diff options
| -rw-r--r-- | pod/perldebguts.pod | 1 | ||||
| -rw-r--r-- | regcomp.c | 36 | ||||
| -rw-r--r-- | regcomp.h | 20 | ||||
| -rw-r--r-- | regcomp.sym | 1 | ||||
| -rw-r--r-- | regexec.c | 3 | ||||
| -rw-r--r-- | regnodes.h | 305 |
6 files changed, 185 insertions, 181 deletions
diff --git a/pod/perldebguts.pod b/pod/perldebguts.pod index 54fc4789f2..9765afbd7d 100644 --- a/pod/perldebguts.pod +++ b/pod/perldebguts.pod @@ -591,7 +591,6 @@ will be lost. CANY no Match any one byte. ANYOF sv Match character in (or not in) this class, single char match only - ANYOF_SYNTHETIC sv Synthetic start class POSIXD none Some [[:class:]] under /d; the FLAGS field gives which one @@ -863,7 +863,7 @@ S_ssc_anything(pTHX_ regnode_ssc *ssc) PERL_ARGS_ASSERT_SSC_ANYTHING; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); ssc->invlist = sv_2mortal(_new_invlist(2)); /* mortalize so won't leak */ _append_range_to_invlist(ssc->invlist, 0, UV_MAX); @@ -883,7 +883,7 @@ S_ssc_is_anything(pTHX_ const regnode_ssc *ssc) PERL_ARGS_ASSERT_SSC_IS_ANYTHING; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); if (! (ANYOF_FLAGS(ssc) & ANYOF_EMPTY_STRING)) { return FALSE; @@ -923,7 +923,7 @@ S_ssc_init(pTHX_ const RExC_state_t *pRExC_state, regnode_ssc *ssc) PERL_ARGS_ASSERT_SSC_INIT; Zero(ssc, 1, regnode_ssc); - OP(ssc) = ANYOF_SYNTHETIC; + set_ANYOF_SYNTHETIC(ssc); ARG_SET(ssc, ANYOF_NONBITMAP_EMPTY); ssc_anything(ssc); @@ -959,7 +959,7 @@ S_ssc_is_cp_posixl_init(pTHX_ const RExC_state_t *pRExC_state, PERL_ARGS_ASSERT_SSC_IS_CP_POSIXL_INIT; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); invlist_iterinit(ssc->invlist); ret = invlist_iternext(ssc->invlist, &start, &end) @@ -1105,11 +1105,11 @@ S_ssc_and(pTHX_ const RExC_state_t *pRExC_state, regnode_ssc *ssc, PERL_ARGS_ASSERT_SSC_AND; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); /* 'and_with' is used as-is if it too is an SSC; otherwise have to extract * the code point inversion list and just the relevant flags */ - if (OP(and_with) == ANYOF_SYNTHETIC) { + if (is_ANYOF_SYNTHETIC(and_with)) { anded_cp_list = and_with->invlist; anded_flags = ANYOF_FLAGS(and_with); @@ -1172,7 +1172,7 @@ S_ssc_and(pTHX_ const RExC_state_t *pRExC_state, regnode_ssc *ssc, * */ if ((ANYOF_FLAGS(and_with) & ANYOF_INVERT) - && OP(and_with) != ANYOF_SYNTHETIC) + && ! is_ANYOF_SYNTHETIC(and_with)) { unsigned int i; @@ -1230,13 +1230,13 @@ S_ssc_and(pTHX_ const RExC_state_t *pRExC_state, regnode_ssc *ssc, } /* else ssc already has no posixes */ } /* else: Not inverted. This routine is a no-op if 'and_with' is an SSC in its initial state */ - else if (OP(and_with) != ANYOF_SYNTHETIC + else if (! is_ANYOF_SYNTHETIC(and_with) || ! ssc_is_cp_posixl_init(pRExC_state, and_with)) { /* But if 'ssc' is in its initial state, the result is just 'and_with'; * copy it over 'ssc' */ if (ssc_is_cp_posixl_init(pRExC_state, ssc)) { - if (OP(and_with) == ANYOF_SYNTHETIC) { + if (is_ANYOF_SYNTHETIC(and_with)) { StructCopy(and_with, ssc, regnode_ssc); } else { @@ -1273,11 +1273,11 @@ S_ssc_or(pTHX_ const RExC_state_t *pRExC_state, regnode_ssc *ssc, PERL_ARGS_ASSERT_SSC_OR; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); /* 'or_with' is used as-is if it too is an SSC; otherwise have to extract * the code point inversion list and just the relevant flags */ - if (OP(or_with) == ANYOF_SYNTHETIC) { + if (is_ANYOF_SYNTHETIC(or_with)) { ored_cp_list = or_with->invlist; ored_flags = ANYOF_FLAGS(or_with); } @@ -1308,7 +1308,7 @@ S_ssc_or(pTHX_ const RExC_state_t *pRExC_state, regnode_ssc *ssc, * */ if ((ANYOF_FLAGS(or_with) & ANYOF_INVERT) - && OP(or_with) != ANYOF_SYNTHETIC) + && ! is_ANYOF_SYNTHETIC(or_with)) { /* We ignore P2, leaving P1 going forward */ } @@ -1341,7 +1341,7 @@ S_ssc_union(pTHX_ regnode_ssc *ssc, SV* const invlist, const bool invert2nd) { PERL_ARGS_ASSERT_SSC_UNION; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); _invlist_union_maybe_complement_2nd(ssc->invlist, invlist, @@ -1356,7 +1356,7 @@ S_ssc_intersection(pTHX_ regnode_ssc *ssc, { PERL_ARGS_ASSERT_SSC_INTERSECTION; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); _invlist_intersection_maybe_complement_2nd(ssc->invlist, invlist, @@ -1369,7 +1369,7 @@ S_ssc_add_range(pTHX_ regnode_ssc *ssc, const UV start, const UV end) { PERL_ARGS_ASSERT_SSC_ADD_RANGE; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); ssc->invlist = _add_range_to_invlist(ssc->invlist, start, end); } @@ -1383,7 +1383,7 @@ S_ssc_cp_and(pTHX_ regnode_ssc *ssc, const UV cp) PERL_ARGS_ASSERT_SSC_CP_AND; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); cp_list = add_cp_to_invlist(cp_list, cp); ssc_intersection(ssc, cp_list, @@ -1399,7 +1399,7 @@ S_ssc_clear_locale(pTHX_ regnode_ssc *ssc) PERL_ARGS_ASSERT_SSC_CLEAR_LOCALE; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); ANYOF_POSIXL_ZERO(ssc); ANYOF_FLAGS(ssc) &= ~ANYOF_LOCALE_FLAGS; @@ -1416,7 +1416,7 @@ S_ssc_finalize(pTHX_ RExC_state_t *pRExC_state, regnode_ssc *ssc) PERL_ARGS_ASSERT_SSC_FINALIZE; - assert(OP(ssc) == ANYOF_SYNTHETIC); + assert(is_ANYOF_SYNTHETIC(ssc)); /* The code in this file assumes that all but these flags aren't relevant * to the SSC, except ANYOF_EMPTY_STRING, which should be cleared by the @@ -204,6 +204,19 @@ struct regnode_ssc { SV* invlist; /* list of code points matched */ }; +/* We take advantage of 'next_off' not otherwise being used in the SSC by + * actually using it: by setting it to 1. This allows us to test and + * distinguish between an SSC and other ANYOF node types, as 'next_off' cannot + * otherwise be 1, because it is the offset to the next regnode expressed in + * units of regnodes. Since an ANYOF node contains extra fields, it adds up + * to 12 regnode units on 32-bit systems, (hence the minimum this can be (if + * not 0) is 11 there. Even if things get tightly packed on a 64-bit system, + * it still would be more than 1. */ +#define set_ANYOF_SYNTHETIC(n) STMT_START{ OP(n) = ANYOF; \ + NEXT_OFF(n) = 1; \ + } STMT_END +#define is_ANYOF_SYNTHETIC(n) (OP(n) == ANYOF && NEXT_OFF(n) == 1) + /* XXX fix this description. Impose a limit of REG_INFTY on various pattern matching operations to limit stack growth and to avoid "infinite" recursions. @@ -322,10 +335,9 @@ struct regnode_ssc { * performance penalty. Better would be to split it off into a separate node, * which actually would improve performance a bit by allowing regexec.c to test * for a UTF-8 character being above 255 without having to call a function nor - * calculate its code point value. However, this solution might need to have a - * second node type, ANYOF_SYNTHETIC_ABOVE_LATIN1_ALL. Several flags are not - * used in synthetic start class (SSC) nodes, so could be shared should new - * flags be needed for SSCs. */ + * calculate its code point value. Several flags are not used in synthetic + * start class (SSC) nodes, so could be shared should new flags be needed for + * SSCs. */ /* regexec.c is expecting this to be in the low bit */ #define ANYOF_INVERT 0x01 diff --git a/regcomp.sym b/regcomp.sym index 09962ddc71..a1981862cc 100644 --- a/regcomp.sym +++ b/regcomp.sym @@ -55,7 +55,6 @@ REG_ANY REG_ANY, no 0 S ; Match any one character (except newline). SANY REG_ANY, no 0 S ; Match any one character. CANY REG_ANY, no 0 S ; Match any one byte. ANYOF ANYOF, sv 0 S ; Match character in (or not in) this class, single char match only -ANYOF_SYNTHETIC ANYOF, sv 0 S ; Synthetic start class # Order of the below is important. See ordering comment above. POSIXD POSIXD, none 0 S ; Some [[:class:]] under /d; the FLAGS field gives which one @@ -1497,7 +1497,6 @@ S_find_byclass(pTHX_ regexp * prog, const regnode *c, char *s, /* We know what class it must start with. */ switch (OP(c)) { case ANYOF: - case ANYOF_SYNTHETIC: if (utf8_target) { REXEC_FBC_UTF8_CLASS_SCAN( reginclass(prog, c, (U8*)s, (U8*) strend, utf8_target)); @@ -7536,7 +7535,7 @@ S_reginclass(pTHX_ regexp * const prog, const regnode * const n, const U8* const || (utf8_target && (c >=256 || (! (flags & ANYOF_LOCALE)) - || OP(n) == ANYOF_SYNTHETIC)))) + || is_ANYOF_SYNTHETIC(n))))) { SV * const sw = core_regclass_swash(prog, n, TRUE, 0); if (sw) { diff --git a/regnodes.h b/regnodes.h index f1a5fe2703..0af89b90f9 100644 --- a/regnodes.h +++ b/regnodes.h @@ -6,8 +6,8 @@ /* Regops and State definitions */ -#define REGNODE_MAX 94 -#define REGMATCH_STATE_MAX 134 +#define REGNODE_MAX 93 +#define REGMATCH_STATE_MAX 133 #define END 0 /* 0000 End of program. */ #define SUCCEED 1 /* 0x01 Return from a subroutine, basically. */ @@ -31,79 +31,78 @@ #define SANY 19 /* 0x13 Match any one character. */ #define CANY 20 /* 0x14 Match any one byte. */ #define ANYOF 21 /* 0x15 Match character in (or not in) this class, single char match only */ -#define ANYOF_SYNTHETIC 22 /* 0x16 Synthetic start class */ -#define POSIXD 23 /* 0x17 Some [[:class:]] under /d; the FLAGS field gives which one */ -#define POSIXL 24 /* 0x18 Some [[:class:]] under /l; the FLAGS field gives which one */ -#define POSIXU 25 /* 0x19 Some [[:class:]] under /u; the FLAGS field gives which one */ -#define POSIXA 26 /* 0x1a Some [[:class:]] under /a; the FLAGS field gives which one */ -#define NPOSIXD 27 /* 0x1b complement of POSIXD, [[:^class:]] */ -#define NPOSIXL 28 /* 0x1c complement of POSIXL, [[:^class:]] */ -#define NPOSIXU 29 /* 0x1d complement of POSIXU, [[:^class:]] */ -#define NPOSIXA 30 /* 0x1e complement of POSIXA, [[:^class:]] */ -#define CLUMP 31 /* 0x1f Match any extended grapheme cluster sequence */ -#define BRANCH 32 /* 0x20 Match this alternative, or the next... */ -#define BACK 33 /* 0x21 Match "", "next" ptr points backward. */ -#define EXACT 34 /* 0x22 Match this string (preceded by length). */ -#define EXACTF 35 /* 0x23 Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ -#define EXACTFL 36 /* 0x24 Match this string (not guaranteed to be folded) using /il rules (w/len). */ -#define EXACTFU 37 /* 0x25 Match this string (folded iff in UTF-8, length in folding doesn't change if not in UTF-8) using /iu rules (w/len). */ -#define EXACTFA 38 /* 0x26 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ -#define EXACTFU_SS 39 /* 0x27 Match this string (folded iff in UTF-8, length in folding may change even if not in UTF-8) using /iu rules (w/len). */ -#define EXACTFA_NO_TRIE 40 /* 0x28 Match this string (which is not trie-able; not guaranteed to be folded) using /iaa rules (w/len). */ -#define NOTHING 41 /* 0x29 Match empty string. */ -#define TAIL 42 /* 0x2a Match empty string. Can jump here from outside. */ -#define STAR 43 /* 0x2b Match this (simple) thing 0 or more times. */ -#define PLUS 44 /* 0x2c Match this (simple) thing 1 or more times. */ -#define CURLY 45 /* 0x2d Match this simple thing {n,m} times. */ -#define CURLYN 46 /* 0x2e Capture next-after-this simple thing */ -#define CURLYM 47 /* 0x2f Capture this medium-complex thing {n,m} times. */ -#define CURLYX 48 /* 0x30 Match this complex thing {n,m} times. */ -#define WHILEM 49 /* 0x31 Do curly processing and see if rest matches. */ -#define OPEN 50 /* 0x32 Mark this point in input as start of #n. */ -#define CLOSE 51 /* 0x33 Analogous to OPEN. */ -#define REF 52 /* 0x34 Match some already matched string */ -#define REFF 53 /* 0x35 Match already matched string, folded using native charset semantics for non-utf8 */ -#define REFFL 54 /* 0x36 Match already matched string, folded in loc. */ -#define REFFU 55 /* 0x37 Match already matched string, folded using unicode semantics for non-utf8 */ -#define REFFA 56 /* 0x38 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ -#define NREF 57 /* 0x39 Match some already matched string */ -#define NREFF 58 /* 0x3a Match already matched string, folded using native charset semantics for non-utf8 */ -#define NREFFL 59 /* 0x3b Match already matched string, folded in loc. */ -#define NREFFU 60 /* 0x3c Match already matched string, folded using unicode semantics for non-utf8 */ -#define NREFFA 61 /* 0x3d Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ -#define IFMATCH 62 /* 0x3e Succeeds if the following matches. */ -#define UNLESSM 63 /* 0x3f Fails if the following matches. */ -#define SUSPEND 64 /* 0x40 "Independent" sub-RE. */ -#define IFTHEN 65 /* 0x41 Switch, should be preceded by switcher. */ -#define GROUPP 66 /* 0x42 Whether the group matched. */ -#define LONGJMP 67 /* 0x43 Jump far away. */ -#define BRANCHJ 68 /* 0x44 BRANCH with long offset. */ -#define EVAL 69 /* 0x45 Execute some Perl code. */ -#define MINMOD 70 /* 0x46 Next operator is not greedy. */ -#define LOGICAL 71 /* 0x47 Next opcode should set the flag only. */ -#define RENUM 72 /* 0x48 Group with independently numbered parens. */ -#define TRIE 73 /* 0x49 Match many EXACT(F[ALU]?)? at once. flags==type */ -#define TRIEC 74 /* 0x4a Same as TRIE, but with embedded charclass data */ -#define AHOCORASICK 75 /* 0x4b Aho Corasick stclass. flags==type */ -#define AHOCORASICKC 76 /* 0x4c Same as AHOCORASICK, but with embedded charclass data */ -#define GOSUB 77 /* 0x4d recurse to paren arg1 at (signed) ofs arg2 */ -#define GOSTART 78 /* 0x4e recurse to start of pattern */ -#define NGROUPP 79 /* 0x4f Whether the group matched. */ -#define INSUBP 80 /* 0x50 Whether we are in a specific recurse. */ -#define DEFINEP 81 /* 0x51 Never execute directly. */ -#define ENDLIKE 82 /* 0x52 Used only for the type field of verbs */ -#define OPFAIL 83 /* 0x53 Same as (?!) */ -#define ACCEPT 84 /* 0x54 Accepts the current matched string. */ -#define VERB 85 /* 0x55 Used only for the type field of verbs */ -#define PRUNE 86 /* 0x56 Pattern fails at this startpoint if no-backtracking through this */ -#define MARKPOINT 87 /* 0x57 Push the current location for rollback by cut. */ -#define SKIP 88 /* 0x58 On failure skip forward (to the mark) before retrying */ -#define COMMIT 89 /* 0x59 Pattern fails outright if backtracking through this */ -#define CUTGROUP 90 /* 0x5a On failure go to the next alternation in the group */ -#define KEEPS 91 /* 0x5b $& begins here. */ -#define LNBREAK 92 /* 0x5c generic newline pattern */ -#define OPTIMIZED 93 /* 0x5d Placeholder for dump. */ -#define PSEUDO 94 /* 0x5e Pseudo opcode for internal use. */ +#define POSIXD 22 /* 0x16 Some [[:class:]] under /d; the FLAGS field gives which one */ +#define POSIXL 23 /* 0x17 Some [[:class:]] under /l; the FLAGS field gives which one */ +#define POSIXU 24 /* 0x18 Some [[:class:]] under /u; the FLAGS field gives which one */ +#define POSIXA 25 /* 0x19 Some [[:class:]] under /a; the FLAGS field gives which one */ +#define NPOSIXD 26 /* 0x1a complement of POSIXD, [[:^class:]] */ +#define NPOSIXL 27 /* 0x1b complement of POSIXL, [[:^class:]] */ +#define NPOSIXU 28 /* 0x1c complement of POSIXU, [[:^class:]] */ +#define NPOSIXA 29 /* 0x1d complement of POSIXA, [[:^class:]] */ +#define CLUMP 30 /* 0x1e Match any extended grapheme cluster sequence */ +#define BRANCH 31 /* 0x1f Match this alternative, or the next... */ +#define BACK 32 /* 0x20 Match "", "next" ptr points backward. */ +#define EXACT 33 /* 0x21 Match this string (preceded by length). */ +#define EXACTF 34 /* 0x22 Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ +#define EXACTFL 35 /* 0x23 Match this string (not guaranteed to be folded) using /il rules (w/len). */ +#define EXACTFU 36 /* 0x24 Match this string (folded iff in UTF-8, length in folding doesn't change if not in UTF-8) using /iu rules (w/len). */ +#define EXACTFA 37 /* 0x25 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ +#define EXACTFU_SS 38 /* 0x26 Match this string (folded iff in UTF-8, length in folding may change even if not in UTF-8) using /iu rules (w/len). */ +#define EXACTFA_NO_TRIE 39 /* 0x27 Match this string (which is not trie-able; not guaranteed to be folded) using /iaa rules (w/len). */ +#define NOTHING 40 /* 0x28 Match empty string. */ +#define TAIL 41 /* 0x29 Match empty string. Can jump here from outside. */ +#define STAR 42 /* 0x2a Match this (simple) thing 0 or more times. */ +#define PLUS 43 /* 0x2b Match this (simple) thing 1 or more times. */ +#define CURLY 44 /* 0x2c Match this simple thing {n,m} times. */ +#define CURLYN 45 /* 0x2d Capture next-after-this simple thing */ +#define CURLYM 46 /* 0x2e Capture this medium-complex thing {n,m} times. */ +#define CURLYX 47 /* 0x2f Match this complex thing {n,m} times. */ +#define WHILEM 48 /* 0x30 Do curly processing and see if rest matches. */ +#define OPEN 49 /* 0x31 Mark this point in input as start of #n. */ +#define CLOSE 50 /* 0x32 Analogous to OPEN. */ +#define REF 51 /* 0x33 Match some already matched string */ +#define REFF 52 /* 0x34 Match already matched string, folded using native charset semantics for non-utf8 */ +#define REFFL 53 /* 0x35 Match already matched string, folded in loc. */ +#define REFFU 54 /* 0x36 Match already matched string, folded using unicode semantics for non-utf8 */ +#define REFFA 55 /* 0x37 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#define NREF 56 /* 0x38 Match some already matched string */ +#define NREFF 57 /* 0x39 Match already matched string, folded using native charset semantics for non-utf8 */ +#define NREFFL 58 /* 0x3a Match already matched string, folded in loc. */ +#define NREFFU 59 /* 0x3b Match already matched string, folded using unicode semantics for non-utf8 */ +#define NREFFA 60 /* 0x3c Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#define IFMATCH 61 /* 0x3d Succeeds if the following matches. */ +#define UNLESSM 62 /* 0x3e Fails if the following matches. */ +#define SUSPEND 63 /* 0x3f "Independent" sub-RE. */ +#define IFTHEN 64 /* 0x40 Switch, should be preceded by switcher. */ +#define GROUPP 65 /* 0x41 Whether the group matched. */ +#define LONGJMP 66 /* 0x42 Jump far away. */ +#define BRANCHJ 67 /* 0x43 BRANCH with long offset. */ +#define EVAL 68 /* 0x44 Execute some Perl code. */ +#define MINMOD 69 /* 0x45 Next operator is not greedy. */ +#define LOGICAL 70 /* 0x46 Next opcode should set the flag only. */ +#define RENUM 71 /* 0x47 Group with independently numbered parens. */ +#define TRIE 72 /* 0x48 Match many EXACT(F[ALU]?)? at once. flags==type */ +#define TRIEC 73 /* 0x49 Same as TRIE, but with embedded charclass data */ +#define AHOCORASICK 74 /* 0x4a Aho Corasick stclass. flags==type */ +#define AHOCORASICKC 75 /* 0x4b Same as AHOCORASICK, but with embedded charclass data */ +#define GOSUB 76 /* 0x4c recurse to paren arg1 at (signed) ofs arg2 */ +#define GOSTART 77 /* 0x4d recurse to start of pattern */ +#define NGROUPP 78 /* 0x4e Whether the group matched. */ +#define INSUBP 79 /* 0x4f Whether we are in a specific recurse. */ +#define DEFINEP 80 /* 0x50 Never execute directly. */ +#define ENDLIKE 81 /* 0x51 Used only for the type field of verbs */ +#define OPFAIL 82 /* 0x52 Same as (?!) */ +#define ACCEPT 83 /* 0x53 Accepts the current matched string. */ +#define VERB 84 /* 0x54 Used only for the type field of verbs */ +#define PRUNE 85 /* 0x55 Pattern fails at this startpoint if no-backtracking through this */ +#define MARKPOINT 86 /* 0x56 Push the current location for rollback by cut. */ +#define SKIP 87 /* 0x57 On failure skip forward (to the mark) before retrying */ +#define COMMIT 88 /* 0x58 Pattern fails outright if backtracking through this */ +#define CUTGROUP 89 /* 0x59 On failure go to the next alternation in the group */ +#define KEEPS 90 /* 0x5a $& begins here. */ +#define LNBREAK 91 /* 0x5b generic newline pattern */ +#define OPTIMIZED 92 /* 0x5c Placeholder for dump. */ +#define PSEUDO 93 /* 0x5d Pseudo opcode for internal use. */ /* ------------ States ------------- */ #define TRIE_next (REGNODE_MAX + 1) /* state for TRIE */ #define TRIE_next_fail (REGNODE_MAX + 2) /* state for TRIE */ @@ -174,7 +173,6 @@ EXTCONST U8 PL_regkind[] = { REG_ANY, /* SANY */ REG_ANY, /* CANY */ ANYOF, /* ANYOF */ - ANYOF, /* ANYOF_SYNTHETIC */ POSIXD, /* POSIXD */ POSIXD, /* POSIXL */ POSIXD, /* POSIXU */ @@ -317,7 +315,6 @@ static const U8 regarglen[] = { 0, /* SANY */ 0, /* CANY */ 0, /* ANYOF */ - 0, /* ANYOF_SYNTHETIC */ 0, /* POSIXD */ 0, /* POSIXL */ 0, /* POSIXU */ @@ -417,7 +414,6 @@ static const char reg_off_by_arg[] = { 0, /* SANY */ 0, /* CANY */ 0, /* ANYOF */ - 0, /* ANYOF_SYNTHETIC */ 0, /* POSIXD */ 0, /* POSIXL */ 0, /* POSIXU */ @@ -522,79 +518,78 @@ EXTCONST char * const PL_reg_name[] = { "SANY", /* 0x13 */ "CANY", /* 0x14 */ "ANYOF", /* 0x15 */ - "ANYOF_SYNTHETIC", /* 0x16 */ - "POSIXD", /* 0x17 */ - "POSIXL", /* 0x18 */ - "POSIXU", /* 0x19 */ - "POSIXA", /* 0x1a */ - "NPOSIXD", /* 0x1b */ - "NPOSIXL", /* 0x1c */ - "NPOSIXU", /* 0x1d */ - "NPOSIXA", /* 0x1e */ - "CLUMP", /* 0x1f */ - "BRANCH", /* 0x20 */ - "BACK", /* 0x21 */ - "EXACT", /* 0x22 */ - "EXACTF", /* 0x23 */ - "EXACTFL", /* 0x24 */ - "EXACTFU", /* 0x25 */ - "EXACTFA", /* 0x26 */ - "EXACTFU_SS", /* 0x27 */ - "EXACTFA_NO_TRIE", /* 0x28 */ - "NOTHING", /* 0x29 */ - "TAIL", /* 0x2a */ - "STAR", /* 0x2b */ - "PLUS", /* 0x2c */ - "CURLY", /* 0x2d */ - "CURLYN", /* 0x2e */ - "CURLYM", /* 0x2f */ - "CURLYX", /* 0x30 */ - "WHILEM", /* 0x31 */ - "OPEN", /* 0x32 */ - "CLOSE", /* 0x33 */ - "REF", /* 0x34 */ - "REFF", /* 0x35 */ - "REFFL", /* 0x36 */ - "REFFU", /* 0x37 */ - "REFFA", /* 0x38 */ - "NREF", /* 0x39 */ - "NREFF", /* 0x3a */ - "NREFFL", /* 0x3b */ - "NREFFU", /* 0x3c */ - "NREFFA", /* 0x3d */ - "IFMATCH", /* 0x3e */ - "UNLESSM", /* 0x3f */ - "SUSPEND", /* 0x40 */ - "IFTHEN", /* 0x41 */ - "GROUPP", /* 0x42 */ - "LONGJMP", /* 0x43 */ - "BRANCHJ", /* 0x44 */ - "EVAL", /* 0x45 */ - "MINMOD", /* 0x46 */ - "LOGICAL", /* 0x47 */ - "RENUM", /* 0x48 */ - "TRIE", /* 0x49 */ - "TRIEC", /* 0x4a */ - "AHOCORASICK", /* 0x4b */ - "AHOCORASICKC", /* 0x4c */ - "GOSUB", /* 0x4d */ - "GOSTART", /* 0x4e */ - "NGROUPP", /* 0x4f */ - "INSUBP", /* 0x50 */ - "DEFINEP", /* 0x51 */ - "ENDLIKE", /* 0x52 */ - "OPFAIL", /* 0x53 */ - "ACCEPT", /* 0x54 */ - "VERB", /* 0x55 */ - "PRUNE", /* 0x56 */ - "MARKPOINT", /* 0x57 */ - "SKIP", /* 0x58 */ - "COMMIT", /* 0x59 */ - "CUTGROUP", /* 0x5a */ - "KEEPS", /* 0x5b */ - "LNBREAK", /* 0x5c */ - "OPTIMIZED", /* 0x5d */ - "PSEUDO", /* 0x5e */ + "POSIXD", /* 0x16 */ + "POSIXL", /* 0x17 */ + "POSIXU", /* 0x18 */ + "POSIXA", /* 0x19 */ + "NPOSIXD", /* 0x1a */ + "NPOSIXL", /* 0x1b */ + "NPOSIXU", /* 0x1c */ + "NPOSIXA", /* 0x1d */ + "CLUMP", /* 0x1e */ + "BRANCH", /* 0x1f */ + "BACK", /* 0x20 */ + "EXACT", /* 0x21 */ + "EXACTF", /* 0x22 */ + "EXACTFL", /* 0x23 */ + "EXACTFU", /* 0x24 */ + "EXACTFA", /* 0x25 */ + "EXACTFU_SS", /* 0x26 */ + "EXACTFA_NO_TRIE", /* 0x27 */ + "NOTHING", /* 0x28 */ + "TAIL", /* 0x29 */ + "STAR", /* 0x2a */ + "PLUS", /* 0x2b */ + "CURLY", /* 0x2c */ + "CURLYN", /* 0x2d */ + "CURLYM", /* 0x2e */ + "CURLYX", /* 0x2f */ + "WHILEM", /* 0x30 */ + "OPEN", /* 0x31 */ + "CLOSE", /* 0x32 */ + "REF", /* 0x33 */ + "REFF", /* 0x34 */ + "REFFL", /* 0x35 */ + "REFFU", /* 0x36 */ + "REFFA", /* 0x37 */ + "NREF", /* 0x38 */ + "NREFF", /* 0x39 */ + "NREFFL", /* 0x3a */ + "NREFFU", /* 0x3b */ + "NREFFA", /* 0x3c */ + "IFMATCH", /* 0x3d */ + "UNLESSM", /* 0x3e */ + "SUSPEND", /* 0x3f */ + "IFTHEN", /* 0x40 */ + "GROUPP", /* 0x41 */ + "LONGJMP", /* 0x42 */ + "BRANCHJ", /* 0x43 */ + "EVAL", /* 0x44 */ + "MINMOD", /* 0x45 */ + "LOGICAL", /* 0x46 */ + "RENUM", /* 0x47 */ + "TRIE", /* 0x48 */ + "TRIEC", /* 0x49 */ + "AHOCORASICK", /* 0x4a */ + "AHOCORASICKC", /* 0x4b */ + "GOSUB", /* 0x4c */ + "GOSTART", /* 0x4d */ + "NGROUPP", /* 0x4e */ + "INSUBP", /* 0x4f */ + "DEFINEP", /* 0x50 */ + "ENDLIKE", /* 0x51 */ + "OPFAIL", /* 0x52 */ + "ACCEPT", /* 0x53 */ + "VERB", /* 0x54 */ + "PRUNE", /* 0x55 */ + "MARKPOINT", /* 0x56 */ + "SKIP", /* 0x57 */ + "COMMIT", /* 0x58 */ + "CUTGROUP", /* 0x59 */ + "KEEPS", /* 0x5a */ + "LNBREAK", /* 0x5b */ + "OPTIMIZED", /* 0x5c */ + "PSEUDO", /* 0x5d */ /* ------------ States ------------- */ "TRIE_next", /* REGNODE_MAX +0x01 */ "TRIE_next_fail", /* REGNODE_MAX +0x02 */ @@ -714,7 +709,7 @@ EXTCONST U8 PL_varies[] __attribute__deprecated__ = { EXTCONST U8 PL_varies_bitmask[]; #else EXTCONST U8 PL_varies_bitmask[] = { - 0x00, 0x00, 0x00, 0x80, 0x03, 0xF8, 0xF3, 0x3F, 0x13, 0x00, 0x00, 0x00 + 0x00, 0x00, 0x00, 0xC0, 0x01, 0xFC, 0xF9, 0x9F, 0x09, 0x00, 0x00, 0x00 }; #endif /* DOINIT */ @@ -726,8 +721,8 @@ EXTCONST U8 PL_varies_bitmask[] = { EXTCONST U8 PL_simple[] __attribute__deprecated__; #else EXTCONST U8 PL_simple[] __attribute__deprecated__ = { - REG_ANY, SANY, CANY, ANYOF, ANYOF_SYNTHETIC, POSIXD, POSIXL, POSIXU, - POSIXA, NPOSIXD, NPOSIXL, NPOSIXU, NPOSIXA, + REG_ANY, SANY, CANY, ANYOF, POSIXD, POSIXL, POSIXU, POSIXA, NPOSIXD, + NPOSIXL, NPOSIXU, NPOSIXA, 0 }; #endif /* DOINIT */ @@ -736,7 +731,7 @@ EXTCONST U8 PL_simple[] __attribute__deprecated__ = { EXTCONST U8 PL_simple_bitmask[]; #else EXTCONST U8 PL_simple_bitmask[] = { - 0x00, 0x00, 0xFC, 0x7F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 + 0x00, 0x00, 0xFC, 0x3F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; #endif /* DOINIT */ |