diff options
Diffstat (limited to 'regnodes.h')
| -rw-r--r-- | regnodes.h | 305 |
1 files changed, 155 insertions, 150 deletions
diff --git a/regnodes.h b/regnodes.h index 475492129d..d1d3cab787 100644 --- a/regnodes.h +++ b/regnodes.h @@ -6,8 +6,8 @@ /* Regops and State definitions */ -#define REGNODE_MAX 93 -#define REGMATCH_STATE_MAX 133 +#define REGNODE_MAX 94 +#define REGMATCH_STATE_MAX 134 #define END 0 /* 0000 End of program. */ #define SUCCEED 1 /* 0x01 Return from a subroutine, basically. */ @@ -31,78 +31,79 @@ #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 POSIXD 22 /* 0x16 Some [[:class:]] under /d; the FLAGS field gives which one */ -#define POSIXL 23 /* 0x17 Some [[:class:]] under /l; the FLAGS field gives which one */ -#define POSIXU 24 /* 0x18 Some [[:class:]] under /u; the FLAGS field gives which one */ -#define POSIXA 25 /* 0x19 Some [[:class:]] under /a; the FLAGS field gives which one */ -#define NPOSIXD 26 /* 0x1a complement of POSIXD, [[:^class:]] */ -#define NPOSIXL 27 /* 0x1b complement of POSIXL, [[:^class:]] */ -#define NPOSIXU 28 /* 0x1c complement of POSIXU, [[:^class:]] */ -#define NPOSIXA 29 /* 0x1d complement of POSIXA, [[:^class:]] */ -#define CLUMP 30 /* 0x1e Match any extended grapheme cluster 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 non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ -#define EXACTFL 35 /* 0x23 Match this string (not guaranteed to be folded) using /il rules (w/len). */ -#define EXACTFU 36 /* 0x24 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 37 /* 0x25 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ -#define EXACTFU_SS 38 /* 0x26 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 39 /* 0x27 Match this folded UTF-8 string using /iu rules */ -#define NOTHING 40 /* 0x28 Match empty string. */ -#define TAIL 41 /* 0x29 Match empty string. Can jump here from outside. */ -#define STAR 42 /* 0x2a Match this (simple) thing 0 or more times. */ -#define PLUS 43 /* 0x2b Match this (simple) thing 1 or more times. */ -#define CURLY 44 /* 0x2c Match this simple thing {n,m} times. */ -#define CURLYN 45 /* 0x2d Capture next-after-this simple thing */ -#define CURLYM 46 /* 0x2e Capture this medium-complex thing {n,m} times. */ -#define CURLYX 47 /* 0x2f Match this complex thing {n,m} times. */ -#define WHILEM 48 /* 0x30 Do curly processing and see if rest matches. */ -#define OPEN 49 /* 0x31 Mark this point in input as start of #n. */ -#define CLOSE 50 /* 0x32 Analogous to OPEN. */ -#define REF 51 /* 0x33 Match some already matched string */ -#define REFF 52 /* 0x34 Match already matched string, folded using native charset semantics for non-utf8 */ -#define REFFL 53 /* 0x35 Match already matched string, folded in loc. */ -#define REFFU 54 /* 0x36 Match already matched string, folded using unicode semantics for non-utf8 */ -#define REFFA 55 /* 0x37 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ -#define NREF 56 /* 0x38 Match some already matched string */ -#define NREFF 57 /* 0x39 Match already matched string, folded using native charset semantics for non-utf8 */ -#define NREFFL 58 /* 0x3a Match already matched string, folded in loc. */ -#define NREFFU 59 /* 0x3b Match already matched string, folded using unicode semantics for non-utf8 */ -#define NREFFA 60 /* 0x3c Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ -#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 LONGJMP 66 /* 0x42 Jump far away. */ -#define BRANCHJ 67 /* 0x43 BRANCH with long offset. */ -#define EVAL 68 /* 0x44 Execute some Perl code. */ -#define MINMOD 69 /* 0x45 Next operator is not greedy. */ -#define LOGICAL 70 /* 0x46 Next opcode should set the flag only. */ -#define RENUM 71 /* 0x47 Group with independently numbered parens. */ -#define TRIE 72 /* 0x48 Match many EXACT(F[ALU]?)? at once. flags==type */ -#define TRIEC 73 /* 0x49 Same as TRIE, but with embedded charclass data */ -#define AHOCORASICK 74 /* 0x4a Aho Corasick stclass. flags==type */ -#define AHOCORASICKC 75 /* 0x4b Same as AHOCORASICK, but with embedded charclass data */ -#define GOSUB 76 /* 0x4c recurse to paren arg1 at (signed) ofs arg2 */ -#define GOSTART 77 /* 0x4d recurse to start of pattern */ -#define NGROUPP 78 /* 0x4e Whether the group matched. */ -#define INSUBP 79 /* 0x4f Whether we are in a specific recurse. */ -#define DEFINEP 80 /* 0x50 Never execute directly. */ -#define ENDLIKE 81 /* 0x51 Used only for the type field of verbs */ -#define OPFAIL 82 /* 0x52 Same as (?!) */ -#define ACCEPT 83 /* 0x53 Accepts the current matched string. */ -#define VERB 84 /* 0x54 Used only for the type field of verbs */ -#define PRUNE 85 /* 0x55 Pattern fails at this startpoint if no-backtracking through this */ -#define MARKPOINT 86 /* 0x56 Push the current location for rollback by cut. */ -#define SKIP 87 /* 0x57 On failure skip forward (to the mark) before retrying */ -#define COMMIT 88 /* 0x58 Pattern fails outright if backtracking through this */ -#define CUTGROUP 89 /* 0x59 On failure go to the next alternation in the group */ -#define KEEPS 90 /* 0x5a $& begins here. */ -#define LNBREAK 91 /* 0x5b generic newline pattern */ -#define OPTIMIZED 92 /* 0x5c Placeholder for dump. */ -#define PSEUDO 93 /* 0x5d 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 POSIXD 23 /* 0x17 Some [[:class:]] under /d; the FLAGS field gives which one */ +#define POSIXL 24 /* 0x18 Some [[:class:]] under /l; the FLAGS field gives which one */ +#define POSIXU 25 /* 0x19 Some [[:class:]] under /u; the FLAGS field gives which one */ +#define POSIXA 26 /* 0x1a Some [[:class:]] under /a; the FLAGS field gives which one */ +#define NPOSIXD 27 /* 0x1b complement of POSIXD, [[:^class:]] */ +#define NPOSIXL 28 /* 0x1c complement of POSIXL, [[:^class:]] */ +#define NPOSIXU 29 /* 0x1d complement of POSIXU, [[:^class:]] */ +#define NPOSIXA 30 /* 0x1e complement of POSIXA, [[:^class:]] */ +#define CLUMP 31 /* 0x1f Match any extended grapheme cluster sequence */ +#define BRANCH 32 /* 0x20 Match this alternative, or the next... */ +#define BACK 33 /* 0x21 Match "", "next" ptr points backward. */ +#define EXACT 34 /* 0x22 Match this string (preceded by length). */ +#define EXACTF 35 /* 0x23 Match this non-UTF-8 string (not guaranteed to be folded) using /id rules (w/len). */ +#define EXACTFL 36 /* 0x24 Match this string (not guaranteed to be folded) using /il rules (w/len). */ +#define EXACTFU 37 /* 0x25 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 38 /* 0x26 Match this string (not guaranteed to be folded) using /iaa rules (w/len). */ +#define EXACTFU_SS 39 /* 0x27 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 40 /* 0x28 Match this folded UTF-8 string using /iu rules */ +#define NOTHING 41 /* 0x29 Match empty string. */ +#define TAIL 42 /* 0x2a Match empty string. Can jump here from outside. */ +#define STAR 43 /* 0x2b Match this (simple) thing 0 or more times. */ +#define PLUS 44 /* 0x2c Match this (simple) thing 1 or more times. */ +#define CURLY 45 /* 0x2d Match this simple thing {n,m} times. */ +#define CURLYN 46 /* 0x2e Capture next-after-this simple thing */ +#define CURLYM 47 /* 0x2f Capture this medium-complex thing {n,m} times. */ +#define CURLYX 48 /* 0x30 Match this complex thing {n,m} times. */ +#define WHILEM 49 /* 0x31 Do curly processing and see if rest matches. */ +#define OPEN 50 /* 0x32 Mark this point in input as start of #n. */ +#define CLOSE 51 /* 0x33 Analogous to OPEN. */ +#define REF 52 /* 0x34 Match some already matched string */ +#define REFF 53 /* 0x35 Match already matched string, folded using native charset semantics for non-utf8 */ +#define REFFL 54 /* 0x36 Match already matched string, folded in loc. */ +#define REFFU 55 /* 0x37 Match already matched string, folded using unicode semantics for non-utf8 */ +#define REFFA 56 /* 0x38 Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#define NREF 57 /* 0x39 Match some already matched string */ +#define NREFF 58 /* 0x3a Match already matched string, folded using native charset semantics for non-utf8 */ +#define NREFFL 59 /* 0x3b Match already matched string, folded in loc. */ +#define NREFFU 60 /* 0x3c Match already matched string, folded using unicode semantics for non-utf8 */ +#define NREFFA 61 /* 0x3d Match already matched string, folded using unicode semantics for non-utf8, no mixing ASCII, non-ASCII */ +#define IFMATCH 62 /* 0x3e Succeeds if the following matches. */ +#define UNLESSM 63 /* 0x3f Fails if the following matches. */ +#define SUSPEND 64 /* 0x40 "Independent" sub-RE. */ +#define IFTHEN 65 /* 0x41 Switch, should be preceded by switcher. */ +#define GROUPP 66 /* 0x42 Whether the group matched. */ +#define LONGJMP 67 /* 0x43 Jump far away. */ +#define BRANCHJ 68 /* 0x44 BRANCH with long offset. */ +#define EVAL 69 /* 0x45 Execute some Perl code. */ +#define MINMOD 70 /* 0x46 Next operator is not greedy. */ +#define LOGICAL 71 /* 0x47 Next opcode should set the flag only. */ +#define RENUM 72 /* 0x48 Group with independently numbered parens. */ +#define TRIE 73 /* 0x49 Match many EXACT(F[ALU]?)? at once. flags==type */ +#define TRIEC 74 /* 0x4a Same as TRIE, but with embedded charclass data */ +#define AHOCORASICK 75 /* 0x4b Aho Corasick stclass. flags==type */ +#define AHOCORASICKC 76 /* 0x4c Same as AHOCORASICK, but with embedded charclass data */ +#define GOSUB 77 /* 0x4d recurse to paren arg1 at (signed) ofs arg2 */ +#define GOSTART 78 /* 0x4e recurse to start of pattern */ +#define NGROUPP 79 /* 0x4f Whether the group matched. */ +#define INSUBP 80 /* 0x50 Whether we are in a specific recurse. */ +#define DEFINEP 81 /* 0x51 Never execute directly. */ +#define ENDLIKE 82 /* 0x52 Used only for the type field of verbs */ +#define OPFAIL 83 /* 0x53 Same as (?!) */ +#define ACCEPT 84 /* 0x54 Accepts the current matched string. */ +#define VERB 85 /* 0x55 Used only for the type field of verbs */ +#define PRUNE 86 /* 0x56 Pattern fails at this startpoint if no-backtracking through this */ +#define MARKPOINT 87 /* 0x57 Push the current location for rollback by cut. */ +#define SKIP 88 /* 0x58 On failure skip forward (to the mark) before retrying */ +#define COMMIT 89 /* 0x59 Pattern fails outright if backtracking through this */ +#define CUTGROUP 90 /* 0x5a On failure go to the next alternation in the group */ +#define KEEPS 91 /* 0x5b $& begins here. */ +#define LNBREAK 92 /* 0x5c generic newline pattern */ +#define OPTIMIZED 93 /* 0x5d Placeholder for dump. */ +#define PSEUDO 94 /* 0x5e 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 */ @@ -173,6 +174,7 @@ EXTCONST U8 PL_regkind[] = { REG_ANY, /* SANY */ REG_ANY, /* CANY */ ANYOF, /* ANYOF */ + ANYOF, /* ANYOF_WARN_SUPER */ POSIXD, /* POSIXD */ POSIXD, /* POSIXL */ POSIXD, /* POSIXU */ @@ -315,6 +317,7 @@ static const U8 regarglen[] = { 0, /* SANY */ 0, /* CANY */ 0, /* ANYOF */ + 0, /* ANYOF_WARN_SUPER */ 0, /* POSIXD */ 0, /* POSIXL */ 0, /* POSIXU */ @@ -414,6 +417,7 @@ static const char reg_off_by_arg[] = { 0, /* SANY */ 0, /* CANY */ 0, /* ANYOF */ + 0, /* ANYOF_WARN_SUPER */ 0, /* POSIXD */ 0, /* POSIXL */ 0, /* POSIXU */ @@ -518,78 +522,79 @@ EXTCONST char * const PL_reg_name[] = { "SANY", /* 0x13 */ "CANY", /* 0x14 */ "ANYOF", /* 0x15 */ - "POSIXD", /* 0x16 */ - "POSIXL", /* 0x17 */ - "POSIXU", /* 0x18 */ - "POSIXA", /* 0x19 */ - "NPOSIXD", /* 0x1a */ - "NPOSIXL", /* 0x1b */ - "NPOSIXU", /* 0x1c */ - "NPOSIXA", /* 0x1d */ - "CLUMP", /* 0x1e */ - "BRANCH", /* 0x1f */ - "BACK", /* 0x20 */ - "EXACT", /* 0x21 */ - "EXACTF", /* 0x22 */ - "EXACTFL", /* 0x23 */ - "EXACTFU", /* 0x24 */ - "EXACTFA", /* 0x25 */ - "EXACTFU_SS", /* 0x26 */ - "EXACTFU_TRICKYFOLD", /* 0x27 */ - "NOTHING", /* 0x28 */ - "TAIL", /* 0x29 */ - "STAR", /* 0x2a */ - "PLUS", /* 0x2b */ - "CURLY", /* 0x2c */ - "CURLYN", /* 0x2d */ - "CURLYM", /* 0x2e */ - "CURLYX", /* 0x2f */ - "WHILEM", /* 0x30 */ - "OPEN", /* 0x31 */ - "CLOSE", /* 0x32 */ - "REF", /* 0x33 */ - "REFF", /* 0x34 */ - "REFFL", /* 0x35 */ - "REFFU", /* 0x36 */ - "REFFA", /* 0x37 */ - "NREF", /* 0x38 */ - "NREFF", /* 0x39 */ - "NREFFL", /* 0x3a */ - "NREFFU", /* 0x3b */ - "NREFFA", /* 0x3c */ - "IFMATCH", /* 0x3d */ - "UNLESSM", /* 0x3e */ - "SUSPEND", /* 0x3f */ - "IFTHEN", /* 0x40 */ - "GROUPP", /* 0x41 */ - "LONGJMP", /* 0x42 */ - "BRANCHJ", /* 0x43 */ - "EVAL", /* 0x44 */ - "MINMOD", /* 0x45 */ - "LOGICAL", /* 0x46 */ - "RENUM", /* 0x47 */ - "TRIE", /* 0x48 */ - "TRIEC", /* 0x49 */ - "AHOCORASICK", /* 0x4a */ - "AHOCORASICKC", /* 0x4b */ - "GOSUB", /* 0x4c */ - "GOSTART", /* 0x4d */ - "NGROUPP", /* 0x4e */ - "INSUBP", /* 0x4f */ - "DEFINEP", /* 0x50 */ - "ENDLIKE", /* 0x51 */ - "OPFAIL", /* 0x52 */ - "ACCEPT", /* 0x53 */ - "VERB", /* 0x54 */ - "PRUNE", /* 0x55 */ - "MARKPOINT", /* 0x56 */ - "SKIP", /* 0x57 */ - "COMMIT", /* 0x58 */ - "CUTGROUP", /* 0x59 */ - "KEEPS", /* 0x5a */ - "LNBREAK", /* 0x5b */ - "OPTIMIZED", /* 0x5c */ - "PSEUDO", /* 0x5d */ + "ANYOF_WARN_SUPER", /* 0x16 */ + "POSIXD", /* 0x17 */ + "POSIXL", /* 0x18 */ + "POSIXU", /* 0x19 */ + "POSIXA", /* 0x1a */ + "NPOSIXD", /* 0x1b */ + "NPOSIXL", /* 0x1c */ + "NPOSIXU", /* 0x1d */ + "NPOSIXA", /* 0x1e */ + "CLUMP", /* 0x1f */ + "BRANCH", /* 0x20 */ + "BACK", /* 0x21 */ + "EXACT", /* 0x22 */ + "EXACTF", /* 0x23 */ + "EXACTFL", /* 0x24 */ + "EXACTFU", /* 0x25 */ + "EXACTFA", /* 0x26 */ + "EXACTFU_SS", /* 0x27 */ + "EXACTFU_TRICKYFOLD", /* 0x28 */ + "NOTHING", /* 0x29 */ + "TAIL", /* 0x2a */ + "STAR", /* 0x2b */ + "PLUS", /* 0x2c */ + "CURLY", /* 0x2d */ + "CURLYN", /* 0x2e */ + "CURLYM", /* 0x2f */ + "CURLYX", /* 0x30 */ + "WHILEM", /* 0x31 */ + "OPEN", /* 0x32 */ + "CLOSE", /* 0x33 */ + "REF", /* 0x34 */ + "REFF", /* 0x35 */ + "REFFL", /* 0x36 */ + "REFFU", /* 0x37 */ + "REFFA", /* 0x38 */ + "NREF", /* 0x39 */ + "NREFF", /* 0x3a */ + "NREFFL", /* 0x3b */ + "NREFFU", /* 0x3c */ + "NREFFA", /* 0x3d */ + "IFMATCH", /* 0x3e */ + "UNLESSM", /* 0x3f */ + "SUSPEND", /* 0x40 */ + "IFTHEN", /* 0x41 */ + "GROUPP", /* 0x42 */ + "LONGJMP", /* 0x43 */ + "BRANCHJ", /* 0x44 */ + "EVAL", /* 0x45 */ + "MINMOD", /* 0x46 */ + "LOGICAL", /* 0x47 */ + "RENUM", /* 0x48 */ + "TRIE", /* 0x49 */ + "TRIEC", /* 0x4a */ + "AHOCORASICK", /* 0x4b */ + "AHOCORASICKC", /* 0x4c */ + "GOSUB", /* 0x4d */ + "GOSTART", /* 0x4e */ + "NGROUPP", /* 0x4f */ + "INSUBP", /* 0x50 */ + "DEFINEP", /* 0x51 */ + "ENDLIKE", /* 0x52 */ + "OPFAIL", /* 0x53 */ + "ACCEPT", /* 0x54 */ + "VERB", /* 0x55 */ + "PRUNE", /* 0x56 */ + "MARKPOINT", /* 0x57 */ + "SKIP", /* 0x58 */ + "COMMIT", /* 0x59 */ + "CUTGROUP", /* 0x5a */ + "KEEPS", /* 0x5b */ + "LNBREAK", /* 0x5c */ + "OPTIMIZED", /* 0x5d */ + "PSEUDO", /* 0x5e */ /* ------------ States ------------- */ "TRIE_next", /* REGNODE_MAX +0x01 */ "TRIE_next_fail", /* REGNODE_MAX +0x02 */ @@ -694,7 +699,7 @@ EXTCONST U8 PL_varies[] __attribute__deprecated__ = { EXTCONST U8 PL_varies_bitmask[]; #else EXTCONST U8 PL_varies_bitmask[] = { - 0x00, 0x00, 0x00, 0xC0, 0x01, 0xFC, 0xF9, 0x9F, 0x09, 0x00, 0x00, 0x00 + 0x00, 0x00, 0x00, 0x80, 0x03, 0xF8, 0xF3, 0x3F, 0x13, 0x00, 0x00, 0x00 }; #endif /* DOINIT */ @@ -706,8 +711,8 @@ EXTCONST U8 PL_varies_bitmask[] = { EXTCONST U8 PL_simple[] __attribute__deprecated__; #else EXTCONST U8 PL_simple[] __attribute__deprecated__ = { - REG_ANY, SANY, CANY, ANYOF, POSIXD, POSIXL, POSIXU, POSIXA, NPOSIXD, - NPOSIXL, NPOSIXU, NPOSIXA, + REG_ANY, SANY, CANY, ANYOF, ANYOF_WARN_SUPER, POSIXD, POSIXL, POSIXU, + POSIXA, NPOSIXD, NPOSIXL, NPOSIXU, NPOSIXA, 0 }; #endif /* DOINIT */ @@ -716,7 +721,7 @@ EXTCONST U8 PL_simple[] __attribute__deprecated__ = { EXTCONST U8 PL_simple_bitmask[]; #else EXTCONST U8 PL_simple_bitmask[] = { - 0x00, 0x00, 0xFC, 0x3F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 + 0x00, 0x00, 0xFC, 0x7F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; #endif /* DOINIT */ |