X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/4764e399f58c2e035a482e18aea2bd8c44fd9083..9e84774b2a68a51af5e246f365059d61ebe86fb0:/regnodes.h diff --git a/regnodes.h b/regnodes.h index e704427..2024d15 100644 --- a/regnodes.h +++ b/regnodes.h @@ -1,13 +1,13 @@ /* -*- 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 121 +#define REGMATCH_STATE_MAX 161 #define END 0 /* 0000 End of program. */ #define SUCCEED 1 /* 0x01 Return from a subroutine, basically. */ @@ -18,82 +18,119 @@ #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 BOUND 9 /* 0x09 Match "" at any word boundary using native charset semantics for non-utf8 */ +#define BOUNDL 10 /* 0x0a Match "" at any locale word boundary */ +#define BOUNDU 11 /* 0x0b Match "" at any word boundary using Unicode semantics */ +#define BOUNDA 12 /* 0x0c Match "" at any word boundary using ASCII semantics */ +#define NBOUND 13 /* 0x0d Match "" at any word non-boundary using native charset semantics for non-utf8 */ +#define NBOUNDL 14 /* 0x0e Match "" at any locale word non-boundary */ +#define NBOUNDU 15 /* 0x0f Match "" at any word non-boundary using Unicode semantics */ +#define NBOUNDA 16 /* 0x10 Match "" at any word non-boundary using ASCII semantics */ +#define GPOS 17 /* 0x11 Matches where last m//g left off. */ +#define REG_ANY 18 /* 0x12 Match any one character (except newline). */ +#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 ALNUM 22 /* 0x16 Match any alphanumeric character using native charset semantics for non-utf8 */ +#define ALNUML 23 /* 0x17 Match any alphanumeric char in locale */ +#define ALNUMU 24 /* 0x18 Match any alphanumeric char using Unicode semantics */ +#define ALNUMA 25 /* 0x19 Match [A-Za-z_0-9] */ +#define NALNUM 26 /* 0x1a Match any non-alphanumeric character using native charset semantics for non-utf8 */ +#define NALNUML 27 /* 0x1b Match any non-alphanumeric char in locale */ +#define NALNUMU 28 /* 0x1c Match any non-alphanumeric char using Unicode semantics */ +#define NALNUMA 29 /* 0x1d Match [^A-Za-z_0-9] */ +#define SPACE 30 /* 0x1e Match any whitespace character using native charset semantics for non-utf8 */ +#define SPACEL 31 /* 0x1f Match any whitespace char in locale */ +#define SPACEU 32 /* 0x20 Match any whitespace char using Unicode semantics */ +#define SPACEA 33 /* 0x21 Match [ \t\n\f\r] */ +#define NSPACE 34 /* 0x22 Match any non-whitespace character using native charset semantics for non-utf8 */ +#define NSPACEL 35 /* 0x23 Match any non-whitespace char in locale */ +#define NSPACEU 36 /* 0x24 Match any non-whitespace char using Unicode semantics */ +#define NSPACEA 37 /* 0x25 Match [^ \t\n\f\r] */ +#define DIGIT 38 /* 0x26 Match any numeric character using native charset semantics for non-utf8 */ +#define DIGITL 39 /* 0x27 Match any numeric character in locale */ +#define PLACEHOLDER1 40 /* 0x28 placeholder for missing DIGITU */ +#define DIGITA 41 /* 0x29 Match [0-9] */ +#define NDIGIT 42 /* 0x2a Match any non-numeric character using native charset semantics for non-utf8 */ +#define NDIGITL 43 /* 0x2b Match any non-numeric character in locale */ +#define PLACEHOLDER2 44 /* 0x2c placeholder for missing NDIGITU */ +#define NDIGITA 45 /* 0x2d Match [^0-9] */ +#define POSIXD 46 /* 0x2e currently unused except as a placeholder */ +#define POSIXL 47 /* 0x2f currently unused except as a placeholder */ +#define POSIXU 48 /* 0x30 currently unused except as a placeholder */ +#define POSIXA 49 /* 0x31 Some [[:class:]] under /a; the FLAGS field gives which one */ +#define NPOSIXD 50 /* 0x32 currently unused except as a placeholder */ +#define NPOSIXL 51 /* 0x33 currently unused except as a placeholder */ +#define NPOSIXU 52 /* 0x34 currently unused except as a placeholder */ +#define NPOSIXA 53 /* 0x35 complement of POSIXA, [[:^class:]] */ +#define CLUMP 54 /* 0x36 Match any extended grapheme cluster sequence */ +#define BRANCH 55 /* 0x37 Match this alternative, or the next... */ +#define BACK 56 /* 0x38 Match "", "next" ptr points backward. */ +#define EXACT 57 /* 0x39 Match this string (preceded by length). */ +#define EXACTF 58 /* 0x3a Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ +#define EXACTFL 59 /* 0x3b Match this string (not guaranteed to be folded) using /il rules (w/len). */ +#define EXACTFU 60 /* 0x3c 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 61 /* 0x3d Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ +#define EXACTFU_SS 62 /* 0x3e 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 EXACTFU_TRICKYFOLD 63 /* 0x3f Match this folded UTF-8 string using /iu rules */ +#define NOTHING 64 /* 0x40 Match empty string. */ +#define TAIL 65 /* 0x41 Match empty string. Can jump here from outside. */ +#define STAR 66 /* 0x42 Match this (simple) thing 0 or more times. */ +#define PLUS 67 /* 0x43 Match this (simple) thing 1 or more times. */ +#define CURLY 68 /* 0x44 Match this simple thing {n,m} times. */ +#define CURLYN 69 /* 0x45 Capture next-after-this simple thing */ +#define CURLYM 70 /* 0x46 Capture this medium-complex thing {n,m} times. */ +#define CURLYX 71 /* 0x47 Match this complex thing {n,m} times. */ +#define WHILEM 72 /* 0x48 Do curly processing and see if rest matches. */ +#define OPEN 73 /* 0x49 Mark this point in input as start of */ +#define CLOSE 74 /* 0x4a Analogous to OPEN. */ +#define REF 75 /* 0x4b Match some already matched string */ +#define REFF 76 /* 0x4c Match already matched string, folded using native charset semantics for non-utf8 */ +#define REFFL 77 /* 0x4d Match already matched string, folded in loc. */ +#define REFFU 78 /* 0x4e Match already matched string, folded using unicode semantics for non-utf8 */ +#define REFFA 79 /* 0x4f Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#define NREF 80 /* 0x50 Match some already matched string */ +#define NREFF 81 /* 0x51 Match already matched string, folded using native charset semantics for non-utf8 */ +#define NREFFL 82 /* 0x52 Match already matched string, folded in loc. */ +#define NREFFU 83 /* 0x53 Match already matched string, folded using unicode semantics for non-utf8 */ +#define NREFFA 84 /* 0x54 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#define IFMATCH 85 /* 0x55 Succeeds if the following matches. */ +#define UNLESSM 86 /* 0x56 Fails if the following matches. */ +#define SUSPEND 87 /* 0x57 "Independent" sub-RE. */ +#define IFTHEN 88 /* 0x58 Switch, should be preceded by switcher . */ +#define GROUPP 89 /* 0x59 Whether the group matched. */ +#define LONGJMP 90 /* 0x5a Jump far away. */ +#define BRANCHJ 91 /* 0x5b BRANCH with long offset. */ +#define EVAL 92 /* 0x5c Execute some Perl code. */ +#define MINMOD 93 /* 0x5d Next operator is not greedy. */ +#define LOGICAL 94 /* 0x5e Next opcode should set the flag only. */ +#define RENUM 95 /* 0x5f Group with independently numbered parens. */ +#define TRIE 96 /* 0x60 Match many EXACT(F[ALU]?)? at once. flags==type */ +#define TRIEC 97 /* 0x61 Same as TRIE, but with embedded charclass data */ +#define AHOCORASICK 98 /* 0x62 Aho Corasick stclass. flags==type */ +#define AHOCORASICKC 99 /* 0x63 Same as AHOCORASICK, but with embedded charclass data */ +#define GOSUB 100 /* 0x64 recurse to paren arg1 at (signed) ofs arg2 */ +#define GOSTART 101 /* 0x65 recurse to start of pattern */ +#define NGROUPP 102 /* 0x66 Whether the group matched. */ +#define INSUBP 103 /* 0x67 Whether we are in a specific recurse. */ +#define DEFINEP 104 /* 0x68 Never execute directly. */ +#define ENDLIKE 105 /* 0x69 Used only for the type field of verbs */ +#define OPFAIL 106 /* 0x6a Same as (?!) */ +#define ACCEPT 107 /* 0x6b Accepts the current matched string. */ +#define VERB 108 /* 0x6c Used only for the type field of verbs */ +#define PRUNE 109 /* 0x6d Pattern fails at this startpoint if no-backtracking through this */ +#define MARKPOINT 110 /* 0x6e Push the current location for rollback by cut. */ +#define SKIP 111 /* 0x6f On failure skip forward (to the mark) before retrying */ +#define COMMIT 112 /* 0x70 Pattern fails outright if backtracking through this */ +#define CUTGROUP 113 /* 0x71 On failure go to the next alternation in the group */ +#define KEEPS 114 /* 0x72 $& begins here. */ +#define LNBREAK 115 /* 0x73 generic newline pattern */ +#define VERTWS 116 /* 0x74 vertical whitespace (Perl 6) */ +#define NVERTWS 117 /* 0x75 not vertical whitespace (Perl 6) */ +#define HORIZWS 118 /* 0x76 horizontal whitespace (Perl 6) */ +#define NHORIZWS 119 /* 0x77 not horizontal whitespace (Perl 6) */ +#define OPTIMIZED 120 /* 0x78 Placeholder for dump. */ +#define PSEUDO 121 /* 0x79 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 */ @@ -153,8 +190,12 @@ EXTCONST U8 PL_regkind[] = { EOL, /* SEOL */ BOUND, /* BOUND */ BOUND, /* BOUNDL */ + BOUND, /* BOUNDU */ + BOUND, /* BOUNDA */ NBOUND, /* NBOUND */ NBOUND, /* NBOUNDL */ + NBOUND, /* NBOUNDU */ + NBOUND, /* NBOUNDA */ GPOS, /* GPOS */ REG_ANY, /* REG_ANY */ REG_ANY, /* SANY */ @@ -162,22 +203,46 @@ EXTCONST U8 PL_regkind[] = { ANYOF, /* ANYOF */ ALNUM, /* ALNUM */ ALNUM, /* ALNUML */ + ALNUM, /* ALNUMU */ + ALNUM, /* ALNUMA */ NALNUM, /* NALNUM */ NALNUM, /* NALNUML */ + NALNUM, /* NALNUMU */ + NALNUM, /* NALNUMA */ SPACE, /* SPACE */ SPACE, /* SPACEL */ + SPACE, /* SPACEU */ + SPACE, /* SPACEA */ NSPACE, /* NSPACE */ NSPACE, /* NSPACEL */ + NSPACE, /* NSPACEU */ + NSPACE, /* NSPACEA */ DIGIT, /* DIGIT */ DIGIT, /* DIGITL */ + NOTHING, /* PLACEHOLDER1 */ + DIGIT, /* DIGITA */ NDIGIT, /* NDIGIT */ NDIGIT, /* NDIGITL */ + NOTHING, /* PLACEHOLDER2 */ + NDIGIT, /* NDIGITA */ + 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, /* EXACTF */ EXACT, /* EXACTFL */ + EXACT, /* EXACTFU */ + EXACT, /* EXACTFA */ + EXACT, /* EXACTFU_SS */ + EXACT, /* EXACTFU_TRICKYFOLD */ NOTHING, /* NOTHING */ NOTHING, /* TAIL */ STAR, /* STAR */ @@ -192,6 +257,13 @@ EXTCONST U8 PL_regkind[] = { REF, /* REF */ REF, /* REFF */ REF, /* REFFL */ + REF, /* REFFU */ + REF, /* REFFA */ + REF, /* NREF */ + REF, /* NREFF */ + REF, /* NREFFL */ + REF, /* NREFFU */ + REF, /* NREFFA */ BRANCHJ, /* IFMATCH */ BRANCHJ, /* UNLESSM */ BRANCHJ, /* SUSPEND */ @@ -209,9 +281,6 @@ EXTCONST U8 PL_regkind[] = { TRIE, /* AHOCORASICKC */ GOSUB, /* GOSUB */ GOSTART, /* GOSTART */ - REF, /* NREF */ - REF, /* NREFF */ - REF, /* NREFFL */ NGROUPP, /* NGROUPP */ INSUBP, /* INSUBP */ DEFINEP, /* DEFINEP */ @@ -225,6 +294,11 @@ EXTCONST U8 PL_regkind[] = { VERB, /* COMMIT */ VERB, /* CUTGROUP */ KEEPS, /* KEEPS */ + LNBREAK, /* LNBREAK */ + VERTWS, /* VERTWS */ + NVERTWS, /* NVERTWS */ + HORIZWS, /* HORIZWS */ + NHORIZWS, /* NHORIZWS */ NOTHING, /* OPTIMIZED */ PSEUDO, /* PSEUDO */ /* ------------ States ------------- */ @@ -286,8 +360,12 @@ static const U8 regarglen[] = { 0, /* SEOL */ 0, /* BOUND */ 0, /* BOUNDL */ + 0, /* BOUNDU */ + 0, /* BOUNDA */ 0, /* NBOUND */ 0, /* NBOUNDL */ + 0, /* NBOUNDU */ + 0, /* NBOUNDA */ 0, /* GPOS */ 0, /* REG_ANY */ 0, /* SANY */ @@ -295,22 +373,46 @@ static const U8 regarglen[] = { 0, /* ANYOF */ 0, /* ALNUM */ 0, /* ALNUML */ + 0, /* ALNUMU */ + 0, /* ALNUMA */ 0, /* NALNUM */ 0, /* NALNUML */ + 0, /* NALNUMU */ + 0, /* NALNUMA */ 0, /* SPACE */ 0, /* SPACEL */ + 0, /* SPACEU */ + 0, /* SPACEA */ 0, /* NSPACE */ 0, /* NSPACEL */ + 0, /* NSPACEU */ + 0, /* NSPACEA */ 0, /* DIGIT */ 0, /* DIGITL */ + 0, /* PLACEHOLDER1 */ + 0, /* DIGITA */ 0, /* NDIGIT */ 0, /* NDIGITL */ + 0, /* PLACEHOLDER2 */ + 0, /* NDIGITA */ + 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, /* EXACTF */ 0, /* EXACTFL */ + 0, /* EXACTFU */ + 0, /* EXACTFA */ + 0, /* EXACTFU_SS */ + 0, /* EXACTFU_TRICKYFOLD */ 0, /* NOTHING */ 0, /* TAIL */ 0, /* STAR */ @@ -325,6 +427,13 @@ 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), /* IFMATCH */ EXTRA_SIZE(struct regnode_1), /* UNLESSM */ EXTRA_SIZE(struct regnode_1), /* SUSPEND */ @@ -342,22 +451,24 @@ static const U8 regarglen[] = { 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), /* 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, /* VERTWS */ + 0, /* NVERTWS */ + 0, /* HORIZWS */ + 0, /* NHORIZWS */ 0, /* OPTIMIZED */ 0, /* PSEUDO */ }; @@ -376,8 +487,12 @@ static const char reg_off_by_arg[] = { 0, /* SEOL */ 0, /* BOUND */ 0, /* BOUNDL */ + 0, /* BOUNDU */ + 0, /* BOUNDA */ 0, /* NBOUND */ 0, /* NBOUNDL */ + 0, /* NBOUNDU */ + 0, /* NBOUNDA */ 0, /* GPOS */ 0, /* REG_ANY */ 0, /* SANY */ @@ -385,22 +500,46 @@ static const char reg_off_by_arg[] = { 0, /* ANYOF */ 0, /* ALNUM */ 0, /* ALNUML */ + 0, /* ALNUMU */ + 0, /* ALNUMA */ 0, /* NALNUM */ 0, /* NALNUML */ + 0, /* NALNUMU */ + 0, /* NALNUMA */ 0, /* SPACE */ 0, /* SPACEL */ + 0, /* SPACEU */ + 0, /* SPACEA */ 0, /* NSPACE */ 0, /* NSPACEL */ + 0, /* NSPACEU */ + 0, /* NSPACEA */ 0, /* DIGIT */ 0, /* DIGITL */ + 0, /* PLACEHOLDER1 */ + 0, /* DIGITA */ 0, /* NDIGIT */ 0, /* NDIGITL */ + 0, /* PLACEHOLDER2 */ + 0, /* NDIGITA */ + 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, /* EXACTF */ 0, /* EXACTFL */ + 0, /* EXACTFU */ + 0, /* EXACTFA */ + 0, /* EXACTFU_SS */ + 0, /* EXACTFU_TRICKYFOLD */ 0, /* NOTHING */ 0, /* TAIL */ 0, /* STAR */ @@ -415,6 +554,13 @@ 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 */ 2, /* IFMATCH */ 2, /* UNLESSM */ 1, /* SUSPEND */ @@ -432,9 +578,6 @@ static const char reg_off_by_arg[] = { 0, /* AHOCORASICKC */ 0, /* GOSUB */ 0, /* GOSTART */ - 0, /* NREF */ - 0, /* NREFF */ - 0, /* NREFFL */ 0, /* NGROUPP */ 0, /* INSUBP */ 0, /* DEFINEP */ @@ -448,6 +591,11 @@ static const char reg_off_by_arg[] = { 0, /* COMMIT */ 0, /* CUTGROUP */ 0, /* KEEPS */ + 0, /* LNBREAK */ + 0, /* VERTWS */ + 0, /* NVERTWS */ + 0, /* HORIZWS */ + 0, /* NHORIZWS */ 0, /* OPTIMIZED */ 0, /* PSEUDO */ }; @@ -471,80 +619,117 @@ EXTCONST char * const PL_reg_name[] = { "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 */ + "BOUNDU", /* 0x0b */ + "BOUNDA", /* 0x0c */ + "NBOUND", /* 0x0d */ + "NBOUNDL", /* 0x0e */ + "NBOUNDU", /* 0x0f */ + "NBOUNDA", /* 0x10 */ + "GPOS", /* 0x11 */ + "REG_ANY", /* 0x12 */ + "SANY", /* 0x13 */ + "CANY", /* 0x14 */ + "ANYOF", /* 0x15 */ + "ALNUM", /* 0x16 */ + "ALNUML", /* 0x17 */ + "ALNUMU", /* 0x18 */ + "ALNUMA", /* 0x19 */ + "NALNUM", /* 0x1a */ + "NALNUML", /* 0x1b */ + "NALNUMU", /* 0x1c */ + "NALNUMA", /* 0x1d */ + "SPACE", /* 0x1e */ + "SPACEL", /* 0x1f */ + "SPACEU", /* 0x20 */ + "SPACEA", /* 0x21 */ + "NSPACE", /* 0x22 */ + "NSPACEL", /* 0x23 */ + "NSPACEU", /* 0x24 */ + "NSPACEA", /* 0x25 */ + "DIGIT", /* 0x26 */ + "DIGITL", /* 0x27 */ + "PLACEHOLDER1", /* 0x28 */ + "DIGITA", /* 0x29 */ + "NDIGIT", /* 0x2a */ + "NDIGITL", /* 0x2b */ + "PLACEHOLDER2", /* 0x2c */ + "NDIGITA", /* 0x2d */ + "POSIXD", /* 0x2e */ + "POSIXL", /* 0x2f */ + "POSIXU", /* 0x30 */ + "POSIXA", /* 0x31 */ + "NPOSIXD", /* 0x32 */ + "NPOSIXL", /* 0x33 */ + "NPOSIXU", /* 0x34 */ + "NPOSIXA", /* 0x35 */ + "CLUMP", /* 0x36 */ + "BRANCH", /* 0x37 */ + "BACK", /* 0x38 */ + "EXACT", /* 0x39 */ + "EXACTF", /* 0x3a */ + "EXACTFL", /* 0x3b */ + "EXACTFU", /* 0x3c */ + "EXACTFA", /* 0x3d */ + "EXACTFU_SS", /* 0x3e */ + "EXACTFU_TRICKYFOLD", /* 0x3f */ + "NOTHING", /* 0x40 */ + "TAIL", /* 0x41 */ + "STAR", /* 0x42 */ + "PLUS", /* 0x43 */ + "CURLY", /* 0x44 */ + "CURLYN", /* 0x45 */ + "CURLYM", /* 0x46 */ + "CURLYX", /* 0x47 */ + "WHILEM", /* 0x48 */ + "OPEN", /* 0x49 */ + "CLOSE", /* 0x4a */ + "REF", /* 0x4b */ + "REFF", /* 0x4c */ + "REFFL", /* 0x4d */ + "REFFU", /* 0x4e */ + "REFFA", /* 0x4f */ + "NREF", /* 0x50 */ + "NREFF", /* 0x51 */ + "NREFFL", /* 0x52 */ + "NREFFU", /* 0x53 */ + "NREFFA", /* 0x54 */ + "IFMATCH", /* 0x55 */ + "UNLESSM", /* 0x56 */ + "SUSPEND", /* 0x57 */ + "IFTHEN", /* 0x58 */ + "GROUPP", /* 0x59 */ + "LONGJMP", /* 0x5a */ + "BRANCHJ", /* 0x5b */ + "EVAL", /* 0x5c */ + "MINMOD", /* 0x5d */ + "LOGICAL", /* 0x5e */ + "RENUM", /* 0x5f */ + "TRIE", /* 0x60 */ + "TRIEC", /* 0x61 */ + "AHOCORASICK", /* 0x62 */ + "AHOCORASICKC", /* 0x63 */ + "GOSUB", /* 0x64 */ + "GOSTART", /* 0x65 */ + "NGROUPP", /* 0x66 */ + "INSUBP", /* 0x67 */ + "DEFINEP", /* 0x68 */ + "ENDLIKE", /* 0x69 */ + "OPFAIL", /* 0x6a */ + "ACCEPT", /* 0x6b */ + "VERB", /* 0x6c */ + "PRUNE", /* 0x6d */ + "MARKPOINT", /* 0x6e */ + "SKIP", /* 0x6f */ + "COMMIT", /* 0x70 */ + "CUTGROUP", /* 0x71 */ + "KEEPS", /* 0x72 */ + "LNBREAK", /* 0x73 */ + "VERTWS", /* 0x74 */ + "NVERTWS", /* 0x75 */ + "HORIZWS", /* 0x76 */ + "NHORIZWS", /* 0x77 */ + "OPTIMIZED", /* 0x78 */ + "PSEUDO", /* 0x79 */ /* ------------ States ------------- */ "TRIE_next", /* REGNODE_MAX +0x01 */ "TRIE_next_fail", /* REGNODE_MAX +0x02 */ @@ -589,4 +774,93 @@ EXTCONST char * const PL_reg_name[] = { }; #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: 11011111111111111111111011111111 */ + "MULTILINE", /* 0x00000001 */ + "SINGLELINE", /* 0x00000002 */ + "FOLD", /* 0x00000004 */ + "EXTENDED", /* 0x00000008 */ + "KEEPCOPY", /* 0x00000010 */ + "CHARSET", /* 0x000000e0 */ + "CHARSET", /* 0x000000e0 */ + "CHARSET", /* 0x000000e0 */ + "UNUSED_BIT_8", /* 0x00000100 */ + "ANCH_BOL", /* 0x00000200 */ + "ANCH_MBOL", /* 0x00000400 */ + "ANCH_SBOL", /* 0x00000800 */ + "ANCH_GPOS", /* 0x00001000 */ + "GPOS_SEEN", /* 0x00002000 */ + "GPOS_FLOAT", /* 0x00004000 */ + "LOOKBEHIND_SEEN", /* 0x00008000 */ + "EVAL_SEEN", /* 0x00010000 */ + "CANY_SEEN", /* 0x00020000 */ + "NOSCAN", /* 0x00040000 */ + "CHECK_ALL", /* 0x00080000 */ + "MATCH_UTF8", /* 0x00100000 */ + "USE_INTUIT_NOML", /* 0x00200000 */ + "USE_INTUIT_ML", /* 0x00400000 */ + "INTUIT_TAIL", /* 0x00800000 */ + "MODIFIES_VARS", /* 0x01000000 */ + "COPY_DONE", /* 0x02000000 */ + "TAINTED_SEEN", /* 0x04000000 */ + "TAINTED", /* 0x08000000 */ + "START_ONLY", /* 0x10000000 */ + "UNUSED_BIT_29", /* 0x20000000 */ + "WHITE", /* 0x40000000 */ + "NULL", /* 0x80000000 */ +}; +#endif /* DOINIT */ + +/* 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, BACK, STAR, PLUS, CURLY, CURLYN, CURLYM, CURLYX, WHILEM, + REF, REFF, REFFL, REFFU, REFFA, NREF, NREFF, NREFFL, NREFFU, NREFFA, + SUSPEND, IFTHEN, BRANCHJ, + 0 +}; +#endif /* DOINIT */ + +#ifndef DOINIT +EXTCONST U8 PL_varies_bitmask[]; +#else +EXTCONST U8 PL_varies_bitmask[] = { + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xC0, 0x01, 0xFC, 0xF9, 0x9F, 0x09, 0x00, 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, CANY, ANYOF, ALNUM, ALNUML, ALNUMU, ALNUMA, NALNUM, + NALNUML, NALNUMU, NALNUMA, SPACE, SPACEL, SPACEU, SPACEA, NSPACE, + NSPACEL, NSPACEU, NSPACEA, DIGIT, DIGITL, DIGITA, NDIGIT, NDIGITL, + NDIGITA, POSIXD, POSIXL, POSIXU, POSIXA, NPOSIXD, NPOSIXL, NPOSIXU, + NPOSIXA, VERTWS, NVERTWS, HORIZWS, NHORIZWS, + 0 +}; +#endif /* DOINIT */ + +#ifndef DOINIT +EXTCONST U8 PL_simple_bitmask[]; +#else +EXTCONST U8 PL_simple_bitmask[] = { + 0x00, 0x00, 0xFC, 0xFF, 0xFF, 0xEE, 0x3F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xF0, 0x00 +}; +#endif /* DOINIT */ + /* ex: set ro: */