X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/3615ea5819e869b314f723b0ce28dfb1d80017ef..8b405cba157a912a7bf5dcc8b16a5f63c220b328:/regnodes.h diff --git a/regnodes.h b/regnodes.h index 42f2452..937dd57 100644 --- a/regnodes.h +++ b/regnodes.h @@ -6,132 +6,102 @@ /* Regops and State definitions */ -#define REGNODE_MAX 122 -#define REGMATCH_STATE_MAX 162 +#define REGNODE_MAX 90 +#define REGMATCH_STATE_MAX 130 #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 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 ANYOFV 22 /* 0x16 Match character in (or not in) this class, can match-multiple chars */ -#define ALNUM 23 /* 0x17 Match any alphanumeric character using native charset semantics for non-utf8 */ -#define ALNUML 24 /* 0x18 Match any alphanumeric char in locale */ -#define ALNUMU 25 /* 0x19 Match any alphanumeric char using Unicode semantics */ -#define ALNUMA 26 /* 0x1a Match [A-Za-z_0-9] */ -#define NALNUM 27 /* 0x1b Match any non-alphanumeric character using native charset semantics for non-utf8 */ -#define NALNUML 28 /* 0x1c Match any non-alphanumeric char in locale */ -#define NALNUMU 29 /* 0x1d Match any non-alphanumeric char using Unicode semantics */ -#define NALNUMA 30 /* 0x1e Match [^A-Za-z_0-9] */ -#define SPACE 31 /* 0x1f Match any whitespace character using native charset semantics for non-utf8 */ -#define SPACEL 32 /* 0x20 Match any whitespace char in locale */ -#define SPACEU 33 /* 0x21 Match any whitespace char using Unicode semantics */ -#define SPACEA 34 /* 0x22 Match [ \t\n\f\r] */ -#define NSPACE 35 /* 0x23 Match any non-whitespace character using native charset semantics for non-utf8 */ -#define NSPACEL 36 /* 0x24 Match any non-whitespace char in locale */ -#define NSPACEU 37 /* 0x25 Match any non-whitespace char using Unicode semantics */ -#define NSPACEA 38 /* 0x26 Match [^ \t\n\f\r] */ -#define DIGIT 39 /* 0x27 Match any numeric character using native charset semantics for non-utf8 */ -#define DIGITL 40 /* 0x28 Match any numeric character in locale */ -#define PLACEHOLDER1 41 /* 0x29 placeholder for missing DIGITU */ -#define DIGITA 42 /* 0x2a Match [0-9] */ -#define NDIGIT 43 /* 0x2b Match any non-numeric character using native charset semantics for non-utf8 */ -#define NDIGITL 44 /* 0x2c Match any non-numeric character in locale */ -#define PLACEHOLDER2 45 /* 0x2d placeholder for missing NDIGITU */ -#define NDIGITA 46 /* 0x2e Match [^0-9] */ -#define POSIXD 47 /* 0x2f currently unused except as a placeholder */ -#define POSIXL 48 /* 0x30 currently unused except as a placeholder */ -#define POSIXU 49 /* 0x31 currently unused except as a placeholder */ -#define POSIXA 50 /* 0x32 Some [[:class:]] under /a; the FLAGS field gives which one */ -#define NPOSIXD 51 /* 0x33 currently unused except as a placeholder */ -#define NPOSIXL 52 /* 0x34 currently unused except as a placeholder */ -#define NPOSIXU 53 /* 0x35 currently unused except as a placeholder */ -#define NPOSIXA 54 /* 0x36 complement of POSIXA, [[:^class:]] */ -#define CLUMP 55 /* 0x37 Match any extended grapheme cluster sequence */ -#define BRANCH 56 /* 0x38 Match this alternative, or the next... */ -#define BACK 57 /* 0x39 Match "", "next" ptr points backward. */ -#define EXACT 58 /* 0x3a Match this string (preceded by length). */ -#define EXACTF 59 /* 0x3b Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ -#define EXACTFL 60 /* 0x3c Match this string (not guaranteed to be folded) using /il rules (w/len). */ -#define EXACTFU 61 /* 0x3d 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 62 /* 0x3e Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ -#define EXACTFU_SS 63 /* 0x3f 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 64 /* 0x40 Match this folded UTF-8 string using /iu rules */ -#define NOTHING 65 /* 0x41 Match empty string. */ -#define TAIL 66 /* 0x42 Match empty string. Can jump here from outside. */ -#define STAR 67 /* 0x43 Match this (simple) thing 0 or more times. */ -#define PLUS 68 /* 0x44 Match this (simple) thing 1 or more times. */ -#define CURLY 69 /* 0x45 Match this simple thing {n,m} times. */ -#define CURLYN 70 /* 0x46 Capture next-after-this simple thing */ -#define CURLYM 71 /* 0x47 Capture this medium-complex thing {n,m} times. */ -#define CURLYX 72 /* 0x48 Match this complex thing {n,m} times. */ -#define WHILEM 73 /* 0x49 Do curly processing and see if rest matches. */ -#define OPEN 74 /* 0x4a Mark this point in input as start of */ -#define CLOSE 75 /* 0x4b Analogous to OPEN. */ -#define REF 76 /* 0x4c Match some already matched string */ -#define REFF 77 /* 0x4d Match already matched string, folded using native charset semantics for non-utf8 */ -#define REFFL 78 /* 0x4e Match already matched string, folded in loc. */ -#define REFFU 79 /* 0x4f Match already matched string, folded using unicode semantics for non-utf8 */ -#define REFFA 80 /* 0x50 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ -#define NREF 81 /* 0x51 Match some already matched string */ -#define NREFF 82 /* 0x52 Match already matched string, folded using native charset semantics for non-utf8 */ -#define NREFFL 83 /* 0x53 Match already matched string, folded in loc. */ -#define NREFFU 84 /* 0x54 Match already matched string, folded using unicode semantics for non-utf8 */ -#define NREFFA 85 /* 0x55 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ -#define IFMATCH 86 /* 0x56 Succeeds if the following matches. */ -#define UNLESSM 87 /* 0x57 Fails if the following matches. */ -#define SUSPEND 88 /* 0x58 "Independent" sub-RE. */ -#define IFTHEN 89 /* 0x59 Switch, should be preceded by switcher . */ -#define GROUPP 90 /* 0x5a Whether the group matched. */ -#define LONGJMP 91 /* 0x5b Jump far away. */ -#define BRANCHJ 92 /* 0x5c BRANCH with long offset. */ -#define EVAL 93 /* 0x5d Execute some Perl code. */ -#define MINMOD 94 /* 0x5e Next operator is not greedy. */ -#define LOGICAL 95 /* 0x5f Next opcode should set the flag only. */ -#define RENUM 96 /* 0x60 Group with independently numbered parens. */ -#define TRIE 97 /* 0x61 Match many EXACT(F[ALU]?)? at once. flags==type */ -#define TRIEC 98 /* 0x62 Same as TRIE, but with embedded charclass data */ -#define AHOCORASICK 99 /* 0x63 Aho Corasick stclass. flags==type */ -#define AHOCORASICKC 100 /* 0x64 Same as AHOCORASICK, but with embedded charclass data */ -#define GOSUB 101 /* 0x65 recurse to paren arg1 at (signed) ofs arg2 */ -#define GOSTART 102 /* 0x66 recurse to start of pattern */ -#define NGROUPP 103 /* 0x67 Whether the group matched. */ -#define INSUBP 104 /* 0x68 Whether we are in a specific recurse. */ -#define DEFINEP 105 /* 0x69 Never execute directly. */ -#define ENDLIKE 106 /* 0x6a Used only for the type field of verbs */ -#define OPFAIL 107 /* 0x6b Same as (?!) */ -#define ACCEPT 108 /* 0x6c Accepts the current matched string. */ -#define VERB 109 /* 0x6d Used only for the type field of verbs */ -#define PRUNE 110 /* 0x6e Pattern fails at this startpoint if no-backtracking through this */ -#define MARKPOINT 111 /* 0x6f Push the current location for rollback by cut. */ -#define SKIP 112 /* 0x70 On failure skip forward (to the mark) before retrying */ -#define COMMIT 113 /* 0x71 Pattern fails outright if backtracking through this */ -#define CUTGROUP 114 /* 0x72 On failure go to the next alternation in the group */ -#define KEEPS 115 /* 0x73 $& begins here. */ -#define LNBREAK 116 /* 0x74 generic newline pattern */ -#define VERTWS 117 /* 0x75 vertical whitespace (Perl 6) */ -#define NVERTWS 118 /* 0x76 not vertical whitespace (Perl 6) */ -#define HORIZWS 119 /* 0x77 horizontal whitespace (Perl 6) */ -#define NHORIZWS 120 /* 0x78 not horizontal whitespace (Perl 6) */ -#define OPTIMIZED 121 /* 0x79 Placeholder for dump. */ -#define PSEUDO 122 /* 0x7a Pseudo opcode for internal use. */ +#define SBOL 2 /* 0x02 Match "" at beginning of line: /^/, /\A/ */ +#define BOL 2 /* 0x02 type alias */ +#define MBOL 3 /* 0x03 Same, assuming multiline: /^/m */ +#define SEOL 4 /* 0x04 Match "" at end of line: /$/ */ +#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 Match "" at any word boundary using native charset rules for non-utf8 */ +#define BOUNDL 9 /* 0x09 Match "" at any locale word boundary */ +#define BOUNDU 10 /* 0x0a Match "" at any word boundary using Unicode rules */ +#define BOUNDA 11 /* 0x0b Match "" at any word boundary using ASCII rules */ +#define NBOUND 12 /* 0x0c Match "" at any word non-boundary using native charset rules for non-utf8 */ +#define NBOUNDL 13 /* 0x0d Match "" at any locale word non-boundary */ +#define NBOUNDU 14 /* 0x0e Match "" at any word non-boundary using Unicode rules */ +#define NBOUNDA 15 /* 0x0f Match "" at any word non-boundary using ASCII rules */ +#define REG_ANY 16 /* 0x10 Match any one character (except newline). */ +#define SANY 17 /* 0x11 Match any one character. */ +#define CANY 18 /* 0x12 Match any one byte. */ +#define ANYOF 19 /* 0x13 Match character in (or not in) this class, single char match only */ +#define POSIXD 20 /* 0x14 Some [[:class:]] under /d; the FLAGS field gives which one */ +#define POSIXL 21 /* 0x15 Some [[:class:]] under /l; the FLAGS field gives which one */ +#define POSIXU 22 /* 0x16 Some [[:class:]] under /u; the FLAGS field gives which one */ +#define POSIXA 23 /* 0x17 Some [[:class:]] under /a; the FLAGS field gives which one */ +#define NPOSIXD 24 /* 0x18 complement of POSIXD, [[:^class:]] */ +#define NPOSIXL 25 /* 0x19 complement of POSIXL, [[:^class:]] */ +#define NPOSIXU 26 /* 0x1a complement of POSIXU, [[:^class:]] */ +#define NPOSIXA 27 /* 0x1b complement of POSIXA, [[:^class:]] */ +#define CLUMP 28 /* 0x1c Match any extended grapheme cluster sequence */ +#define BRANCH 29 /* 0x1d Match this alternative, or the next... */ +#define EXACT 30 /* 0x1e Match this string (preceded by length). */ +#define EXACTF 31 /* 0x1f Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ +#define EXACTFL 32 /* 0x20 Match this string (not guaranteed to be folded) using /il rules (w/len). */ +#define EXACTFU 33 /* 0x21 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 34 /* 0x22 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ +#define EXACTFU_SS 35 /* 0x23 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 36 /* 0x24 Match this string (which is not trie-able; not guaranteed to be folded) using /iaa rules (w/len). */ +#define NOTHING 37 /* 0x25 Match empty string. */ +#define TAIL 38 /* 0x26 Match empty string. Can jump here from outside. */ +#define STAR 39 /* 0x27 Match this (simple) thing 0 or more times. */ +#define PLUS 40 /* 0x28 Match this (simple) thing 1 or more times. */ +#define CURLY 41 /* 0x29 Match this simple thing {n,m} times. */ +#define CURLYN 42 /* 0x2a Capture next-after-this simple thing */ +#define CURLYM 43 /* 0x2b Capture this medium-complex thing {n,m} times. */ +#define CURLYX 44 /* 0x2c Match this complex thing {n,m} times. */ +#define WHILEM 45 /* 0x2d Do curly processing and see if rest matches. */ +#define OPEN 46 /* 0x2e Mark this point in input as start of #n. */ +#define CLOSE 47 /* 0x2f Analogous to OPEN. */ +#define REF 48 /* 0x30 Match some already matched string */ +#define REFF 49 /* 0x31 Match already matched string, folded using native charset rules for non-utf8 */ +#define REFFL 50 /* 0x32 Match already matched string, folded in loc. */ +#define REFFU 51 /* 0x33 Match already matched string, folded using unicode rules for non-utf8 */ +#define REFFA 52 /* 0x34 Match already matched string, folded using unicode rules for non-utf8, no mixing ASCII, non-ASCII */ +#define NREF 53 /* 0x35 Match some already matched string */ +#define NREFF 54 /* 0x36 Match already matched string, folded using native charset rules for non-utf8 */ +#define NREFFL 55 /* 0x37 Match already matched string, folded in loc. */ +#define NREFFU 56 /* 0x38 Match already matched string, folded using unicode rules for non-utf8 */ +#define NREFFA 57 /* 0x39 Match already matched string, folded using unicode rules for non-utf8, no mixing ASCII, non-ASCII */ +#define LONGJMP 58 /* 0x3a Jump far away. */ +#define BRANCHJ 59 /* 0x3b BRANCH with long offset. */ +#define IFMATCH 60 /* 0x3c Succeeds if the following matches. */ +#define UNLESSM 61 /* 0x3d Fails if the following matches. */ +#define SUSPEND 62 /* 0x3e "Independent" sub-RE. */ +#define IFTHEN 63 /* 0x3f Switch, should be preceded by switcher. */ +#define GROUPP 64 /* 0x40 Whether the group matched. */ +#define EVAL 65 /* 0x41 Execute some Perl code. */ +#define MINMOD 66 /* 0x42 Next operator is not greedy. */ +#define LOGICAL 67 /* 0x43 Next opcode should set the flag only. */ +#define RENUM 68 /* 0x44 Group with independently numbered parens. */ +#define TRIE 69 /* 0x45 Match many EXACT(F[ALU]?)? at once. flags==type */ +#define TRIEC 70 /* 0x46 Same as TRIE, but with embedded charclass data */ +#define AHOCORASICK 71 /* 0x47 Aho Corasick stclass. flags==type */ +#define AHOCORASICKC 72 /* 0x48 Same as AHOCORASICK, but with embedded charclass data */ +#define GOSUB 73 /* 0x49 recurse to paren arg1 at (signed) ofs arg2 */ +#define GOSTART 74 /* 0x4a recurse to start of pattern */ +#define NGROUPP 75 /* 0x4b Whether the group matched. */ +#define INSUBP 76 /* 0x4c Whether we are in a specific recurse. */ +#define DEFINEP 77 /* 0x4d Never execute directly. */ +#define ENDLIKE 78 /* 0x4e Used only for the type field of verbs */ +#define OPFAIL 79 /* 0x4f Same as (?!) */ +#define ACCEPT 80 /* 0x50 Accepts the current matched string. */ +#define VERB 81 /* 0x51 Used only for the type field of verbs */ +#define PRUNE 82 /* 0x52 Pattern fails at this startpoint if no-backtracking through this */ +#define MARKPOINT 83 /* 0x53 Push the current location for rollback by cut. */ +#define SKIP 84 /* 0x54 On failure skip forward (to the mark) before retrying */ +#define COMMIT 85 /* 0x55 Pattern fails outright if backtracking through this */ +#define CUTGROUP 86 /* 0x56 On failure go to the next alternation in the group */ +#define KEEPS 87 /* 0x57 $& begins here. */ +#define LNBREAK 88 /* 0x58 generic newline pattern */ +#define OPTIMIZED 89 /* 0x59 Placeholder for dump. */ +#define PSEUDO 90 /* 0x5a 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 */ @@ -182,13 +152,12 @@ 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 */ @@ -197,54 +166,27 @@ EXTCONST U8 PL_regkind[] = { NBOUND, /* NBOUNDL */ NBOUND, /* NBOUNDU */ NBOUND, /* NBOUNDA */ - GPOS, /* GPOS */ REG_ANY, /* REG_ANY */ REG_ANY, /* SANY */ REG_ANY, /* CANY */ ANYOF, /* ANYOF */ - ANYOF, /* ANYOFV */ - 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 */ - POSIXD, /* NPOSIXD */ - POSIXD, /* NPOSIXL */ - POSIXD, /* NPOSIXU */ - POSIXD, /* NPOSIXA */ + 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 */ + EXACT, /* EXACTFA_NO_TRIE */ NOTHING, /* NOTHING */ NOTHING, /* TAIL */ STAR, /* STAR */ @@ -266,13 +208,13 @@ EXTCONST U8 PL_regkind[] = { 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 */ @@ -297,10 +239,6 @@ EXTCONST U8 PL_regkind[] = { VERB, /* CUTGROUP */ KEEPS, /* KEEPS */ LNBREAK, /* LNBREAK */ - VERTWS, /* VERTWS */ - NVERTWS, /* NVERTWS */ - HORIZWS, /* HORIZWS */ - NHORIZWS, /* NHORIZWS */ NOTHING, /* OPTIMIZED */ PSEUDO, /* PSEUDO */ /* ------------ States ------------- */ @@ -353,13 +291,12 @@ EXTCONST U8 PL_regkind[] = { 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 */ @@ -368,36 +305,10 @@ static const U8 regarglen[] = { 0, /* NBOUNDL */ 0, /* NBOUNDU */ 0, /* NBOUNDA */ - 0, /* GPOS */ 0, /* REG_ANY */ 0, /* SANY */ 0, /* CANY */ - 0, /* ANYOF */ - 0, /* ANYOFV */ - 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 */ + EXTRA_SIZE(struct regnode_1), /* ANYOF */ 0, /* POSIXD */ 0, /* POSIXL */ 0, /* POSIXU */ @@ -408,14 +319,13 @@ static const U8 regarglen[] = { 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, /* EXACTFA_NO_TRIE */ 0, /* NOTHING */ 0, /* TAIL */ 0, /* STAR */ @@ -437,14 +347,14 @@ static const U8 regarglen[] = { 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 */ @@ -468,10 +378,6 @@ static const U8 regarglen[] = { EXTRA_SIZE(struct regnode_1), /* CUTGROUP */ 0, /* KEEPS */ 0, /* LNBREAK */ - 0, /* VERTWS */ - 0, /* NVERTWS */ - 0, /* HORIZWS */ - 0, /* NHORIZWS */ 0, /* OPTIMIZED */ 0, /* PSEUDO */ }; @@ -481,13 +387,12 @@ 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 */ @@ -496,36 +401,10 @@ static const char reg_off_by_arg[] = { 0, /* NBOUNDL */ 0, /* NBOUNDU */ 0, /* NBOUNDA */ - 0, /* GPOS */ 0, /* REG_ANY */ 0, /* SANY */ 0, /* CANY */ 0, /* ANYOF */ - 0, /* ANYOFV */ - 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 */ @@ -536,14 +415,13 @@ static const char reg_off_by_arg[] = { 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, /* EXACTFA_NO_TRIE */ 0, /* NOTHING */ 0, /* TAIL */ 0, /* STAR */ @@ -565,13 +443,13 @@ static const char reg_off_by_arg[] = { 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 */ @@ -596,10 +474,6 @@ static const char reg_off_by_arg[] = { 0, /* CUTGROUP */ 0, /* KEEPS */ 0, /* LNBREAK */ - 0, /* VERTWS */ - 0, /* NVERTWS */ - 0, /* HORIZWS */ - 0, /* NHORIZWS */ 0, /* OPTIMIZED */ 0, /* PSEUDO */ }; @@ -614,127 +488,95 @@ EXTCONST char * PL_reg_name[]; 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 */ - "BOUNDU", /* 0x0b */ - "BOUNDA", /* 0x0c */ - "NBOUND", /* 0x0d */ - "NBOUNDL", /* 0x0e */ - "NBOUNDU", /* 0x0f */ - "NBOUNDA", /* 0x10 */ - "GPOS", /* 0x11 */ - "REG_ANY", /* 0x12 */ - "SANY", /* 0x13 */ - "CANY", /* 0x14 */ - "ANYOF", /* 0x15 */ - "ANYOFV", /* 0x16 */ - "ALNUM", /* 0x17 */ - "ALNUML", /* 0x18 */ - "ALNUMU", /* 0x19 */ - "ALNUMA", /* 0x1a */ - "NALNUM", /* 0x1b */ - "NALNUML", /* 0x1c */ - "NALNUMU", /* 0x1d */ - "NALNUMA", /* 0x1e */ - "SPACE", /* 0x1f */ - "SPACEL", /* 0x20 */ - "SPACEU", /* 0x21 */ - "SPACEA", /* 0x22 */ - "NSPACE", /* 0x23 */ - "NSPACEL", /* 0x24 */ - "NSPACEU", /* 0x25 */ - "NSPACEA", /* 0x26 */ - "DIGIT", /* 0x27 */ - "DIGITL", /* 0x28 */ - "PLACEHOLDER1", /* 0x29 */ - "DIGITA", /* 0x2a */ - "NDIGIT", /* 0x2b */ - "NDIGITL", /* 0x2c */ - "PLACEHOLDER2", /* 0x2d */ - "NDIGITA", /* 0x2e */ - "POSIXD", /* 0x2f */ - "POSIXL", /* 0x30 */ - "POSIXU", /* 0x31 */ - "POSIXA", /* 0x32 */ - "NPOSIXD", /* 0x33 */ - "NPOSIXL", /* 0x34 */ - "NPOSIXU", /* 0x35 */ - "NPOSIXA", /* 0x36 */ - "CLUMP", /* 0x37 */ - "BRANCH", /* 0x38 */ - "BACK", /* 0x39 */ - "EXACT", /* 0x3a */ - "EXACTF", /* 0x3b */ - "EXACTFL", /* 0x3c */ - "EXACTFU", /* 0x3d */ - "EXACTFA", /* 0x3e */ - "EXACTFU_SS", /* 0x3f */ - "EXACTFU_TRICKYFOLD", /* 0x40 */ - "NOTHING", /* 0x41 */ - "TAIL", /* 0x42 */ - "STAR", /* 0x43 */ - "PLUS", /* 0x44 */ - "CURLY", /* 0x45 */ - "CURLYN", /* 0x46 */ - "CURLYM", /* 0x47 */ - "CURLYX", /* 0x48 */ - "WHILEM", /* 0x49 */ - "OPEN", /* 0x4a */ - "CLOSE", /* 0x4b */ - "REF", /* 0x4c */ - "REFF", /* 0x4d */ - "REFFL", /* 0x4e */ - "REFFU", /* 0x4f */ - "REFFA", /* 0x50 */ - "NREF", /* 0x51 */ - "NREFF", /* 0x52 */ - "NREFFL", /* 0x53 */ - "NREFFU", /* 0x54 */ - "NREFFA", /* 0x55 */ - "IFMATCH", /* 0x56 */ - "UNLESSM", /* 0x57 */ - "SUSPEND", /* 0x58 */ - "IFTHEN", /* 0x59 */ - "GROUPP", /* 0x5a */ - "LONGJMP", /* 0x5b */ - "BRANCHJ", /* 0x5c */ - "EVAL", /* 0x5d */ - "MINMOD", /* 0x5e */ - "LOGICAL", /* 0x5f */ - "RENUM", /* 0x60 */ - "TRIE", /* 0x61 */ - "TRIEC", /* 0x62 */ - "AHOCORASICK", /* 0x63 */ - "AHOCORASICKC", /* 0x64 */ - "GOSUB", /* 0x65 */ - "GOSTART", /* 0x66 */ - "NGROUPP", /* 0x67 */ - "INSUBP", /* 0x68 */ - "DEFINEP", /* 0x69 */ - "ENDLIKE", /* 0x6a */ - "OPFAIL", /* 0x6b */ - "ACCEPT", /* 0x6c */ - "VERB", /* 0x6d */ - "PRUNE", /* 0x6e */ - "MARKPOINT", /* 0x6f */ - "SKIP", /* 0x70 */ - "COMMIT", /* 0x71 */ - "CUTGROUP", /* 0x72 */ - "KEEPS", /* 0x73 */ - "LNBREAK", /* 0x74 */ - "VERTWS", /* 0x75 */ - "NVERTWS", /* 0x76 */ - "HORIZWS", /* 0x77 */ - "NHORIZWS", /* 0x78 */ - "OPTIMIZED", /* 0x79 */ - "PSEUDO", /* 0x7a */ + "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 */ + "CANY", /* 0x12 */ + "ANYOF", /* 0x13 */ + "POSIXD", /* 0x14 */ + "POSIXL", /* 0x15 */ + "POSIXU", /* 0x16 */ + "POSIXA", /* 0x17 */ + "NPOSIXD", /* 0x18 */ + "NPOSIXL", /* 0x19 */ + "NPOSIXU", /* 0x1a */ + "NPOSIXA", /* 0x1b */ + "CLUMP", /* 0x1c */ + "BRANCH", /* 0x1d */ + "EXACT", /* 0x1e */ + "EXACTF", /* 0x1f */ + "EXACTFL", /* 0x20 */ + "EXACTFU", /* 0x21 */ + "EXACTFA", /* 0x22 */ + "EXACTFU_SS", /* 0x23 */ + "EXACTFA_NO_TRIE", /* 0x24 */ + "NOTHING", /* 0x25 */ + "TAIL", /* 0x26 */ + "STAR", /* 0x27 */ + "PLUS", /* 0x28 */ + "CURLY", /* 0x29 */ + "CURLYN", /* 0x2a */ + "CURLYM", /* 0x2b */ + "CURLYX", /* 0x2c */ + "WHILEM", /* 0x2d */ + "OPEN", /* 0x2e */ + "CLOSE", /* 0x2f */ + "REF", /* 0x30 */ + "REFF", /* 0x31 */ + "REFFL", /* 0x32 */ + "REFFU", /* 0x33 */ + "REFFA", /* 0x34 */ + "NREF", /* 0x35 */ + "NREFF", /* 0x36 */ + "NREFFL", /* 0x37 */ + "NREFFU", /* 0x38 */ + "NREFFA", /* 0x39 */ + "LONGJMP", /* 0x3a */ + "BRANCHJ", /* 0x3b */ + "IFMATCH", /* 0x3c */ + "UNLESSM", /* 0x3d */ + "SUSPEND", /* 0x3e */ + "IFTHEN", /* 0x3f */ + "GROUPP", /* 0x40 */ + "EVAL", /* 0x41 */ + "MINMOD", /* 0x42 */ + "LOGICAL", /* 0x43 */ + "RENUM", /* 0x44 */ + "TRIE", /* 0x45 */ + "TRIEC", /* 0x46 */ + "AHOCORASICK", /* 0x47 */ + "AHOCORASICKC", /* 0x48 */ + "GOSUB", /* 0x49 */ + "GOSTART", /* 0x4a */ + "NGROUPP", /* 0x4b */ + "INSUBP", /* 0x4c */ + "DEFINEP", /* 0x4d */ + "ENDLIKE", /* 0x4e */ + "OPFAIL", /* 0x4f */ + "ACCEPT", /* 0x50 */ + "VERB", /* 0x51 */ + "PRUNE", /* 0x52 */ + "MARKPOINT", /* 0x53 */ + "SKIP", /* 0x54 */ + "COMMIT", /* 0x55 */ + "CUTGROUP", /* 0x56 */ + "KEEPS", /* 0x57 */ + "LNBREAK", /* 0x58 */ + "OPTIMIZED", /* 0x59 */ + "PSEUDO", /* 0x5a */ /* ------------ States ------------- */ "TRIE_next", /* REGNODE_MAX +0x01 */ "TRIE_next_fail", /* REGNODE_MAX +0x02 */ @@ -785,32 +627,32 @@ EXTCONST char * const PL_reg_name[] = { EXTCONST char * PL_reg_extflags_name[]; #else EXTCONST char * const PL_reg_extflags_name[] = { - /* Bits in extflags defined: 11111111111111111111111011111111 */ + /* Bits in extflags defined: 11111111111111110000001111111111 */ "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 */ + "EXTENDED_MORE", /* 0x00000010 */ + "KEEPCOPY", /* 0x00000020 */ + "CHARSET0", /* 0x00000040 : "CHARSET" - 0x000001c0 */ + "CHARSET1", /* 0x00000080 : "CHARSET" - 0x000001c0 */ + "CHARSET2", /* 0x00000100 : "CHARSET" - 0x000001c0 */ + "SPLIT", /* 0x00000200 */ + "UNUSED_BIT_10", /* 0x00000400 */ + "UNUSED_BIT_11", /* 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 */ - "SPLIT", /* 0x01000000 */ + "IS_ANCHORED", /* 0x01000000 */ "COPY_DONE", /* 0x02000000 */ "TAINTED_SEEN", /* 0x04000000 */ "TAINTED", /* 0x08000000 */ @@ -821,6 +663,36 @@ EXTCONST char * const PL_reg_extflags_name[] = { }; #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 */ + "CANY_SEEN", /* 0x00000080 - PREGf_CANY_SEEN */ + "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 */ +}; +#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))) @@ -828,9 +700,9 @@ EXTCONST char * const PL_reg_extflags_name[] = { EXTCONST U8 PL_varies[] __attribute__deprecated__; #else EXTCONST U8 PL_varies[] __attribute__deprecated__ = { - ANYOFV, CLUMP, BRANCH, BACK, STAR, PLUS, CURLY, CURLYN, CURLYM, CURLYX, - WHILEM, REF, REFF, REFFL, REFFU, REFFA, NREF, NREFF, NREFFL, NREFFU, - NREFFA, SUSPEND, IFTHEN, BRANCHJ, + 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 */ @@ -839,7 +711,7 @@ EXTCONST U8 PL_varies[] __attribute__deprecated__ = { EXTCONST U8 PL_varies_bitmask[]; #else EXTCONST U8 PL_varies_bitmask[] = { - 0x00, 0x00, 0x40, 0x00, 0x00, 0x00, 0x80, 0x03, 0xF8, 0xF3, 0x3F, 0x13, 0x00, 0x00, 0x00, 0x00 + 0x00, 0x00, 0x00, 0x30, 0x80, 0x3F, 0xFF, 0xCB, 0x00, 0x00, 0x00, 0x00 }; #endif /* DOINIT */ @@ -851,11 +723,8 @@ EXTCONST U8 PL_varies_bitmask[] = { 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, + REG_ANY, SANY, CANY, ANYOF, POSIXD, POSIXL, POSIXU, POSIXA, NPOSIXD, + NPOSIXL, NPOSIXU, NPOSIXA, 0 }; #endif /* DOINIT */ @@ -864,7 +733,7 @@ EXTCONST U8 PL_simple[] __attribute__deprecated__ = { EXTCONST U8 PL_simple_bitmask[]; #else EXTCONST U8 PL_simple_bitmask[] = { - 0x00, 0x00, 0xBC, 0xFF, 0xFF, 0xDD, 0x7F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xE0, 0x01 + 0x00, 0x00, 0xFF, 0x0F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; #endif /* DOINIT */