/* Regops and State definitions */
-#define REGNODE_MAX 114
-#define REGMATCH_STATE_MAX 154
+#define REGNODE_MAX 91
+#define REGMATCH_STATE_MAX 131
#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 CLUMP 47 /* 0x2f Match any extended grapheme cluster sequence */
-#define BRANCH 48 /* 0x30 Match this alternative, or the next... */
-#define BACK 49 /* 0x31 Match "", "next" ptr points backward. */
-#define EXACT 50 /* 0x32 Match this string (preceded by length). */
-#define EXACTF 51 /* 0x33 Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */
-#define EXACTFL 52 /* 0x34 Match this string (not guaranteed to be folded) using /il rules (w/len). */
-#define EXACTFU 53 /* 0x35 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 54 /* 0x36 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */
-#define EXACTFU_SS 55 /* 0x37 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 56 /* 0x38 Match this folded UTF-8 string using /iu rules */
-#define NOTHING 57 /* 0x39 Match empty string. */
-#define TAIL 58 /* 0x3a Match empty string. Can jump here from outside. */
-#define STAR 59 /* 0x3b Match this (simple) thing 0 or more times. */
-#define PLUS 60 /* 0x3c Match this (simple) thing 1 or more times. */
-#define CURLY 61 /* 0x3d Match this simple thing {n,m} times. */
-#define CURLYN 62 /* 0x3e Capture next-after-this simple thing */
-#define CURLYM 63 /* 0x3f Capture this medium-complex thing {n,m} times. */
-#define CURLYX 64 /* 0x40 Match this complex thing {n,m} times. */
-#define WHILEM 65 /* 0x41 Do curly processing and see if rest matches. */
-#define OPEN 66 /* 0x42 Mark this point in input as start of */
-#define CLOSE 67 /* 0x43 Analogous to OPEN. */
-#define REF 68 /* 0x44 Match some already matched string */
-#define REFF 69 /* 0x45 Match already matched string, folded using native charset semantics for non-utf8 */
-#define REFFL 70 /* 0x46 Match already matched string, folded in loc. */
-#define REFFU 71 /* 0x47 Match already matched string, folded using unicode semantics for non-utf8 */
-#define REFFA 72 /* 0x48 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */
-#define NREF 73 /* 0x49 Match some already matched string */
-#define NREFF 74 /* 0x4a Match already matched string, folded using native charset semantics for non-utf8 */
-#define NREFFL 75 /* 0x4b Match already matched string, folded in loc. */
-#define NREFFU 76 /* 0x4c Match already matched string, folded using unicode semantics for non-utf8 */
-#define NREFFA 77 /* 0x4d Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */
-#define IFMATCH 78 /* 0x4e Succeeds if the following matches. */
-#define UNLESSM 79 /* 0x4f Fails if the following matches. */
-#define SUSPEND 80 /* 0x50 "Independent" sub-RE. */
-#define IFTHEN 81 /* 0x51 Switch, should be preceded by switcher . */
-#define GROUPP 82 /* 0x52 Whether the group matched. */
-#define LONGJMP 83 /* 0x53 Jump far away. */
-#define BRANCHJ 84 /* 0x54 BRANCH with long offset. */
-#define EVAL 85 /* 0x55 Execute some Perl code. */
-#define MINMOD 86 /* 0x56 Next operator is not greedy. */
-#define LOGICAL 87 /* 0x57 Next opcode should set the flag only. */
-#define RENUM 88 /* 0x58 Group with independently numbered parens. */
-#define TRIE 89 /* 0x59 Match many EXACT(F[ALU]?)? at once. flags==type */
-#define TRIEC 90 /* 0x5a Same as TRIE, but with embedded charclass data */
-#define AHOCORASICK 91 /* 0x5b Aho Corasick stclass. flags==type */
-#define AHOCORASICKC 92 /* 0x5c Same as AHOCORASICK, but with embedded charclass data */
-#define GOSUB 93 /* 0x5d recurse to paren arg1 at (signed) ofs arg2 */
-#define GOSTART 94 /* 0x5e recurse to start of pattern */
-#define NGROUPP 95 /* 0x5f Whether the group matched. */
-#define INSUBP 96 /* 0x60 Whether we are in a specific recurse. */
-#define DEFINEP 97 /* 0x61 Never execute directly. */
-#define ENDLIKE 98 /* 0x62 Used only for the type field of verbs */
-#define OPFAIL 99 /* 0x63 Same as (?!) */
-#define ACCEPT 100 /* 0x64 Accepts the current matched string. */
-#define VERB 101 /* 0x65 Used only for the type field of verbs */
-#define PRUNE 102 /* 0x66 Pattern fails at this startpoint if no-backtracking through this */
-#define MARKPOINT 103 /* 0x67 Push the current location for rollback by cut. */
-#define SKIP 104 /* 0x68 On failure skip forward (to the mark) before retrying */
-#define COMMIT 105 /* 0x69 Pattern fails outright if backtracking through this */
-#define CUTGROUP 106 /* 0x6a On failure go to the next alternation in the group */
-#define KEEPS 107 /* 0x6b $& begins here. */
-#define LNBREAK 108 /* 0x6c generic newline pattern */
-#define VERTWS 109 /* 0x6d vertical whitespace (Perl 6) */
-#define NVERTWS 110 /* 0x6e not vertical whitespace (Perl 6) */
-#define HORIZWS 111 /* 0x6f horizontal whitespace (Perl 6) */
-#define NHORIZWS 112 /* 0x70 not horizontal whitespace (Perl 6) */
-#define OPTIMIZED 113 /* 0x71 Placeholder for dump. */
-#define PSEUDO 114 /* 0x72 Pseudo opcode for internal use. */
+#define SBOL 2 /* 0x02 Match "" at beginning of line: /^/, /\A/ */
+#define BOL 2 /* 0x02 type alias */
+#define MBOL 3 /* 0x03 Same, assuming multiline: /^/m */
+#define SEOL 4 /* 0x04 Match "" at end of line: /$/ */
+#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 BACK 30 /* 0x1e Match "", "next" ptr points backward. */
+#define EXACT 31 /* 0x1f Match this string (preceded by length). */
+#define EXACTF 32 /* 0x20 Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */
+#define EXACTFL 33 /* 0x21 Match this string (not guaranteed to be folded) using /il rules (w/len). */
+#define EXACTFU 34 /* 0x22 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 35 /* 0x23 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */
+#define EXACTFU_SS 36 /* 0x24 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 37 /* 0x25 Match this string (which is not trie-able; not guaranteed to be folded) using /iaa rules (w/len). */
+#define NOTHING 38 /* 0x26 Match empty string. */
+#define TAIL 39 /* 0x27 Match empty string. Can jump here from outside. */
+#define STAR 40 /* 0x28 Match this (simple) thing 0 or more times. */
+#define PLUS 41 /* 0x29 Match this (simple) thing 1 or more times. */
+#define CURLY 42 /* 0x2a Match this simple thing {n,m} times. */
+#define CURLYN 43 /* 0x2b Capture next-after-this simple thing */
+#define CURLYM 44 /* 0x2c Capture this medium-complex thing {n,m} times. */
+#define CURLYX 45 /* 0x2d Match this complex thing {n,m} times. */
+#define WHILEM 46 /* 0x2e Do curly processing and see if rest matches. */
+#define OPEN 47 /* 0x2f Mark this point in input as start of #n. */
+#define CLOSE 48 /* 0x30 Analogous to OPEN. */
+#define REF 49 /* 0x31 Match some already matched string */
+#define REFF 50 /* 0x32 Match already matched string, folded using native charset rules for non-utf8 */
+#define REFFL 51 /* 0x33 Match already matched string, folded in loc. */
+#define REFFU 52 /* 0x34 Match already matched string, folded using unicode rules for non-utf8 */
+#define REFFA 53 /* 0x35 Match already matched string, folded using unicode rules for non-utf8, no mixing ASCII, non-ASCII */
+#define NREF 54 /* 0x36 Match some already matched string */
+#define NREFF 55 /* 0x37 Match already matched string, folded using native charset rules for non-utf8 */
+#define NREFFL 56 /* 0x38 Match already matched string, folded in loc. */
+#define NREFFU 57 /* 0x39 Match already matched string, folded using unicode rules for non-utf8 */
+#define NREFFA 58 /* 0x3a Match already matched string, folded using unicode rules for non-utf8, no mixing ASCII, non-ASCII */
+#define LONGJMP 59 /* 0x3b Jump far away. */
+#define BRANCHJ 60 /* 0x3c BRANCH with long offset. */
+#define IFMATCH 61 /* 0x3d Succeeds if the following matches. */
+#define UNLESSM 62 /* 0x3e Fails if the following matches. */
+#define SUSPEND 63 /* 0x3f "Independent" sub-RE. */
+#define IFTHEN 64 /* 0x40 Switch, should be preceded by switcher. */
+#define GROUPP 65 /* 0x41 Whether the group matched. */
+#define EVAL 66 /* 0x42 Execute some Perl code. */
+#define MINMOD 67 /* 0x43 Next operator is not greedy. */
+#define LOGICAL 68 /* 0x44 Next opcode should set the flag only. */
+#define RENUM 69 /* 0x45 Group with independently numbered parens. */
+#define TRIE 70 /* 0x46 Match many EXACT(F[ALU]?)? at once. flags==type */
+#define TRIEC 71 /* 0x47 Same as TRIE, but with embedded charclass data */
+#define AHOCORASICK 72 /* 0x48 Aho Corasick stclass. flags==type */
+#define AHOCORASICKC 73 /* 0x49 Same as AHOCORASICK, but with embedded charclass data */
+#define GOSUB 74 /* 0x4a recurse to paren arg1 at (signed) ofs arg2 */
+#define GOSTART 75 /* 0x4b recurse to start of pattern */
+#define NGROUPP 76 /* 0x4c Whether the group matched. */
+#define INSUBP 77 /* 0x4d Whether we are in a specific recurse. */
+#define DEFINEP 78 /* 0x4e Never execute directly. */
+#define ENDLIKE 79 /* 0x4f Used only for the type field of verbs */
+#define OPFAIL 80 /* 0x50 Same as (?!) */
+#define ACCEPT 81 /* 0x51 Accepts the current matched string. */
+#define VERB 82 /* 0x52 Used only for the type field of verbs */
+#define PRUNE 83 /* 0x53 Pattern fails at this startpoint if no-backtracking through this */
+#define MARKPOINT 84 /* 0x54 Push the current location for rollback by cut. */
+#define SKIP 85 /* 0x55 On failure skip forward (to the mark) before retrying */
+#define COMMIT 86 /* 0x56 Pattern fails outright if backtracking through this */
+#define CUTGROUP 87 /* 0x57 On failure go to the next alternation in the group */
+#define KEEPS 88 /* 0x58 $& begins here. */
+#define LNBREAK 89 /* 0x59 generic newline pattern */
+#define OPTIMIZED 90 /* 0x5a Placeholder for dump. */
+#define PSEUDO 91 /* 0x5b 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 */
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 */
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 */
+ NPOSIXD, /* NPOSIXD */
+ NPOSIXD, /* NPOSIXL */
+ NPOSIXD, /* NPOSIXU */
+ NPOSIXD, /* NPOSIXA */
CLUMP, /* CLUMP */
BRANCH, /* BRANCH */
BACK, /* BACK */
EXACT, /* EXACTFU */
EXACT, /* EXACTFA */
EXACT, /* EXACTFU_SS */
- EXACT, /* EXACTFU_TRICKYFOLD */
+ EXACT, /* EXACTFA_NO_TRIE */
NOTHING, /* NOTHING */
NOTHING, /* TAIL */
STAR, /* STAR */
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 */
VERB, /* CUTGROUP */
KEEPS, /* KEEPS */
LNBREAK, /* LNBREAK */
- VERTWS, /* VERTWS */
- NVERTWS, /* NVERTWS */
- HORIZWS, /* HORIZWS */
- NHORIZWS, /* NHORIZWS */
NOTHING, /* OPTIMIZED */
PSEUDO, /* PSEUDO */
/* ------------ States ------------- */
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, /* 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 */
+ 0, /* POSIXA */
+ 0, /* NPOSIXD */
+ 0, /* NPOSIXL */
+ 0, /* NPOSIXU */
+ 0, /* NPOSIXA */
0, /* CLUMP */
0, /* BRANCH */
0, /* BACK */
0, /* EXACTFU */
0, /* EXACTFA */
0, /* EXACTFU_SS */
- 0, /* EXACTFU_TRICKYFOLD */
+ 0, /* EXACTFA_NO_TRIE */
0, /* NOTHING */
0, /* TAIL */
0, /* STAR */
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 */
0, /* MINMOD */
0, /* LOGICAL */
EXTRA_SIZE(struct regnode_1), /* CUTGROUP */
0, /* KEEPS */
0, /* LNBREAK */
- 0, /* VERTWS */
- 0, /* NVERTWS */
- 0, /* HORIZWS */
- 0, /* NHORIZWS */
0, /* OPTIMIZED */
0, /* PSEUDO */
};
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, /* 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 */
+ 0, /* POSIXA */
+ 0, /* NPOSIXD */
+ 0, /* NPOSIXL */
+ 0, /* NPOSIXU */
+ 0, /* NPOSIXA */
0, /* CLUMP */
0, /* BRANCH */
0, /* BACK */
0, /* EXACTFU */
0, /* EXACTFA */
0, /* EXACTFU_SS */
- 0, /* EXACTFU_TRICKYFOLD */
+ 0, /* EXACTFA_NO_TRIE */
0, /* NOTHING */
0, /* TAIL */
0, /* STAR */
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 */
0, /* CUTGROUP */
0, /* KEEPS */
0, /* LNBREAK */
- 0, /* VERTWS */
- 0, /* NVERTWS */
- 0, /* HORIZWS */
- 0, /* NHORIZWS */
0, /* OPTIMIZED */
0, /* PSEUDO */
};
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 */
- "CLUMP", /* 0x2f */
- "BRANCH", /* 0x30 */
- "BACK", /* 0x31 */
- "EXACT", /* 0x32 */
- "EXACTF", /* 0x33 */
- "EXACTFL", /* 0x34 */
- "EXACTFU", /* 0x35 */
- "EXACTFA", /* 0x36 */
- "EXACTFU_SS", /* 0x37 */
- "EXACTFU_TRICKYFOLD", /* 0x38 */
- "NOTHING", /* 0x39 */
- "TAIL", /* 0x3a */
- "STAR", /* 0x3b */
- "PLUS", /* 0x3c */
- "CURLY", /* 0x3d */
- "CURLYN", /* 0x3e */
- "CURLYM", /* 0x3f */
- "CURLYX", /* 0x40 */
- "WHILEM", /* 0x41 */
- "OPEN", /* 0x42 */
- "CLOSE", /* 0x43 */
- "REF", /* 0x44 */
- "REFF", /* 0x45 */
- "REFFL", /* 0x46 */
- "REFFU", /* 0x47 */
- "REFFA", /* 0x48 */
- "NREF", /* 0x49 */
- "NREFF", /* 0x4a */
- "NREFFL", /* 0x4b */
- "NREFFU", /* 0x4c */
- "NREFFA", /* 0x4d */
- "IFMATCH", /* 0x4e */
- "UNLESSM", /* 0x4f */
- "SUSPEND", /* 0x50 */
- "IFTHEN", /* 0x51 */
- "GROUPP", /* 0x52 */
- "LONGJMP", /* 0x53 */
- "BRANCHJ", /* 0x54 */
- "EVAL", /* 0x55 */
- "MINMOD", /* 0x56 */
- "LOGICAL", /* 0x57 */
- "RENUM", /* 0x58 */
- "TRIE", /* 0x59 */
- "TRIEC", /* 0x5a */
- "AHOCORASICK", /* 0x5b */
- "AHOCORASICKC", /* 0x5c */
- "GOSUB", /* 0x5d */
- "GOSTART", /* 0x5e */
- "NGROUPP", /* 0x5f */
- "INSUBP", /* 0x60 */
- "DEFINEP", /* 0x61 */
- "ENDLIKE", /* 0x62 */
- "OPFAIL", /* 0x63 */
- "ACCEPT", /* 0x64 */
- "VERB", /* 0x65 */
- "PRUNE", /* 0x66 */
- "MARKPOINT", /* 0x67 */
- "SKIP", /* 0x68 */
- "COMMIT", /* 0x69 */
- "CUTGROUP", /* 0x6a */
- "KEEPS", /* 0x6b */
- "LNBREAK", /* 0x6c */
- "VERTWS", /* 0x6d */
- "NVERTWS", /* 0x6e */
- "HORIZWS", /* 0x6f */
- "NHORIZWS", /* 0x70 */
- "OPTIMIZED", /* 0x71 */
- "PSEUDO", /* 0x72 */
+ "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 */
+ "BACK", /* 0x1e */
+ "EXACT", /* 0x1f */
+ "EXACTF", /* 0x20 */
+ "EXACTFL", /* 0x21 */
+ "EXACTFU", /* 0x22 */
+ "EXACTFA", /* 0x23 */
+ "EXACTFU_SS", /* 0x24 */
+ "EXACTFA_NO_TRIE", /* 0x25 */
+ "NOTHING", /* 0x26 */
+ "TAIL", /* 0x27 */
+ "STAR", /* 0x28 */
+ "PLUS", /* 0x29 */
+ "CURLY", /* 0x2a */
+ "CURLYN", /* 0x2b */
+ "CURLYM", /* 0x2c */
+ "CURLYX", /* 0x2d */
+ "WHILEM", /* 0x2e */
+ "OPEN", /* 0x2f */
+ "CLOSE", /* 0x30 */
+ "REF", /* 0x31 */
+ "REFF", /* 0x32 */
+ "REFFL", /* 0x33 */
+ "REFFU", /* 0x34 */
+ "REFFA", /* 0x35 */
+ "NREF", /* 0x36 */
+ "NREFF", /* 0x37 */
+ "NREFFL", /* 0x38 */
+ "NREFFU", /* 0x39 */
+ "NREFFA", /* 0x3a */
+ "LONGJMP", /* 0x3b */
+ "BRANCHJ", /* 0x3c */
+ "IFMATCH", /* 0x3d */
+ "UNLESSM", /* 0x3e */
+ "SUSPEND", /* 0x3f */
+ "IFTHEN", /* 0x40 */
+ "GROUPP", /* 0x41 */
+ "EVAL", /* 0x42 */
+ "MINMOD", /* 0x43 */
+ "LOGICAL", /* 0x44 */
+ "RENUM", /* 0x45 */
+ "TRIE", /* 0x46 */
+ "TRIEC", /* 0x47 */
+ "AHOCORASICK", /* 0x48 */
+ "AHOCORASICKC", /* 0x49 */
+ "GOSUB", /* 0x4a */
+ "GOSTART", /* 0x4b */
+ "NGROUPP", /* 0x4c */
+ "INSUBP", /* 0x4d */
+ "DEFINEP", /* 0x4e */
+ "ENDLIKE", /* 0x4f */
+ "OPFAIL", /* 0x50 */
+ "ACCEPT", /* 0x51 */
+ "VERB", /* 0x52 */
+ "PRUNE", /* 0x53 */
+ "MARKPOINT", /* 0x54 */
+ "SKIP", /* 0x55 */
+ "COMMIT", /* 0x56 */
+ "CUTGROUP", /* 0x57 */
+ "KEEPS", /* 0x58 */
+ "LNBREAK", /* 0x59 */
+ "OPTIMIZED", /* 0x5a */
+ "PSEUDO", /* 0x5b */
/* ------------ States ------------- */
"TRIE_next", /* REGNODE_MAX +0x01 */
"TRIE_next_fail", /* REGNODE_MAX +0x02 */
EXTCONST char * PL_reg_extflags_name[];
#else
EXTCONST char * const PL_reg_extflags_name[] = {
- /* Bits in extflags defined: 11111111111111111111111011111111 */
+ /* Bits in extflags defined: 11111111111111110000000111111111 */
"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 */
+ "CHARSET0", /* 0x00000020 : "CHARSET" - 0x000000e0 */
+ "CHARSET1", /* 0x00000040 : "CHARSET" - 0x000000e0 */
+ "CHARSET2", /* 0x00000080 : "CHARSET" - 0x000000e0 */
+ "SPLIT", /* 0x00000100 */
+ "UNUSED_BIT_9", /* 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 */
};
#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)))
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, BACK, STAR, PLUS, CURLY, CURLYN, CURLYM, CURLYX, WHILEM,
+ REF, REFF, REFFL, REFFU, REFFA, NREF, NREFF, NREFFL, NREFFU, NREFFA,
+ BRANCHJ, SUSPEND, IFTHEN,
0
};
#endif /* DOINIT */
EXTCONST U8 PL_varies_bitmask[];
#else
EXTCONST U8 PL_varies_bitmask[] = {
- 0x00, 0x00, 0x40, 0x00, 0x00, 0x80, 0x03, 0xF8, 0xF3, 0x3F, 0x13, 0x00, 0x00, 0x00, 0x00
+ 0x00, 0x00, 0x00, 0x70, 0x00, 0x7F, 0xFE, 0x97, 0x01, 0x00, 0x00, 0x00
};
#endif /* 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, VERTWS, NVERTWS, HORIZWS, NHORIZWS,
+ REG_ANY, SANY, CANY, ANYOF, POSIXD, POSIXL, POSIXU, POSIXA, NPOSIXD,
+ NPOSIXL, NPOSIXU, NPOSIXA,
0
};
#endif /* DOINIT */
EXTCONST U8 PL_simple_bitmask[];
#else
EXTCONST U8 PL_simple_bitmask[] = {
- 0x00, 0x00, 0xBC, 0xFF, 0xFF, 0x5D, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xE0, 0x01
+ 0x00, 0x00, 0xFF, 0x0F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00
};
#endif /* DOINIT */
https://perl5.git.perl.org / perl5.git / blobdiff