X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/9e84774b2a68a51af5e246f365059d61ebe86fb0..f1fb9b037a1fe86e0ccdfbf27affa94647af2a37:/regnodes.h?ds=sidebyside diff --git a/regnodes.h b/regnodes.h index 2024d15..b302f5d 100644 --- a/regnodes.h +++ b/regnodes.h @@ -6,8 +6,8 @@ /* Regops and State definitions */ -#define REGNODE_MAX 121 -#define REGMATCH_STATE_MAX 161 +#define REGNODE_MAX 95 +#define REGMATCH_STATE_MAX 135 #define END 0 /* 0000 End of program. */ #define SUCCEED 1 /* 0x01 Return from a subroutine, basically. */ @@ -31,106 +31,80 @@ #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. */ +#define ANYOF_WARN_SUPER 22 /* 0x16 Match character in (or not in) this class, warn (if enabled) upon matching a char above Unicode max; */ +#define ANYOF_SYNTHETIC 23 /* 0x17 Synthetic start class */ +#define POSIXD 24 /* 0x18 Some [[:class:]] under /d; the FLAGS field gives which one */ +#define POSIXL 25 /* 0x19 Some [[:class:]] under /l; the FLAGS field gives which one */ +#define POSIXU 26 /* 0x1a Some [[:class:]] under /u; the FLAGS field gives which one */ +#define POSIXA 27 /* 0x1b Some [[:class:]] under /a; the FLAGS field gives which one */ +#define NPOSIXD 28 /* 0x1c complement of POSIXD, [[:^class:]] */ +#define NPOSIXL 29 /* 0x1d complement of POSIXL, [[:^class:]] */ +#define NPOSIXU 30 /* 0x1e complement of POSIXU, [[:^class:]] */ +#define NPOSIXA 31 /* 0x1f complement of POSIXA, [[:^class:]] */ +#define CLUMP 32 /* 0x20 Match any extended grapheme cluster sequence */ +#define BRANCH 33 /* 0x21 Match this alternative, or the next... */ +#define BACK 34 /* 0x22 Match "", "next" ptr points backward. */ +#define EXACT 35 /* 0x23 Match this string (preceded by length). */ +#define EXACTF 36 /* 0x24 Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ +#define EXACTFL 37 /* 0x25 Match this string (not guaranteed to be folded) using /il rules (w/len). */ +#define EXACTFU 38 /* 0x26 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 39 /* 0x27 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ +#define EXACTFU_SS 40 /* 0x28 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 41 /* 0x29 Match this folded UTF-8 string using /iu rules */ +#define NOTHING 42 /* 0x2a Match empty string. */ +#define TAIL 43 /* 0x2b Match empty string. Can jump here from outside. */ +#define STAR 44 /* 0x2c Match this (simple) thing 0 or more times. */ +#define PLUS 45 /* 0x2d Match this (simple) thing 1 or more times. */ +#define CURLY 46 /* 0x2e Match this simple thing {n,m} times. */ +#define CURLYN 47 /* 0x2f Capture next-after-this simple thing */ +#define CURLYM 48 /* 0x30 Capture this medium-complex thing {n,m} times. */ +#define CURLYX 49 /* 0x31 Match this complex thing {n,m} times. */ +#define WHILEM 50 /* 0x32 Do curly processing and see if rest matches. */ +#define OPEN 51 /* 0x33 Mark this point in input as start of #n. */ +#define CLOSE 52 /* 0x34 Analogous to OPEN. */ +#define REF 53 /* 0x35 Match some already matched string */ +#define REFF 54 /* 0x36 Match already matched string, folded using native charset semantics for non-utf8 */ +#define REFFL 55 /* 0x37 Match already matched string, folded in loc. */ +#define REFFU 56 /* 0x38 Match already matched string, folded using unicode semantics for non-utf8 */ +#define REFFA 57 /* 0x39 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#define NREF 58 /* 0x3a Match some already matched string */ +#define NREFF 59 /* 0x3b Match already matched string, folded using native charset semantics for non-utf8 */ +#define NREFFL 60 /* 0x3c Match already matched string, folded in loc. */ +#define NREFFU 61 /* 0x3d Match already matched string, folded using unicode semantics for non-utf8 */ +#define NREFFA 62 /* 0x3e Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#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 LONGJMP 68 /* 0x44 Jump far away. */ +#define BRANCHJ 69 /* 0x45 BRANCH with long offset. */ +#define EVAL 70 /* 0x46 Execute some Perl code. */ +#define MINMOD 71 /* 0x47 Next operator is not greedy. */ +#define LOGICAL 72 /* 0x48 Next opcode should set the flag only. */ +#define RENUM 73 /* 0x49 Group with independently numbered parens. */ +#define TRIE 74 /* 0x4a Match many EXACT(F[ALU]?)? at once. flags==type */ +#define TRIEC 75 /* 0x4b Same as TRIE, but with embedded charclass data */ +#define AHOCORASICK 76 /* 0x4c Aho Corasick stclass. flags==type */ +#define AHOCORASICKC 77 /* 0x4d Same as AHOCORASICK, but with embedded charclass data */ +#define GOSUB 78 /* 0x4e recurse to paren arg1 at (signed) ofs arg2 */ +#define GOSTART 79 /* 0x4f recurse to start of pattern */ +#define NGROUPP 80 /* 0x50 Whether the group matched. */ +#define INSUBP 81 /* 0x51 Whether we are in a specific recurse. */ +#define DEFINEP 82 /* 0x52 Never execute directly. */ +#define ENDLIKE 83 /* 0x53 Used only for the type field of verbs */ +#define OPFAIL 84 /* 0x54 Same as (?!) */ +#define ACCEPT 85 /* 0x55 Accepts the current matched string. */ +#define VERB 86 /* 0x56 Used only for the type field of verbs */ +#define PRUNE 87 /* 0x57 Pattern fails at this startpoint if no-backtracking through this */ +#define MARKPOINT 88 /* 0x58 Push the current location for rollback by cut. */ +#define SKIP 89 /* 0x59 On failure skip forward (to the mark) before retrying */ +#define COMMIT 90 /* 0x5a Pattern fails outright if backtracking through this */ +#define CUTGROUP 91 /* 0x5b On failure go to the next alternation in the group */ +#define KEEPS 92 /* 0x5c $& begins here. */ +#define LNBREAK 93 /* 0x5d generic newline pattern */ +#define OPTIMIZED 94 /* 0x5e Placeholder for dump. */ +#define PSEUDO 95 /* 0x5f 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 */ @@ -201,30 +175,8 @@ EXTCONST U8 PL_regkind[] = { REG_ANY, /* SANY */ REG_ANY, /* CANY */ 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 */ + ANYOF, /* ANYOF_WARN_SUPER */ + ANYOF, /* ANYOF_SYNTHETIC */ POSIXD, /* POSIXD */ POSIXD, /* POSIXL */ POSIXD, /* POSIXU */ @@ -295,10 +247,6 @@ EXTCONST U8 PL_regkind[] = { VERB, /* CUTGROUP */ KEEPS, /* KEEPS */ LNBREAK, /* LNBREAK */ - VERTWS, /* VERTWS */ - NVERTWS, /* NVERTWS */ - HORIZWS, /* HORIZWS */ - NHORIZWS, /* NHORIZWS */ NOTHING, /* OPTIMIZED */ PSEUDO, /* PSEUDO */ /* ------------ States ------------- */ @@ -371,30 +319,8 @@ static const U8 regarglen[] = { 0, /* SANY */ 0, /* CANY */ 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, /* ANYOF_WARN_SUPER */ + 0, /* ANYOF_SYNTHETIC */ 0, /* POSIXD */ 0, /* POSIXL */ 0, /* POSIXU */ @@ -465,10 +391,6 @@ 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 */ }; @@ -498,30 +420,8 @@ static const char reg_off_by_arg[] = { 0, /* SANY */ 0, /* CANY */ 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, /* ANYOF_WARN_SUPER */ + 0, /* ANYOF_SYNTHETIC */ 0, /* POSIXD */ 0, /* POSIXL */ 0, /* POSIXU */ @@ -592,10 +492,6 @@ 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 */ }; @@ -630,106 +526,80 @@ EXTCONST char * const PL_reg_name[] = { "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 */ + "ANYOF_WARN_SUPER", /* 0x16 */ + "ANYOF_SYNTHETIC", /* 0x17 */ + "POSIXD", /* 0x18 */ + "POSIXL", /* 0x19 */ + "POSIXU", /* 0x1a */ + "POSIXA", /* 0x1b */ + "NPOSIXD", /* 0x1c */ + "NPOSIXL", /* 0x1d */ + "NPOSIXU", /* 0x1e */ + "NPOSIXA", /* 0x1f */ + "CLUMP", /* 0x20 */ + "BRANCH", /* 0x21 */ + "BACK", /* 0x22 */ + "EXACT", /* 0x23 */ + "EXACTF", /* 0x24 */ + "EXACTFL", /* 0x25 */ + "EXACTFU", /* 0x26 */ + "EXACTFA", /* 0x27 */ + "EXACTFU_SS", /* 0x28 */ + "EXACTFU_TRICKYFOLD", /* 0x29 */ + "NOTHING", /* 0x2a */ + "TAIL", /* 0x2b */ + "STAR", /* 0x2c */ + "PLUS", /* 0x2d */ + "CURLY", /* 0x2e */ + "CURLYN", /* 0x2f */ + "CURLYM", /* 0x30 */ + "CURLYX", /* 0x31 */ + "WHILEM", /* 0x32 */ + "OPEN", /* 0x33 */ + "CLOSE", /* 0x34 */ + "REF", /* 0x35 */ + "REFF", /* 0x36 */ + "REFFL", /* 0x37 */ + "REFFU", /* 0x38 */ + "REFFA", /* 0x39 */ + "NREF", /* 0x3a */ + "NREFF", /* 0x3b */ + "NREFFL", /* 0x3c */ + "NREFFU", /* 0x3d */ + "NREFFA", /* 0x3e */ + "IFMATCH", /* 0x3f */ + "UNLESSM", /* 0x40 */ + "SUSPEND", /* 0x41 */ + "IFTHEN", /* 0x42 */ + "GROUPP", /* 0x43 */ + "LONGJMP", /* 0x44 */ + "BRANCHJ", /* 0x45 */ + "EVAL", /* 0x46 */ + "MINMOD", /* 0x47 */ + "LOGICAL", /* 0x48 */ + "RENUM", /* 0x49 */ + "TRIE", /* 0x4a */ + "TRIEC", /* 0x4b */ + "AHOCORASICK", /* 0x4c */ + "AHOCORASICKC", /* 0x4d */ + "GOSUB", /* 0x4e */ + "GOSTART", /* 0x4f */ + "NGROUPP", /* 0x50 */ + "INSUBP", /* 0x51 */ + "DEFINEP", /* 0x52 */ + "ENDLIKE", /* 0x53 */ + "OPFAIL", /* 0x54 */ + "ACCEPT", /* 0x55 */ + "VERB", /* 0x56 */ + "PRUNE", /* 0x57 */ + "MARKPOINT", /* 0x58 */ + "SKIP", /* 0x59 */ + "COMMIT", /* 0x5a */ + "CUTGROUP", /* 0x5b */ + "KEEPS", /* 0x5c */ + "LNBREAK", /* 0x5d */ + "OPTIMIZED", /* 0x5e */ + "PSEUDO", /* 0x5f */ /* ------------ States ------------- */ "TRIE_next", /* REGNODE_MAX +0x01 */ "TRIE_next_fail", /* REGNODE_MAX +0x02 */ @@ -780,23 +650,23 @@ EXTCONST char * const PL_reg_name[] = { EXTCONST char * PL_reg_extflags_name[]; #else EXTCONST char * const PL_reg_extflags_name[] = { - /* Bits in extflags defined: 11011111111111111111111011111111 */ + /* Bits in extflags defined: 11111110111111111111111111111111 */ "MULTILINE", /* 0x00000001 */ "SINGLELINE", /* 0x00000002 */ "FOLD", /* 0x00000004 */ "EXTENDED", /* 0x00000008 */ "KEEPCOPY", /* 0x00000010 */ - "CHARSET", /* 0x000000e0 */ - "CHARSET", /* 0x000000e0 */ - "CHARSET", /* 0x000000e0 */ - "UNUSED_BIT_8", /* 0x00000100 */ + "CHARSET0", /* 0x00000020 : "CHARSET" - 0x000000e0 */ + "CHARSET1", /* 0x00000040 : "CHARSET" - 0x000000e0 */ + "CHARSET2", /* 0x00000080 : "CHARSET" - 0x000000e0 */ + "SPLIT", /* 0x00000100 */ "ANCH_BOL", /* 0x00000200 */ "ANCH_MBOL", /* 0x00000400 */ "ANCH_SBOL", /* 0x00000800 */ "ANCH_GPOS", /* 0x00001000 */ "GPOS_SEEN", /* 0x00002000 */ "GPOS_FLOAT", /* 0x00004000 */ - "LOOKBEHIND_SEEN", /* 0x00008000 */ + "NO_INPLACE_SUBST", /* 0x00008000 */ "EVAL_SEEN", /* 0x00010000 */ "CANY_SEEN", /* 0x00020000 */ "NOSCAN", /* 0x00040000 */ @@ -805,17 +675,32 @@ EXTCONST char * const PL_reg_extflags_name[] = { "USE_INTUIT_NOML", /* 0x00200000 */ "USE_INTUIT_ML", /* 0x00400000 */ "INTUIT_TAIL", /* 0x00800000 */ - "MODIFIES_VARS", /* 0x01000000 */ + "UNUSED_BIT_24", /* 0x01000000 */ "COPY_DONE", /* 0x02000000 */ "TAINTED_SEEN", /* 0x04000000 */ "TAINTED", /* 0x08000000 */ "START_ONLY", /* 0x10000000 */ - "UNUSED_BIT_29", /* 0x20000000 */ + "SKIPWHITE", /* 0x20000000 */ "WHITE", /* 0x40000000 */ "NULL", /* 0x80000000 */ }; #endif /* DOINIT */ +/* 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'" */ +}; +#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))) @@ -834,7 +719,7 @@ EXTCONST U8 PL_varies[] __attribute__deprecated__ = { 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 + 0x00, 0x00, 0x00, 0x00, 0x07, 0xF0, 0xE7, 0x7F, 0x26, 0x00, 0x00, 0x00 }; #endif /* DOINIT */ @@ -846,11 +731,8 @@ 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, ANYOF_WARN_SUPER, ANYOF_SYNTHETIC, POSIXD, + POSIXL, POSIXU, POSIXA, NPOSIXD, NPOSIXL, NPOSIXU, NPOSIXA, 0 }; #endif /* DOINIT */ @@ -859,7 +741,7 @@ EXTCONST U8 PL_simple[] __attribute__deprecated__ = { 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 + 0x00, 0x00, 0xFC, 0xFF, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; #endif /* DOINIT */