_regexec_flags(pTHX_ register regexp *prog, char *stringarg, register char * last = strend; /* matching `$' */ } } - if (last == NULL) goto phooey; /* Should not happen! */ + if (last == NULL) { + DEBUG_r(PerlIO_printf(Perl_debug_log, + "%sCan't trim the tail, match fails (should not happen)%s\n", + PL_colors[4],PL_colors[5])); + goto phooey; /* Should not happen! */ + } dontbother = strend - last + prog->float_min_offset; } if (minlen && (dontbother < minlen)) @@ -1616,6 +1671,8 @@ got_it: return 1; phooey: + DEBUG_r(PerlIO_printf(Perl_debug_log, "%sMatch failed%s\n", + PL_colors[4],PL_colors[5])); if (PL_reg_eval_set) restore_pos(aTHXo_ 0); return 0; @@ -1840,7 +1897,7 @@ S_regmatch(pTHX_ regnode *prog) } sayNO; case SBOL: - if (locinput == PL_regbol && PL_regprev == '\n') + if (locinput == PL_bostr) break; sayNO; case GPOS: @@ -2994,16 +3051,7 @@ S_regmatch(pTHX_ regnode *prog) sayNO; } /* PL_reginput == locinput now */ - if (paren) { - if (ln) { - PL_regstartp[paren] = HOPc(locinput, -1) - PL_bostr; - PL_regendp[paren] = locinput - PL_bostr; - } - else - PL_regendp[paren] = -1; - } - if (regmatch(next)) - sayYES; + TRYPAREN(paren, ln, locinput); PL_reginput = locinput; /* Could be reset... */ REGCP_UNWIND; /* Couldn't or didn't -- move forward. */ @@ -3017,16 +3065,7 @@ S_regmatch(pTHX_ regnode *prog) UCHARAT(PL_reginput) == c1 || UCHARAT(PL_reginput) == c2) { - if (paren) { - if (n) { - PL_regstartp[paren] = HOPc(PL_reginput, -1) - PL_bostr; - PL_regendp[paren] = PL_reginput - PL_bostr; - } - else - PL_regendp[paren] = -1; - } - if (regmatch(next)) - sayYES; + TRYPAREN(paren, n, PL_reginput); REGCP_UNWIND; } /* Couldn't or didn't -- move forward. */ @@ -3060,16 +3099,7 @@ S_regmatch(pTHX_ regnode *prog) UCHARAT(PL_reginput) == c1 || UCHARAT(PL_reginput) == c2) { - if (paren && n) { - if (n) { - PL_regstartp[paren] = HOPc(PL_reginput, -1) - PL_bostr; - PL_regendp[paren] = PL_reginput - PL_bostr; - } - else - PL_regendp[paren] = -1; - } - if (regmatch(next)) - sayYES; + TRYPAREN(paren, n, PL_reginput); REGCP_UNWIND; } /* Couldn't or didn't -- back up. */ @@ -3084,8 +3114,7 @@ S_regmatch(pTHX_ regnode *prog) UCHARAT(PL_reginput) == c1 || UCHARAT(PL_reginput) == c2) { - if (regmatch(next)) - sayYES; + TRYPAREN(paren, n, PL_reginput); REGCP_UNWIND; } /* Couldn't or didn't -- back up. */