X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/13d6edb45df716e9bd8a40bf4d0dcc0834c8e1d8..a06de4dcf2448ead9a5f5dbd87366727ebd91aca:/regnodes.h diff --git a/regnodes.h b/regnodes.h index 2f21dd9..8fe0f41 100644 --- a/regnodes.h +++ b/regnodes.h @@ -1,140 +1,152 @@ /* -*- buffer-read-only: t -*- !!!!!!! DO NOT EDIT THIS FILE !!!!!!! - This file is built by regcomp.pl from regcomp.sym. + This file is built by regen/regcomp.pl from regcomp.sym. Any changes made here will be lost! -*/ + */ /* Regops and State definitions */ -#define REGNODE_MAX 84 -#define REGMATCH_STATE_MAX 124 +#define REGNODE_MAX 92 +#define REGMATCH_STATE_MAX 134 #define END 0 /* 0000 End of program. */ #define SUCCEED 1 /* 0x01 Return from a subroutine, basically. */ -#define BOL 2 /* 0x02 Match "" at beginning of line. */ -#define MBOL 3 /* 0x03 Same, assuming multiline. */ -#define SBOL 4 /* 0x04 Same, assuming singleline. */ -#define EOS 5 /* 0x05 Match "" at end of string. */ -#define EOL 6 /* 0x06 Match "" at end of line. */ -#define MEOL 7 /* 0x07 Same, assuming multiline. */ -#define SEOL 8 /* 0x08 Same, assuming singleline. */ -#define BOUND 9 /* 0x09 Match "" at any word boundary */ -#define BOUNDL 10 /* 0x0a Match "" at any word boundary */ -#define NBOUND 11 /* 0x0b Match "" at any word non-boundary */ -#define NBOUNDL 12 /* 0x0c Match "" at any word non-boundary */ -#define GPOS 13 /* 0x0d Matches where last m//g left off. */ -#define REG_ANY 14 /* 0x0e Match any one character (except newline). */ -#define SANY 15 /* 0x0f Match any one character. */ -#define CANY 16 /* 0x10 Match any one byte. */ -#define ANYOF 17 /* 0x11 Match character in (or not in) this class. */ -#define ALNUM 18 /* 0x12 Match any alphanumeric character */ -#define ALNUML 19 /* 0x13 Match any alphanumeric char in locale */ -#define NALNUM 20 /* 0x14 Match any non-alphanumeric character */ -#define NALNUML 21 /* 0x15 Match any non-alphanumeric char in locale */ -#define SPACE 22 /* 0x16 Match any whitespace character */ -#define SPACEL 23 /* 0x17 Match any whitespace char in locale */ -#define NSPACE 24 /* 0x18 Match any non-whitespace character */ -#define NSPACEL 25 /* 0x19 Match any non-whitespace char in locale */ -#define DIGIT 26 /* 0x1a Match any numeric character */ -#define DIGITL 27 /* 0x1b Match any numeric character in locale */ -#define NDIGIT 28 /* 0x1c Match any non-numeric character */ -#define NDIGITL 29 /* 0x1d Match any non-numeric character in locale */ -#define CLUMP 30 /* 0x1e Match any combining character 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 string, folded (prec. by length). */ -#define EXACTFL 35 /* 0x23 Match this string, folded in locale (w/len). */ -#define NOTHING 36 /* 0x24 Match empty string. */ -#define TAIL 37 /* 0x25 Match empty string. Can jump here from outside. */ -#define STAR 38 /* 0x26 Match this (simple) thing 0 or more times. */ -#define PLUS 39 /* 0x27 Match this (simple) thing 1 or more times. */ -#define CURLY 40 /* 0x28 Match this simple thing {n,m} times. */ -#define CURLYN 41 /* 0x29 Capture next-after-this simple thing */ -#define CURLYM 42 /* 0x2a Capture this medium-complex thing {n,m} times. */ -#define CURLYX 43 /* 0x2b Match this complex thing {n,m} times. */ -#define WHILEM 44 /* 0x2c Do curly processing and see if rest matches. */ -#define OPEN 45 /* 0x2d Mark this point in input as start of */ -#define CLOSE 46 /* 0x2e Analogous to OPEN. */ -#define REF 47 /* 0x2f Match some already matched string */ -#define REFF 48 /* 0x30 Match already matched string, folded */ -#define REFFL 49 /* 0x31 Match already matched string, folded in loc. */ -#define IFMATCH 50 /* 0x32 Succeeds if the following matches. */ -#define UNLESSM 51 /* 0x33 Fails if the following matches. */ -#define SUSPEND 52 /* 0x34 "Independent" sub-RE. */ -#define IFTHEN 53 /* 0x35 Switch, should be preceeded by switcher . */ -#define GROUPP 54 /* 0x36 Whether the group matched. */ -#define LONGJMP 55 /* 0x37 Jump far away. */ -#define BRANCHJ 56 /* 0x38 BRANCH with long offset. */ -#define EVAL 57 /* 0x39 Execute some Perl code. */ -#define MINMOD 58 /* 0x3a Next operator is not greedy. */ -#define LOGICAL 59 /* 0x3b Next opcode should set the flag only. */ -#define RENUM 60 /* 0x3c Group with independently numbered parens. */ -#define TRIE 61 /* 0x3d Match many EXACT(FL?)? at once. flags==type */ -#define TRIEC 62 /* 0x3e Same as TRIE, but with embedded charclass data */ -#define AHOCORASICK 63 /* 0x3f Aho Corasick stclass. flags==type */ -#define AHOCORASICKC 64 /* 0x40 Same as AHOCORASICK, but with embedded charclass data */ -#define GOSUB 65 /* 0x41 recurse to paren arg1 at (signed) ofs arg2 */ -#define GOSTART 66 /* 0x42 recurse to start of pattern */ -#define NREF 67 /* 0x43 Match some already matched string */ -#define NREFF 68 /* 0x44 Match already matched string, folded */ -#define NREFFL 69 /* 0x45 Match already matched string, folded in loc. */ -#define NGROUPP 70 /* 0x46 Whether the group matched. */ -#define INSUBP 71 /* 0x47 Whether we are in a specific recurse. */ -#define DEFINEP 72 /* 0x48 Never execute directly. */ -#define ENDLIKE 73 /* 0x49 Used only for the type field of verbs */ -#define OPFAIL 74 /* 0x4a Same as (?!) */ -#define ACCEPT 75 /* 0x4b Accepts the current matched string. */ -#define VERB 76 /* 0x4c no-sv 1 Used only for the type field of verbs */ -#define PRUNE 77 /* 0x4d Pattern fails at this startpoint if no-backtracking through this */ -#define MARKPOINT 78 /* 0x4e Push the current location for rollback by cut. */ -#define SKIP 79 /* 0x4f On failure skip forward (to the mark) before retrying */ -#define COMMIT 80 /* 0x50 Pattern fails outright if backtracking through this */ -#define CUTGROUP 81 /* 0x51 On failure go to the next alternation in the group */ -#define KEEPS 82 /* 0x52 $& begins here. */ -#define OPTIMIZED 83 /* 0x53 Placeholder for dump. */ -#define PSEUDO 84 /* 0x54 Pseudo opcode for internal use. */ +#define SBOL 2 /* 0x02 Match "" at beginning of line: /^/, /\A/ */ +#define BOL 2 /* 0x02 type alias */ +#define MBOL 3 /* 0x03 Same, assuming multiline: /^/m */ +#define SEOL 4 /* 0x04 Match "" at end of line: /$/ */ +#define EOL 4 /* 0x04 type alias */ +#define MEOL 5 /* 0x05 Same, assuming multiline: /$/m */ +#define EOS 6 /* 0x06 Match "" at end of string: /\z/ */ +#define GPOS 7 /* 0x07 Matches where last m//g left off. */ +#define BOUND 8 /* 0x08 Like BOUNDA for non-utf8, otherwise match "" between any Unicode \w\W or \W\w */ +#define BOUNDL 9 /* 0x09 Like BOUND/BOUNDU, but \w and \W are defined by current locale */ +#define BOUNDU 10 /* 0x0a Match "" at any boundary of a given type using Unicode rules */ +#define BOUNDA 11 /* 0x0b Match "" at any boundary between \w\W or \W\w, where \w is [_a-zA-Z0-9] */ +#define NBOUND 12 /* 0x0c Like NBOUNDA for non-utf8, otherwise match "" between any Unicode \w\w or \W\W */ +#define NBOUNDL 13 /* 0x0d Like NBOUND/NBOUNDU, but \w and \W are defined by current locale */ +#define NBOUNDU 14 /* 0x0e Match "" at any non-boundary of a given type using using Unicode rules */ +#define NBOUNDA 15 /* 0x0f Match "" betweeen any \w\w or \W\W, where \w is [_a-zA-Z0-9] */ +#define REG_ANY 16 /* 0x10 Match any one character (except newline). */ +#define SANY 17 /* 0x11 Match any one character. */ +#define ANYOF 18 /* 0x12 Match character in (or not in) this class, single char match only */ +#define ANYOFD 19 /* 0x13 Like ANYOF, but /d is in effect */ +#define ANYOFL 20 /* 0x14 Like ANYOF, but /l is in effect */ +#define POSIXD 21 /* 0x15 Some [[:class:]] under /d; the FLAGS field gives which one */ +#define POSIXL 22 /* 0x16 Some [[:class:]] under /l; the FLAGS field gives which one */ +#define POSIXU 23 /* 0x17 Some [[:class:]] under /u; the FLAGS field gives which one */ +#define POSIXA 24 /* 0x18 Some [[:class:]] under /a; the FLAGS field gives which one */ +#define NPOSIXD 25 /* 0x19 complement of POSIXD, [[:^class:]] */ +#define NPOSIXL 26 /* 0x1a complement of POSIXL, [[:^class:]] */ +#define NPOSIXU 27 /* 0x1b complement of POSIXU, [[:^class:]] */ +#define NPOSIXA 28 /* 0x1c complement of POSIXA, [[:^class:]] */ +#define CLUMP 29 /* 0x1d Match any extended grapheme cluster sequence */ +#define BRANCH 30 /* 0x1e Match this alternative, or the next... */ +#define EXACT 31 /* 0x1f Match this string (preceded by length). */ +#define EXACTL 32 /* 0x20 Like EXACT, but /l is in effect (used so locale-related warnings can be checked for). */ +#define EXACTF 33 /* 0x21 Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ +#define EXACTFL 34 /* 0x22 Match this string (not guaranteed to be folded) using /il rules (w/len). */ +#define EXACTFU 35 /* 0x23 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 36 /* 0x24 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ +#define EXACTFU_SS 37 /* 0x25 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 EXACTFLU8 38 /* 0x26 Rare cirucmstances: like EXACTFU, but is under /l, UTF-8, folded, and everything in it is above 255. */ +#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 rules 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 rules for non-utf8 */ +#define REFFA 55 /* 0x37 Match already matched string, folded using unicode rules 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 rules 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 rules for non-utf8 */ +#define NREFFA 60 /* 0x3c Match already matched string, folded using unicode rules for non-utf8, no mixing ASCII, non-ASCII */ +#define LONGJMP 61 /* 0x3d Jump far away. */ +#define BRANCHJ 62 /* 0x3e BRANCH with long offset. */ +#define IFMATCH 63 /* 0x3f Succeeds if the following matches. */ +#define UNLESSM 64 /* 0x40 Fails if the following matches. */ +#define SUSPEND 65 /* 0x41 "Independent" sub-RE. */ +#define IFTHEN 66 /* 0x42 Switch, should be preceded by switcher. */ +#define GROUPP 67 /* 0x43 Whether the group matched. */ +#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 NGROUPP 77 /* 0x4d Whether the group matched. */ +#define INSUBP 78 /* 0x4e Whether we are in a specific recurse. */ +#define DEFINEP 79 /* 0x4f Never execute directly. */ +#define ENDLIKE 80 /* 0x50 Used only for the type field of verbs */ +#define OPFAIL 81 /* 0x51 Same as (?!), but with verb arg */ +#define ACCEPT 82 /* 0x52 Accepts the current matched string, with verbar */ +#define VERB 83 /* 0x53 Used only for the type field of verbs */ +#define PRUNE 84 /* 0x54 Pattern fails at this startpoint if no-backtracking through this */ +#define MARKPOINT 85 /* 0x55 Push the current location for rollback by cut. */ +#define SKIP 86 /* 0x56 On failure skip forward (to the mark) before retrying */ +#define COMMIT 87 /* 0x57 Pattern fails outright if backtracking through this */ +#define CUTGROUP 88 /* 0x58 On failure go to the next alternation in the group */ +#define KEEPS 89 /* 0x59 $& begins here. */ +#define LNBREAK 90 /* 0x5a generic newline pattern */ +#define OPTIMIZED 91 /* 0x5b Placeholder for dump. */ +#define PSEUDO 92 /* 0x5c 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 */ -#define EVAL_AB (REGNODE_MAX + 3) /* state for EVAL */ -#define EVAL_AB_fail (REGNODE_MAX + 4) /* state for EVAL */ -#define CURLYX_end (REGNODE_MAX + 5) /* state for CURLYX */ -#define CURLYX_end_fail (REGNODE_MAX + 6) /* state for CURLYX */ -#define WHILEM_A_pre (REGNODE_MAX + 7) /* state for WHILEM */ -#define WHILEM_A_pre_fail (REGNODE_MAX + 8) /* state for WHILEM */ -#define WHILEM_A_min (REGNODE_MAX + 9) /* state for WHILEM */ -#define WHILEM_A_min_fail (REGNODE_MAX + 10) /* state for WHILEM */ -#define WHILEM_A_max (REGNODE_MAX + 11) /* state for WHILEM */ -#define WHILEM_A_max_fail (REGNODE_MAX + 12) /* state for WHILEM */ -#define WHILEM_B_min (REGNODE_MAX + 13) /* state for WHILEM */ -#define WHILEM_B_min_fail (REGNODE_MAX + 14) /* state for WHILEM */ -#define WHILEM_B_max (REGNODE_MAX + 15) /* state for WHILEM */ -#define WHILEM_B_max_fail (REGNODE_MAX + 16) /* state for WHILEM */ -#define BRANCH_next (REGNODE_MAX + 17) /* state for BRANCH */ -#define BRANCH_next_fail (REGNODE_MAX + 18) /* state for BRANCH */ -#define CURLYM_A (REGNODE_MAX + 19) /* state for CURLYM */ -#define CURLYM_A_fail (REGNODE_MAX + 20) /* state for CURLYM */ -#define CURLYM_B (REGNODE_MAX + 21) /* state for CURLYM */ -#define CURLYM_B_fail (REGNODE_MAX + 22) /* state for CURLYM */ -#define IFMATCH_A (REGNODE_MAX + 23) /* state for IFMATCH */ -#define IFMATCH_A_fail (REGNODE_MAX + 24) /* state for IFMATCH */ -#define CURLY_B_min_known (REGNODE_MAX + 25) /* state for CURLY */ -#define CURLY_B_min_known_fail (REGNODE_MAX + 26) /* state for CURLY */ -#define CURLY_B_min (REGNODE_MAX + 27) /* state for CURLY */ -#define CURLY_B_min_fail (REGNODE_MAX + 28) /* state for CURLY */ -#define CURLY_B_max (REGNODE_MAX + 29) /* state for CURLY */ -#define CURLY_B_max_fail (REGNODE_MAX + 30) /* state for CURLY */ -#define COMMIT_next (REGNODE_MAX + 31) /* state for COMMIT */ -#define COMMIT_next_fail (REGNODE_MAX + 32) /* state for COMMIT */ -#define MARKPOINT_next (REGNODE_MAX + 33) /* state for MARKPOINT */ -#define MARKPOINT_next_fail (REGNODE_MAX + 34) /* state for MARKPOINT */ -#define SKIP_next (REGNODE_MAX + 35) /* state for SKIP */ -#define SKIP_next_fail (REGNODE_MAX + 36) /* state for SKIP */ -#define CUTGROUP_next (REGNODE_MAX + 37) /* state for CUTGROUP */ -#define CUTGROUP_next_fail (REGNODE_MAX + 38) /* state for CUTGROUP */ -#define KEEPS_next (REGNODE_MAX + 39) /* state for KEEPS */ -#define KEEPS_next_fail (REGNODE_MAX + 40) /* state for KEEPS */ +#define EVAL_B (REGNODE_MAX + 3) /* state for EVAL */ +#define EVAL_B_fail (REGNODE_MAX + 4) /* state for EVAL */ +#define EVAL_postponed_AB (REGNODE_MAX + 5) /* state for EVAL */ +#define EVAL_postponed_AB_fail (REGNODE_MAX + 6) /* state for EVAL */ +#define CURLYX_end (REGNODE_MAX + 7) /* state for CURLYX */ +#define CURLYX_end_fail (REGNODE_MAX + 8) /* state for CURLYX */ +#define WHILEM_A_pre (REGNODE_MAX + 9) /* state for WHILEM */ +#define WHILEM_A_pre_fail (REGNODE_MAX + 10) /* state for WHILEM */ +#define WHILEM_A_min (REGNODE_MAX + 11) /* state for WHILEM */ +#define WHILEM_A_min_fail (REGNODE_MAX + 12) /* state for WHILEM */ +#define WHILEM_A_max (REGNODE_MAX + 13) /* state for WHILEM */ +#define WHILEM_A_max_fail (REGNODE_MAX + 14) /* state for WHILEM */ +#define WHILEM_B_min (REGNODE_MAX + 15) /* state for WHILEM */ +#define WHILEM_B_min_fail (REGNODE_MAX + 16) /* state for WHILEM */ +#define WHILEM_B_max (REGNODE_MAX + 17) /* state for WHILEM */ +#define WHILEM_B_max_fail (REGNODE_MAX + 18) /* state for WHILEM */ +#define BRANCH_next (REGNODE_MAX + 19) /* state for BRANCH */ +#define BRANCH_next_fail (REGNODE_MAX + 20) /* state for BRANCH */ +#define CURLYM_A (REGNODE_MAX + 21) /* state for CURLYM */ +#define CURLYM_A_fail (REGNODE_MAX + 22) /* state for CURLYM */ +#define CURLYM_B (REGNODE_MAX + 23) /* state for CURLYM */ +#define CURLYM_B_fail (REGNODE_MAX + 24) /* state for CURLYM */ +#define IFMATCH_A (REGNODE_MAX + 25) /* state for IFMATCH */ +#define IFMATCH_A_fail (REGNODE_MAX + 26) /* state for IFMATCH */ +#define CURLY_B_min_known (REGNODE_MAX + 27) /* state for CURLY */ +#define CURLY_B_min_known_fail (REGNODE_MAX + 28) /* state for CURLY */ +#define CURLY_B_min (REGNODE_MAX + 29) /* state for CURLY */ +#define CURLY_B_min_fail (REGNODE_MAX + 30) /* state for CURLY */ +#define CURLY_B_max (REGNODE_MAX + 31) /* state for CURLY */ +#define CURLY_B_max_fail (REGNODE_MAX + 32) /* state for CURLY */ +#define COMMIT_next (REGNODE_MAX + 33) /* state for COMMIT */ +#define COMMIT_next_fail (REGNODE_MAX + 34) /* state for COMMIT */ +#define MARKPOINT_next (REGNODE_MAX + 35) /* state for MARKPOINT */ +#define MARKPOINT_next_fail (REGNODE_MAX + 36) /* state for MARKPOINT */ +#define SKIP_next (REGNODE_MAX + 37) /* state for SKIP */ +#define SKIP_next_fail (REGNODE_MAX + 38) /* state for SKIP */ +#define CUTGROUP_next (REGNODE_MAX + 39) /* state for CUTGROUP */ +#define CUTGROUP_next_fail (REGNODE_MAX + 40) /* state for CUTGROUP */ +#define KEEPS_next (REGNODE_MAX + 41) /* state for KEEPS */ +#define KEEPS_next_fail (REGNODE_MAX + 42) /* state for KEEPS */ /* PL_regkind[] What type of regop or state is this. */ @@ -144,40 +156,44 @@ EXTCONST U8 PL_regkind[]; EXTCONST U8 PL_regkind[] = { END, /* END */ END, /* SUCCEED */ - BOL, /* BOL */ - BOL, /* MBOL */ BOL, /* SBOL */ - EOL, /* EOS */ - EOL, /* EOL */ - EOL, /* MEOL */ + BOL, /* MBOL */ EOL, /* SEOL */ + EOL, /* MEOL */ + EOL, /* EOS */ + GPOS, /* GPOS */ BOUND, /* BOUND */ BOUND, /* BOUNDL */ + BOUND, /* BOUNDU */ + BOUND, /* BOUNDA */ NBOUND, /* NBOUND */ NBOUND, /* NBOUNDL */ - GPOS, /* GPOS */ + NBOUND, /* NBOUNDU */ + NBOUND, /* NBOUNDA */ REG_ANY, /* REG_ANY */ REG_ANY, /* SANY */ - REG_ANY, /* CANY */ ANYOF, /* ANYOF */ - ALNUM, /* ALNUM */ - ALNUM, /* ALNUML */ - NALNUM, /* NALNUM */ - NALNUM, /* NALNUML */ - SPACE, /* SPACE */ - SPACE, /* SPACEL */ - NSPACE, /* NSPACE */ - NSPACE, /* NSPACEL */ - DIGIT, /* DIGIT */ - DIGIT, /* DIGITL */ - NDIGIT, /* NDIGIT */ - NDIGIT, /* NDIGITL */ + ANYOF, /* ANYOFD */ + ANYOF, /* ANYOFL */ + POSIXD, /* POSIXD */ + POSIXD, /* POSIXL */ + POSIXD, /* POSIXU */ + POSIXD, /* POSIXA */ + NPOSIXD, /* NPOSIXD */ + NPOSIXD, /* NPOSIXL */ + NPOSIXD, /* NPOSIXU */ + NPOSIXD, /* NPOSIXA */ CLUMP, /* CLUMP */ BRANCH, /* BRANCH */ - BACK, /* BACK */ EXACT, /* EXACT */ + EXACT, /* EXACTL */ EXACT, /* EXACTF */ EXACT, /* EXACTFL */ + EXACT, /* EXACTFU */ + EXACT, /* EXACTFA */ + EXACT, /* EXACTFU_SS */ + EXACT, /* EXACTFLU8 */ + EXACT, /* EXACTFA_NO_TRIE */ NOTHING, /* NOTHING */ NOTHING, /* TAIL */ STAR, /* STAR */ @@ -192,13 +208,20 @@ EXTCONST U8 PL_regkind[] = { REF, /* REF */ REF, /* REFF */ REF, /* REFFL */ + REF, /* REFFU */ + REF, /* REFFA */ + REF, /* NREF */ + REF, /* NREFF */ + REF, /* NREFFL */ + REF, /* NREFFU */ + REF, /* NREFFA */ + LONGJMP, /* LONGJMP */ + BRANCHJ, /* BRANCHJ */ BRANCHJ, /* IFMATCH */ BRANCHJ, /* UNLESSM */ BRANCHJ, /* SUSPEND */ BRANCHJ, /* IFTHEN */ GROUPP, /* GROUPP */ - LONGJMP, /* LONGJMP */ - BRANCHJ, /* BRANCHJ */ EVAL, /* EVAL */ MINMOD, /* MINMOD */ LOGICAL, /* LOGICAL */ @@ -208,10 +231,6 @@ EXTCONST U8 PL_regkind[] = { TRIE, /* AHOCORASICK */ TRIE, /* AHOCORASICKC */ GOSUB, /* GOSUB */ - GOSTART, /* GOSTART */ - REF, /* NREF */ - REF, /* NREFF */ - REF, /* NREFFL */ NGROUPP, /* NGROUPP */ INSUBP, /* INSUBP */ DEFINEP, /* DEFINEP */ @@ -225,13 +244,16 @@ EXTCONST U8 PL_regkind[] = { VERB, /* COMMIT */ VERB, /* CUTGROUP */ KEEPS, /* KEEPS */ + LNBREAK, /* LNBREAK */ NOTHING, /* OPTIMIZED */ PSEUDO, /* PSEUDO */ /* ------------ States ------------- */ TRIE, /* TRIE_next */ TRIE, /* TRIE_next_fail */ - EVAL, /* EVAL_AB */ - EVAL, /* EVAL_AB_fail */ + EVAL, /* EVAL_B */ + EVAL, /* EVAL_B_fail */ + EVAL, /* EVAL_postponed_AB */ + EVAL, /* EVAL_postponed_AB_fail */ CURLYX, /* CURLYX_end */ CURLYX, /* CURLYX_end_fail */ WHILEM, /* WHILEM_A_pre */ @@ -271,46 +293,51 @@ EXTCONST U8 PL_regkind[] = { }; #endif +#ifdef REG_COMP_C + /* regarglen[] - How large is the argument part of the node (in regnodes) */ -#ifdef REG_COMP_C static const U8 regarglen[] = { 0, /* END */ 0, /* SUCCEED */ - 0, /* BOL */ - 0, /* MBOL */ 0, /* SBOL */ - 0, /* EOS */ - 0, /* EOL */ - 0, /* MEOL */ + 0, /* MBOL */ 0, /* SEOL */ + 0, /* MEOL */ + 0, /* EOS */ + 0, /* GPOS */ 0, /* BOUND */ 0, /* BOUNDL */ + 0, /* BOUNDU */ + 0, /* BOUNDA */ 0, /* NBOUND */ 0, /* NBOUNDL */ - 0, /* GPOS */ + 0, /* NBOUNDU */ + 0, /* NBOUNDA */ 0, /* REG_ANY */ 0, /* SANY */ - 0, /* CANY */ - 0, /* ANYOF */ - 0, /* ALNUM */ - 0, /* ALNUML */ - 0, /* NALNUM */ - 0, /* NALNUML */ - 0, /* SPACE */ - 0, /* SPACEL */ - 0, /* NSPACE */ - 0, /* NSPACEL */ - 0, /* DIGIT */ - 0, /* DIGITL */ - 0, /* NDIGIT */ - 0, /* NDIGITL */ + EXTRA_SIZE(struct regnode_1), /* ANYOF */ + EXTRA_SIZE(struct regnode_1), /* ANYOFD */ + EXTRA_SIZE(struct regnode_1), /* ANYOFL */ + 0, /* POSIXD */ + 0, /* POSIXL */ + 0, /* POSIXU */ + 0, /* POSIXA */ + 0, /* NPOSIXD */ + 0, /* NPOSIXL */ + 0, /* NPOSIXU */ + 0, /* NPOSIXA */ 0, /* CLUMP */ 0, /* BRANCH */ - 0, /* BACK */ 0, /* EXACT */ + 0, /* EXACTL */ 0, /* EXACTF */ 0, /* EXACTFL */ + 0, /* EXACTFU */ + 0, /* EXACTFA */ + 0, /* EXACTFU_SS */ + 0, /* EXACTFLU8 */ + 0, /* EXACTFA_NO_TRIE */ 0, /* NOTHING */ 0, /* TAIL */ 0, /* STAR */ @@ -325,14 +352,21 @@ static const U8 regarglen[] = { EXTRA_SIZE(struct regnode_1), /* REF */ EXTRA_SIZE(struct regnode_1), /* REFF */ EXTRA_SIZE(struct regnode_1), /* REFFL */ + EXTRA_SIZE(struct regnode_1), /* REFFU */ + EXTRA_SIZE(struct regnode_1), /* REFFA */ + EXTRA_SIZE(struct regnode_1), /* NREF */ + EXTRA_SIZE(struct regnode_1), /* NREFF */ + EXTRA_SIZE(struct regnode_1), /* NREFFL */ + EXTRA_SIZE(struct regnode_1), /* NREFFU */ + EXTRA_SIZE(struct regnode_1), /* NREFFA */ + EXTRA_SIZE(struct regnode_1), /* LONGJMP */ + EXTRA_SIZE(struct regnode_1), /* BRANCHJ */ EXTRA_SIZE(struct regnode_1), /* IFMATCH */ EXTRA_SIZE(struct regnode_1), /* UNLESSM */ EXTRA_SIZE(struct regnode_1), /* SUSPEND */ EXTRA_SIZE(struct regnode_1), /* IFTHEN */ EXTRA_SIZE(struct regnode_1), /* GROUPP */ - EXTRA_SIZE(struct regnode_1), /* LONGJMP */ - EXTRA_SIZE(struct regnode_1), /* BRANCHJ */ - EXTRA_SIZE(struct regnode_1), /* EVAL */ + EXTRA_SIZE(struct regnode_2L), /* EVAL */ 0, /* MINMOD */ 0, /* LOGICAL */ EXTRA_SIZE(struct regnode_1), /* RENUM */ @@ -341,23 +375,20 @@ static const U8 regarglen[] = { EXTRA_SIZE(struct regnode_1), /* AHOCORASICK */ EXTRA_SIZE(struct regnode_charclass), /* AHOCORASICKC */ EXTRA_SIZE(struct regnode_2L), /* GOSUB */ - 0, /* GOSTART */ - EXTRA_SIZE(struct regnode_1), /* NREF */ - EXTRA_SIZE(struct regnode_1), /* NREFF */ - EXTRA_SIZE(struct regnode_1), /* NREFFL */ EXTRA_SIZE(struct regnode_1), /* NGROUPP */ EXTRA_SIZE(struct regnode_1), /* INSUBP */ EXTRA_SIZE(struct regnode_1), /* DEFINEP */ 0, /* ENDLIKE */ - 0, /* OPFAIL */ - EXTRA_SIZE(struct regnode_1), /* ACCEPT */ - 0, /* VERB */ + EXTRA_SIZE(struct regnode_1), /* OPFAIL */ + EXTRA_SIZE(struct regnode_2L), /* ACCEPT */ + EXTRA_SIZE(struct regnode_1), /* VERB */ EXTRA_SIZE(struct regnode_1), /* PRUNE */ EXTRA_SIZE(struct regnode_1), /* MARKPOINT */ EXTRA_SIZE(struct regnode_1), /* SKIP */ EXTRA_SIZE(struct regnode_1), /* COMMIT */ EXTRA_SIZE(struct regnode_1), /* CUTGROUP */ 0, /* KEEPS */ + 0, /* LNBREAK */ 0, /* OPTIMIZED */ 0, /* PSEUDO */ }; @@ -367,40 +398,44 @@ static const U8 regarglen[] = { static const char reg_off_by_arg[] = { 0, /* END */ 0, /* SUCCEED */ - 0, /* BOL */ - 0, /* MBOL */ 0, /* SBOL */ - 0, /* EOS */ - 0, /* EOL */ - 0, /* MEOL */ + 0, /* MBOL */ 0, /* SEOL */ + 0, /* MEOL */ + 0, /* EOS */ + 0, /* GPOS */ 0, /* BOUND */ 0, /* BOUNDL */ + 0, /* BOUNDU */ + 0, /* BOUNDA */ 0, /* NBOUND */ 0, /* NBOUNDL */ - 0, /* GPOS */ + 0, /* NBOUNDU */ + 0, /* NBOUNDA */ 0, /* REG_ANY */ 0, /* SANY */ - 0, /* CANY */ 0, /* ANYOF */ - 0, /* ALNUM */ - 0, /* ALNUML */ - 0, /* NALNUM */ - 0, /* NALNUML */ - 0, /* SPACE */ - 0, /* SPACEL */ - 0, /* NSPACE */ - 0, /* NSPACEL */ - 0, /* DIGIT */ - 0, /* DIGITL */ - 0, /* NDIGIT */ - 0, /* NDIGITL */ + 0, /* ANYOFD */ + 0, /* ANYOFL */ + 0, /* POSIXD */ + 0, /* POSIXL */ + 0, /* POSIXU */ + 0, /* POSIXA */ + 0, /* NPOSIXD */ + 0, /* NPOSIXL */ + 0, /* NPOSIXU */ + 0, /* NPOSIXA */ 0, /* CLUMP */ 0, /* BRANCH */ - 0, /* BACK */ 0, /* EXACT */ + 0, /* EXACTL */ 0, /* EXACTF */ 0, /* EXACTFL */ + 0, /* EXACTFU */ + 0, /* EXACTFA */ + 0, /* EXACTFU_SS */ + 0, /* EXACTFLU8 */ + 0, /* EXACTFA_NO_TRIE */ 0, /* NOTHING */ 0, /* TAIL */ 0, /* STAR */ @@ -415,13 +450,20 @@ static const char reg_off_by_arg[] = { 0, /* REF */ 0, /* REFF */ 0, /* REFFL */ + 0, /* REFFU */ + 0, /* REFFA */ + 0, /* NREF */ + 0, /* NREFF */ + 0, /* NREFFL */ + 0, /* NREFFU */ + 0, /* NREFFA */ + 1, /* LONGJMP */ + 1, /* BRANCHJ */ 2, /* IFMATCH */ 2, /* UNLESSM */ 1, /* SUSPEND */ 1, /* IFTHEN */ 0, /* GROUPP */ - 1, /* LONGJMP */ - 1, /* BRANCHJ */ 0, /* EVAL */ 0, /* MINMOD */ 0, /* LOGICAL */ @@ -431,10 +473,6 @@ static const char reg_off_by_arg[] = { 0, /* AHOCORASICK */ 0, /* AHOCORASICKC */ 0, /* GOSUB */ - 0, /* GOSTART */ - 0, /* NREF */ - 0, /* NREFF */ - 0, /* NREFFL */ 0, /* NGROUPP */ 0, /* INSUBP */ 0, /* DEFINEP */ @@ -448,147 +486,273 @@ static const char reg_off_by_arg[] = { 0, /* COMMIT */ 0, /* CUTGROUP */ 0, /* KEEPS */ + 0, /* LNBREAK */ 0, /* OPTIMIZED */ 0, /* PSEUDO */ }; #endif /* REG_COMP_C */ + /* reg_name[] - Opcode/state names in string form, for debugging */ -#ifdef DEBUGGING -# ifndef DOINIT +#ifndef DOINIT EXTCONST char * PL_reg_name[]; -# else -EXTCONST char * PL_reg_name[] = { +#else +EXTCONST char * const PL_reg_name[] = { "END", /* 0000 */ "SUCCEED", /* 0x01 */ - "BOL", /* 0x02 */ + "SBOL", /* 0x02 */ "MBOL", /* 0x03 */ - "SBOL", /* 0x04 */ - "EOS", /* 0x05 */ - "EOL", /* 0x06 */ - "MEOL", /* 0x07 */ - "SEOL", /* 0x08 */ - "BOUND", /* 0x09 */ - "BOUNDL", /* 0x0a */ - "NBOUND", /* 0x0b */ - "NBOUNDL", /* 0x0c */ - "GPOS", /* 0x0d */ - "REG_ANY", /* 0x0e */ - "SANY", /* 0x0f */ - "CANY", /* 0x10 */ - "ANYOF", /* 0x11 */ - "ALNUM", /* 0x12 */ - "ALNUML", /* 0x13 */ - "NALNUM", /* 0x14 */ - "NALNUML", /* 0x15 */ - "SPACE", /* 0x16 */ - "SPACEL", /* 0x17 */ - "NSPACE", /* 0x18 */ - "NSPACEL", /* 0x19 */ - "DIGIT", /* 0x1a */ - "DIGITL", /* 0x1b */ - "NDIGIT", /* 0x1c */ - "NDIGITL", /* 0x1d */ - "CLUMP", /* 0x1e */ - "BRANCH", /* 0x1f */ - "BACK", /* 0x20 */ - "EXACT", /* 0x21 */ - "EXACTF", /* 0x22 */ - "EXACTFL", /* 0x23 */ - "NOTHING", /* 0x24 */ - "TAIL", /* 0x25 */ - "STAR", /* 0x26 */ - "PLUS", /* 0x27 */ - "CURLY", /* 0x28 */ - "CURLYN", /* 0x29 */ - "CURLYM", /* 0x2a */ - "CURLYX", /* 0x2b */ - "WHILEM", /* 0x2c */ - "OPEN", /* 0x2d */ - "CLOSE", /* 0x2e */ - "REF", /* 0x2f */ - "REFF", /* 0x30 */ - "REFFL", /* 0x31 */ - "IFMATCH", /* 0x32 */ - "UNLESSM", /* 0x33 */ - "SUSPEND", /* 0x34 */ - "IFTHEN", /* 0x35 */ - "GROUPP", /* 0x36 */ - "LONGJMP", /* 0x37 */ - "BRANCHJ", /* 0x38 */ - "EVAL", /* 0x39 */ - "MINMOD", /* 0x3a */ - "LOGICAL", /* 0x3b */ - "RENUM", /* 0x3c */ - "TRIE", /* 0x3d */ - "TRIEC", /* 0x3e */ - "AHOCORASICK", /* 0x3f */ - "AHOCORASICKC", /* 0x40 */ - "GOSUB", /* 0x41 */ - "GOSTART", /* 0x42 */ - "NREF", /* 0x43 */ - "NREFF", /* 0x44 */ - "NREFFL", /* 0x45 */ - "NGROUPP", /* 0x46 */ - "INSUBP", /* 0x47 */ - "DEFINEP", /* 0x48 */ - "ENDLIKE", /* 0x49 */ - "OPFAIL", /* 0x4a */ - "ACCEPT", /* 0x4b */ - "VERB", /* 0x4c */ - "PRUNE", /* 0x4d */ - "MARKPOINT", /* 0x4e */ - "SKIP", /* 0x4f */ - "COMMIT", /* 0x50 */ - "CUTGROUP", /* 0x51 */ - "KEEPS", /* 0x52 */ - "OPTIMIZED", /* 0x53 */ - "PSEUDO", /* 0x54 */ + "SEOL", /* 0x04 */ + "MEOL", /* 0x05 */ + "EOS", /* 0x06 */ + "GPOS", /* 0x07 */ + "BOUND", /* 0x08 */ + "BOUNDL", /* 0x09 */ + "BOUNDU", /* 0x0a */ + "BOUNDA", /* 0x0b */ + "NBOUND", /* 0x0c */ + "NBOUNDL", /* 0x0d */ + "NBOUNDU", /* 0x0e */ + "NBOUNDA", /* 0x0f */ + "REG_ANY", /* 0x10 */ + "SANY", /* 0x11 */ + "ANYOF", /* 0x12 */ + "ANYOFD", /* 0x13 */ + "ANYOFL", /* 0x14 */ + "POSIXD", /* 0x15 */ + "POSIXL", /* 0x16 */ + "POSIXU", /* 0x17 */ + "POSIXA", /* 0x18 */ + "NPOSIXD", /* 0x19 */ + "NPOSIXL", /* 0x1a */ + "NPOSIXU", /* 0x1b */ + "NPOSIXA", /* 0x1c */ + "CLUMP", /* 0x1d */ + "BRANCH", /* 0x1e */ + "EXACT", /* 0x1f */ + "EXACTL", /* 0x20 */ + "EXACTF", /* 0x21 */ + "EXACTFL", /* 0x22 */ + "EXACTFU", /* 0x23 */ + "EXACTFA", /* 0x24 */ + "EXACTFU_SS", /* 0x25 */ + "EXACTFLU8", /* 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 */ + "LONGJMP", /* 0x3d */ + "BRANCHJ", /* 0x3e */ + "IFMATCH", /* 0x3f */ + "UNLESSM", /* 0x40 */ + "SUSPEND", /* 0x41 */ + "IFTHEN", /* 0x42 */ + "GROUPP", /* 0x43 */ + "EVAL", /* 0x44 */ + "MINMOD", /* 0x45 */ + "LOGICAL", /* 0x46 */ + "RENUM", /* 0x47 */ + "TRIE", /* 0x48 */ + "TRIEC", /* 0x49 */ + "AHOCORASICK", /* 0x4a */ + "AHOCORASICKC", /* 0x4b */ + "GOSUB", /* 0x4c */ + "NGROUPP", /* 0x4d */ + "INSUBP", /* 0x4e */ + "DEFINEP", /* 0x4f */ + "ENDLIKE", /* 0x50 */ + "OPFAIL", /* 0x51 */ + "ACCEPT", /* 0x52 */ + "VERB", /* 0x53 */ + "PRUNE", /* 0x54 */ + "MARKPOINT", /* 0x55 */ + "SKIP", /* 0x56 */ + "COMMIT", /* 0x57 */ + "CUTGROUP", /* 0x58 */ + "KEEPS", /* 0x59 */ + "LNBREAK", /* 0x5a */ + "OPTIMIZED", /* 0x5b */ + "PSEUDO", /* 0x5c */ /* ------------ States ------------- */ "TRIE_next", /* REGNODE_MAX +0x01 */ "TRIE_next_fail", /* REGNODE_MAX +0x02 */ - "EVAL_AB", /* REGNODE_MAX +0x03 */ - "EVAL_AB_fail", /* REGNODE_MAX +0x04 */ - "CURLYX_end", /* REGNODE_MAX +0x05 */ - "CURLYX_end_fail", /* REGNODE_MAX +0x06 */ - "WHILEM_A_pre", /* REGNODE_MAX +0x07 */ - "WHILEM_A_pre_fail", /* REGNODE_MAX +0x08 */ - "WHILEM_A_min", /* REGNODE_MAX +0x09 */ - "WHILEM_A_min_fail", /* REGNODE_MAX +0x0a */ - "WHILEM_A_max", /* REGNODE_MAX +0x0b */ - "WHILEM_A_max_fail", /* REGNODE_MAX +0x0c */ - "WHILEM_B_min", /* REGNODE_MAX +0x0d */ - "WHILEM_B_min_fail", /* REGNODE_MAX +0x0e */ - "WHILEM_B_max", /* REGNODE_MAX +0x0f */ - "WHILEM_B_max_fail", /* REGNODE_MAX +0x10 */ - "BRANCH_next", /* REGNODE_MAX +0x11 */ - "BRANCH_next_fail", /* REGNODE_MAX +0x12 */ - "CURLYM_A", /* REGNODE_MAX +0x13 */ - "CURLYM_A_fail", /* REGNODE_MAX +0x14 */ - "CURLYM_B", /* REGNODE_MAX +0x15 */ - "CURLYM_B_fail", /* REGNODE_MAX +0x16 */ - "IFMATCH_A", /* REGNODE_MAX +0x17 */ - "IFMATCH_A_fail", /* REGNODE_MAX +0x18 */ - "CURLY_B_min_known", /* REGNODE_MAX +0x19 */ - "CURLY_B_min_known_fail", /* REGNODE_MAX +0x1a */ - "CURLY_B_min", /* REGNODE_MAX +0x1b */ - "CURLY_B_min_fail", /* REGNODE_MAX +0x1c */ - "CURLY_B_max", /* REGNODE_MAX +0x1d */ - "CURLY_B_max_fail", /* REGNODE_MAX +0x1e */ - "COMMIT_next", /* REGNODE_MAX +0x1f */ - "COMMIT_next_fail", /* REGNODE_MAX +0x20 */ - "MARKPOINT_next", /* REGNODE_MAX +0x21 */ - "MARKPOINT_next_fail", /* REGNODE_MAX +0x22 */ - "SKIP_next", /* REGNODE_MAX +0x23 */ - "SKIP_next_fail", /* REGNODE_MAX +0x24 */ - "CUTGROUP_next", /* REGNODE_MAX +0x25 */ - "CUTGROUP_next_fail", /* REGNODE_MAX +0x26 */ - "KEEPS_next", /* REGNODE_MAX +0x27 */ - "KEEPS_next_fail", /* REGNODE_MAX +0x28 */ + "EVAL_B", /* REGNODE_MAX +0x03 */ + "EVAL_B_fail", /* REGNODE_MAX +0x04 */ + "EVAL_postponed_AB", /* REGNODE_MAX +0x05 */ + "EVAL_postponed_AB_fail", /* REGNODE_MAX +0x06 */ + "CURLYX_end", /* REGNODE_MAX +0x07 */ + "CURLYX_end_fail", /* REGNODE_MAX +0x08 */ + "WHILEM_A_pre", /* REGNODE_MAX +0x09 */ + "WHILEM_A_pre_fail", /* REGNODE_MAX +0x0a */ + "WHILEM_A_min", /* REGNODE_MAX +0x0b */ + "WHILEM_A_min_fail", /* REGNODE_MAX +0x0c */ + "WHILEM_A_max", /* REGNODE_MAX +0x0d */ + "WHILEM_A_max_fail", /* REGNODE_MAX +0x0e */ + "WHILEM_B_min", /* REGNODE_MAX +0x0f */ + "WHILEM_B_min_fail", /* REGNODE_MAX +0x10 */ + "WHILEM_B_max", /* REGNODE_MAX +0x11 */ + "WHILEM_B_max_fail", /* REGNODE_MAX +0x12 */ + "BRANCH_next", /* REGNODE_MAX +0x13 */ + "BRANCH_next_fail", /* REGNODE_MAX +0x14 */ + "CURLYM_A", /* REGNODE_MAX +0x15 */ + "CURLYM_A_fail", /* REGNODE_MAX +0x16 */ + "CURLYM_B", /* REGNODE_MAX +0x17 */ + "CURLYM_B_fail", /* REGNODE_MAX +0x18 */ + "IFMATCH_A", /* REGNODE_MAX +0x19 */ + "IFMATCH_A_fail", /* REGNODE_MAX +0x1a */ + "CURLY_B_min_known", /* REGNODE_MAX +0x1b */ + "CURLY_B_min_known_fail", /* REGNODE_MAX +0x1c */ + "CURLY_B_min", /* REGNODE_MAX +0x1d */ + "CURLY_B_min_fail", /* REGNODE_MAX +0x1e */ + "CURLY_B_max", /* REGNODE_MAX +0x1f */ + "CURLY_B_max_fail", /* REGNODE_MAX +0x20 */ + "COMMIT_next", /* REGNODE_MAX +0x21 */ + "COMMIT_next_fail", /* REGNODE_MAX +0x22 */ + "MARKPOINT_next", /* REGNODE_MAX +0x23 */ + "MARKPOINT_next_fail", /* REGNODE_MAX +0x24 */ + "SKIP_next", /* REGNODE_MAX +0x25 */ + "SKIP_next_fail", /* REGNODE_MAX +0x26 */ + "CUTGROUP_next", /* REGNODE_MAX +0x27 */ + "CUTGROUP_next_fail", /* REGNODE_MAX +0x28 */ + "KEEPS_next", /* REGNODE_MAX +0x29 */ + "KEEPS_next_fail", /* REGNODE_MAX +0x2a */ +}; +#endif /* DOINIT */ + +/* PL_reg_extflags_name[] - Opcode/state names in string form, for debugging */ + +#ifndef DOINIT +EXTCONST char * PL_reg_extflags_name[]; +#else +EXTCONST char * const PL_reg_extflags_name[] = { + /* Bits in extflags defined: 11111111111111110000111111111111 */ + "MULTILINE", /* 0x00000001 */ + "SINGLELINE", /* 0x00000002 */ + "FOLD", /* 0x00000004 */ + "EXTENDED", /* 0x00000008 */ + "EXTENDED_MORE", /* 0x00000010 */ + "NOCAPTURE", /* 0x00000020 */ + "KEEPCOPY", /* 0x00000040 */ + "CHARSET0", /* 0x00000080 : "CHARSET" - 0x00000380 */ + "CHARSET1", /* 0x00000100 : "CHARSET" - 0x00000380 */ + "CHARSET2", /* 0x00000200 : "CHARSET" - 0x00000380 */ + "STRICT", /* 0x00000400 */ + "SPLIT", /* 0x00000800 */ + "UNUSED_BIT_12", /* 0x00001000 */ + "UNUSED_BIT_13", /* 0x00002000 */ + "UNUSED_BIT_14", /* 0x00004000 */ + "UNUSED_BIT_15", /* 0x00008000 */ + "NO_INPLACE_SUBST", /* 0x00010000 */ + "EVAL_SEEN", /* 0x00020000 */ + "UNBOUNDED_QUANTIFIER_SEEN",/* 0x00040000 */ + "CHECK_ALL", /* 0x00080000 */ + "MATCH_UTF8", /* 0x00100000 */ + "USE_INTUIT_NOML", /* 0x00200000 */ + "USE_INTUIT_ML", /* 0x00400000 */ + "INTUIT_TAIL", /* 0x00800000 */ + "IS_ANCHORED", /* 0x01000000 */ + "COPY_DONE", /* 0x02000000 */ + "TAINTED_SEEN", /* 0x04000000 */ + "TAINTED", /* 0x08000000 */ + "START_ONLY", /* 0x10000000 */ + "SKIPWHITE", /* 0x20000000 */ + "WHITE", /* 0x40000000 */ + "NULL", /* 0x80000000 */ +}; +#endif /* DOINIT */ + +#ifdef DEBUGGING +# define REG_EXTFLAGS_NAME_SIZE 32 +#endif + +/* PL_reg_intflags_name[] - Opcode/state names in string form, for debugging */ + +#ifndef DOINIT +EXTCONST char * PL_reg_intflags_name[]; +#else +EXTCONST char * const PL_reg_intflags_name[] = { + "SKIP", /* 0x00000001 - PREGf_SKIP */ + "IMPLICIT", /* 0x00000002 - PREGf_IMPLICIT - Converted .* to ^.* */ + "NAUGHTY", /* 0x00000004 - PREGf_NAUGHTY - how exponential is this pattern? */ + "VERBARG_SEEN", /* 0x00000008 - PREGf_VERBARG_SEEN */ + "CUTGROUP_SEEN", /* 0x00000010 - PREGf_CUTGROUP_SEEN */ + "USE_RE_EVAL", /* 0x00000020 - PREGf_USE_RE_EVAL - compiled with "use re 'eval'" */ + "NOSCAN", /* 0x00000040 - PREGf_NOSCAN */ + "GPOS_SEEN", /* 0x00000100 - PREGf_GPOS_SEEN */ + "GPOS_FLOAT", /* 0x00000200 - PREGf_GPOS_FLOAT */ + "ANCH_MBOL", /* 0x00000400 - PREGf_ANCH_MBOL */ + "ANCH_SBOL", /* 0x00000800 - PREGf_ANCH_SBOL */ + "ANCH_GPOS", /* 0x00001000 - PREGf_ANCH_GPOS */ + "RECURSE_SEEN", /* 0x00002000 - PREGf_RECURSE_SEEN */ +}; +#endif /* DOINIT */ + +#ifdef DEBUGGING +# define REG_INTFLAGS_NAME_SIZE 13 +#endif + +/* The following have no fixed length. U8 so we can do strchr() on it. */ +#define REGNODE_VARIES(node) (PL_varies_bitmask[(node) >> 3] & (1 << ((node) & 7))) + +#ifndef DOINIT +EXTCONST U8 PL_varies[] __attribute__deprecated__; +#else +EXTCONST U8 PL_varies[] __attribute__deprecated__ = { + CLUMP, BRANCH, STAR, PLUS, CURLY, CURLYN, CURLYM, CURLYX, WHILEM, REF, + REFF, REFFL, REFFU, REFFA, NREF, NREFF, NREFFL, NREFFU, NREFFA, + BRANCHJ, SUSPEND, IFTHEN, + 0 +}; +#endif /* DOINIT */ + +#ifndef DOINIT +EXTCONST U8 PL_varies_bitmask[]; +#else +EXTCONST U8 PL_varies_bitmask[] = { + 0x00, 0x00, 0x00, 0x60, 0x00, 0xFC, 0xF9, 0x5F, 0x06, 0x00, 0x00, 0x00 +}; +#endif /* DOINIT */ + +/* The following always have a length of 1. U8 we can do strchr() on it. */ +/* (Note that length 1 means "one character" under UTF8, not "one octet".) */ +#define REGNODE_SIMPLE(node) (PL_simple_bitmask[(node) >> 3] & (1 << ((node) & 7))) + +#ifndef DOINIT +EXTCONST U8 PL_simple[] __attribute__deprecated__; +#else +EXTCONST U8 PL_simple[] __attribute__deprecated__ = { + REG_ANY, SANY, ANYOF, ANYOFD, ANYOFL, POSIXD, POSIXL, POSIXU, POSIXA, + NPOSIXD, NPOSIXL, NPOSIXU, NPOSIXA, + 0 +}; +#endif /* DOINIT */ + +#ifndef DOINIT +EXTCONST U8 PL_simple_bitmask[]; +#else +EXTCONST U8 PL_simple_bitmask[] = { + 0x00, 0x00, 0xFF, 0x1F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; -# endif /* DOINIT */ -#endif /* DEBUGGING */ +#endif /* DOINIT */ /* ex: set ro: */