/* regcomp.h
+ *
+ * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
+ * 2000, 2001, 2002, 2003, 2005 by Larry Wall and others
+ *
+ * You may distribute under the terms of either the GNU General Public
+ * License or the Artistic License, as specified in the README file.
+ *
*/
+typedef OP OP_4tree; /* Will be redefined later. */
+
+
+#define PERL_ENABLE_TRIE_OPTIMISATION 1
+#define PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION 1
+#define PERL_ENABLE_POSITIVE_ASSERTION_STUDY 1
+#define PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS 0
+/* Unless the next line is uncommented it is illegal to combine lazy
+ matching with possessive matching. Frankly it doesn't make much sense
+ to allow it as X*?+ matches nothing, X+?+ matches a single char only,
+ and X{min,max}?+ matches min times only.
+ */
+/* #define REG_ALLOW_MINMOD_SUSPEND */
+
/*
* The "internal use only" fields in regexp.h are present to pass info from
* compile to execute that permits the execute phase to run lots faster on
* simple cases. They are:
*
- * regstart sv that must begin a match; Nullch if none obvious
+ * regstart sv that must begin a match; NULL if none obvious
* reganch is the match anchored (at beginning-of-line only)?
* regmust string (pointer into program) that match must include, or NULL
* [regmust changed to SV* for bminstr()--law]
* a literal string; for others, it is a node leading into a sub-FSM. In
* particular, the operand of a BRANCH node is the first node of the branch.
* (NB this is *not* a tree structure: the tail of the branch connects
- * to the thing following the set of BRANCHes.) The opcodes are:
+ * to the thing following the set of BRANCHes.) The opcodes are defined
+ * in regnodes.h which is generated from regcomp.sym by regcomp.pl.
*/
-/* definition number opnd? meaning */
-#define END 0 /* no End of program. */
-#define BOL 1 /* no Match "" at beginning of line. */
-#define MBOL 2 /* no Same, assuming multiline. */
-#define SBOL 3 /* no Same, assuming singleline. */
-#define EOL 4 /* no Match "" at end of line. */
-#define MEOL 5 /* no Same, assuming multiline. */
-#define SEOL 6 /* no Same, assuming singleline. */
-#define ANY 7 /* no Match any one character (except newline). */
-#define SANY 8 /* no Match any one character. */
-#define ANYOF 9 /* sv Match character in (or not in) this class. */
-#define CURLY 10 /* sv Match this simple thing {n,m} times. */
-#define CURLYX 11 /* sv Match this complex thing {n,m} times. */
-#define BRANCH 12 /* node Match this alternative, or the next... */
-#define BACK 13 /* no Match "", "next" ptr points backward. */
-#define EXACT 14 /* sv Match this string (preceded by length). */
-#define EXACTF 15 /* sv Match this string, folded (prec. by length). */
-#define EXACTFL 16 /* sv Match this string, folded in locale (w/len). */
-#define NOTHING 17 /* no Match empty string. */
-#define STAR 18 /* node Match this (simple) thing 0 or more times. */
-#define PLUS 19 /* node Match this (simple) thing 1 or more times. */
-#define BOUND 20 /* no Match "" at any word boundary */
-#define BOUNDL 21 /* no Match "" at any word boundary */
-#define NBOUND 22 /* no Match "" at any word non-boundary */
-#define NBOUNDL 23 /* no Match "" at any word non-boundary */
-#define REF 24 /* num Match already matched string */
-#define REFF 25 /* num Match already matched string, folded */
-#define REFFL 26 /* num Match already matched string, folded in loc. */
-#define OPEN 27 /* num Mark this point in input as start of #n. */
-#define CLOSE 28 /* num Analogous to OPEN. */
-#define MINMOD 29 /* no Next operator is not greedy. */
-#define GPOS 30 /* no Matches where last m//g left off. */
-#define IFMATCH 31 /* no Succeeds if the following matches. */
-#define UNLESSM 32 /* no Fails if the following matches. */
-#define SUCCEED 33 /* no Return from a subroutine, basically. */
-#define WHILEM 34 /* no Do curly processing and see if rest matches. */
-#define ALNUM 35 /* no Match any alphanumeric character */
-#define ALNUML 36 /* no Match any alphanumeric char in locale */
-#define NALNUM 37 /* no Match any non-alphanumeric character */
-#define NALNUML 38 /* no Match any non-alphanumeric char in locale */
-#define SPACE 39 /* no Match any whitespace character */
-#define SPACEL 40 /* no Match any whitespace char in locale */
-#define NSPACE 41 /* no Match any non-whitespace character */
-#define NSPACEL 42 /* no Match any non-whitespace char in locale */
-#define DIGIT 43 /* no Match any numeric character */
-#define NDIGIT 44 /* no Match any non-numeric character */
-
/*
- * Opcode notes:
- *
- * BRANCH The set of branches constituting a single choice are hooked
- * together with their "next" pointers, since precedence prevents
- * anything being concatenated to any individual branch. The
- * "next" pointer of the last BRANCH in a choice points to the
- * thing following the whole choice. This is also where the
- * final "next" pointer of each individual branch points; each
- * branch starts with the operand node of a BRANCH node.
- *
- * BACK Normal "next" pointers all implicitly point forward; BACK
- * exists to make loop structures possible.
+ * A node is one char of opcode followed by two chars of "next" pointer.
+ * "Next" pointers are stored as two 8-bit pieces, high order first. The
+ * value is a positive offset from the opcode of the node containing it.
+ * An operand, if any, simply follows the node. (Note that much of the
+ * code generation knows about this implicit relationship.)
*
- * STAR,PLUS '?', and complex '*' and '+', are implemented as circular
- * BRANCH structures using BACK. Simple cases (one character
- * per match) are implemented with STAR and PLUS for speed
- * and to minimize recursive plunges.
+ * Using two bytes for the "next" pointer is vast overkill for most things,
+ * but allows patterns to get big without disasters.
*
- * OPEN,CLOSE ...are numbered at compile time.
+ * [The "next" pointer is always aligned on an even
+ * boundary, and reads the offset directly as a short. Also, there is no
+ * special test to reverse the sign of BACK pointers since the offset is
+ * stored negative.]
*/
-#ifndef DOINIT
-EXT char regarglen[];
-#else
-EXT char regarglen[] = {
- 0,0,0,0,0,0,0,0,0,0,
- /*CURLY*/ 4, /*CURLYX*/ 4,
- 0,0,0,0,0,0,0,0,0,0,0,0,
- /*REF*/ 2, 2, 2, /*OPEN*/ 2, /*CLOSE*/ 2,
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
+struct regnode_string {
+ U8 str_len;
+ U8 type;
+ U16 next_off;
+ char string[1];
+};
+
+/* Argument bearing node - workhorse,
+ arg1 is often for the data field */
+struct regnode_1 {
+ U8 flags;
+ U8 type;
+ U16 next_off;
+ U32 arg1;
+};
+
+/* Similar to a regnode_1 but with an extra signed argument */
+struct regnode_2L {
+ U8 flags;
+ U8 type;
+ U16 next_off;
+ U32 arg1;
+ I32 arg2;
+};
+
+/* 'Two field' -- Two 16 bit unsigned args */
+struct regnode_2 {
+ U8 flags;
+ U8 type;
+ U16 next_off;
+ U16 arg1;
+ U16 arg2;
+};
+
+
+#define ANYOF_BITMAP_SIZE 32 /* 256 b/(8 b/B) */
+#define ANYOF_CLASSBITMAP_SIZE 4 /* up to 32 (8*4) named classes */
+
+/* also used by trie */
+struct regnode_charclass {
+ U8 flags;
+ U8 type;
+ U16 next_off;
+ U32 arg1;
+ char bitmap[ANYOF_BITMAP_SIZE]; /* only compile-time */
};
+
+struct regnode_charclass_class { /* has [[:blah:]] classes */
+ U8 flags; /* should have ANYOF_CLASS here */
+ U8 type;
+ U16 next_off;
+ U32 arg1;
+ char bitmap[ANYOF_BITMAP_SIZE]; /* both compile-time */
+ char classflags[ANYOF_CLASSBITMAP_SIZE]; /* and run-time */
+};
+
+/* XXX fix this description.
+ Impose a limit of REG_INFTY on various pattern matching operations
+ to limit stack growth and to avoid "infinite" recursions.
+*/
+/* The default size for REG_INFTY is I16_MAX, which is the same as
+ SHORT_MAX (see perl.h). Unfortunately I16 isn't necessarily 16 bits
+ (see handy.h). On the Cray C90, sizeof(short)==4 and hence I16_MAX is
+ ((1<<31)-1), while on the Cray T90, sizeof(short)==8 and I16_MAX is
+ ((1<<63)-1). To limit stack growth to reasonable sizes, supply a
+ smaller default.
+ --Andy Dougherty 11 June 1998
+*/
+#if SHORTSIZE > 2
+# ifndef REG_INFTY
+# define REG_INFTY ((1<<15)-1)
+# endif
#endif
-#ifndef DOINIT
-EXT char regkind[];
+#ifndef REG_INFTY
+# define REG_INFTY I16_MAX
+#endif
+
+#define ARG_VALUE(arg) (arg)
+#define ARG__SET(arg,val) ((arg) = (val))
+
+#undef ARG
+#undef ARG1
+#undef ARG2
+
+#define ARG(p) ARG_VALUE(ARG_LOC(p))
+#define ARG1(p) ARG_VALUE(ARG1_LOC(p))
+#define ARG2(p) ARG_VALUE(ARG2_LOC(p))
+#define ARG2L(p) ARG_VALUE(ARG2L_LOC(p))
+
+#define ARG_SET(p, val) ARG__SET(ARG_LOC(p), (val))
+#define ARG1_SET(p, val) ARG__SET(ARG1_LOC(p), (val))
+#define ARG2_SET(p, val) ARG__SET(ARG2_LOC(p), (val))
+#define ARG2L_SET(p, val) ARG__SET(ARG2L_LOC(p), (val))
+
+#undef NEXT_OFF
+#undef NODE_ALIGN
+
+#define NEXT_OFF(p) ((p)->next_off)
+#define NODE_ALIGN(node)
+#define NODE_ALIGN_FILL(node) ((node)->flags = 0xde) /* deadbeef */
+
+#define SIZE_ALIGN NODE_ALIGN
+
+#undef OP
+#undef OPERAND
+#undef MASK
+#undef STRING
+
+#define OP(p) ((p)->type)
+#define OPERAND(p) (((struct regnode_string *)p)->string)
+#define MASK(p) ((char*)OPERAND(p))
+#define STR_LEN(p) (((struct regnode_string *)p)->str_len)
+#define STRING(p) (((struct regnode_string *)p)->string)
+#define STR_SZ(l) ((l + sizeof(regnode) - 1) / sizeof(regnode))
+#define NODE_SZ_STR(p) (STR_SZ(STR_LEN(p))+1)
+
+#undef NODE_ALIGN
+#undef ARG_LOC
+#undef NEXTOPER
+#undef PREVOPER
+
+#define NODE_ALIGN(node)
+#define ARG_LOC(p) (((struct regnode_1 *)p)->arg1)
+#define ARG1_LOC(p) (((struct regnode_2 *)p)->arg1)
+#define ARG2_LOC(p) (((struct regnode_2 *)p)->arg2)
+#define ARG2L_LOC(p) (((struct regnode_2L *)p)->arg2)
+
+#define NODE_STEP_REGNODE 1 /* sizeof(regnode)/sizeof(regnode) */
+#define EXTRA_STEP_2ARGS EXTRA_SIZE(struct regnode_2)
+
+#define NODE_STEP_B 4
+
+#define NEXTOPER(p) ((p) + NODE_STEP_REGNODE)
+#define PREVOPER(p) ((p) - NODE_STEP_REGNODE)
+
+#define FILL_ADVANCE_NODE(ptr, op) STMT_START { \
+ (ptr)->type = op; (ptr)->next_off = 0; (ptr)++; } STMT_END
+#define FILL_ADVANCE_NODE_ARG(ptr, op, arg) STMT_START { \
+ ARG_SET(ptr, arg); FILL_ADVANCE_NODE(ptr, op); (ptr) += 1; } STMT_END
+
+#define REG_MAGIC 0234
+
+#define SIZE_ONLY (RExC_emit == &PL_regdummy)
+
+/* Flags for node->flags of ANYOF */
+
+#define ANYOF_CLASS 0x08 /* has [[:blah:]] classes */
+#define ANYOF_INVERT 0x04
+#define ANYOF_FOLD 0x02
+#define ANYOF_LOCALE 0x01
+
+/* Used for regstclass only */
+#define ANYOF_EOS 0x10 /* Can match an empty string too */
+
+/* There is a character or a range past 0xff */
+#define ANYOF_UNICODE 0x20
+#define ANYOF_UNICODE_ALL 0x40 /* Can match any char past 0xff */
+
+/* size of node is large (includes class pointer) */
+#define ANYOF_LARGE 0x80
+
+/* Are there any runtime flags on in this node? */
+#define ANYOF_RUNTIME(s) (ANYOF_FLAGS(s) & 0x0f)
+
+#define ANYOF_FLAGS_ALL 0xff
+
+/* Character classes for node->classflags of ANYOF */
+/* Should be synchronized with a table in regprop() */
+/* 2n should pair with 2n+1 */
+
+#define ANYOF_ALNUM 0 /* \w, PL_utf8_alnum, utf8::IsWord, ALNUM */
+#define ANYOF_NALNUM 1
+#define ANYOF_SPACE 2 /* \s */
+#define ANYOF_NSPACE 3
+#define ANYOF_DIGIT 4
+#define ANYOF_NDIGIT 5
+#define ANYOF_ALNUMC 6 /* isalnum(3), utf8::IsAlnum, ALNUMC */
+#define ANYOF_NALNUMC 7
+#define ANYOF_ALPHA 8
+#define ANYOF_NALPHA 9
+#define ANYOF_ASCII 10
+#define ANYOF_NASCII 11
+#define ANYOF_CNTRL 12
+#define ANYOF_NCNTRL 13
+#define ANYOF_GRAPH 14
+#define ANYOF_NGRAPH 15
+#define ANYOF_LOWER 16
+#define ANYOF_NLOWER 17
+#define ANYOF_PRINT 18
+#define ANYOF_NPRINT 19
+#define ANYOF_PUNCT 20
+#define ANYOF_NPUNCT 21
+#define ANYOF_UPPER 22
+#define ANYOF_NUPPER 23
+#define ANYOF_XDIGIT 24
+#define ANYOF_NXDIGIT 25
+#define ANYOF_PSXSPC 26 /* POSIX space: \s plus the vertical tab */
+#define ANYOF_NPSXSPC 27
+#define ANYOF_BLANK 28 /* GNU extension: space and tab: non-vertical space */
+#define ANYOF_NBLANK 29
+
+#define ANYOF_MAX 32
+
+/* Backward source code compatibility. */
+
+#define ANYOF_ALNUML ANYOF_ALNUM
+#define ANYOF_NALNUML ANYOF_NALNUM
+#define ANYOF_SPACEL ANYOF_SPACE
+#define ANYOF_NSPACEL ANYOF_NSPACE
+
+/* Utility macros for the bitmap and classes of ANYOF */
+
+#define ANYOF_SIZE (sizeof(struct regnode_charclass))
+#define ANYOF_CLASS_SIZE (sizeof(struct regnode_charclass_class))
+
+#define ANYOF_FLAGS(p) ((p)->flags)
+
+#define ANYOF_BIT(c) (1 << ((c) & 7))
+
+#define ANYOF_CLASS_BYTE(p, c) (((struct regnode_charclass_class*)(p))->classflags[((c) >> 3) & 3])
+#define ANYOF_CLASS_SET(p, c) (ANYOF_CLASS_BYTE(p, c) |= ANYOF_BIT(c))
+#define ANYOF_CLASS_CLEAR(p, c) (ANYOF_CLASS_BYTE(p, c) &= ~ANYOF_BIT(c))
+#define ANYOF_CLASS_TEST(p, c) (ANYOF_CLASS_BYTE(p, c) & ANYOF_BIT(c))
+
+#define ANYOF_CLASS_ZERO(ret) Zero(((struct regnode_charclass_class*)(ret))->classflags, ANYOF_CLASSBITMAP_SIZE, char)
+#define ANYOF_BITMAP_ZERO(ret) Zero(((struct regnode_charclass*)(ret))->bitmap, ANYOF_BITMAP_SIZE, char)
+
+#define ANYOF_BITMAP(p) (((struct regnode_charclass*)(p))->bitmap)
+#define ANYOF_BITMAP_BYTE(p, c) (ANYOF_BITMAP(p)[(((U8)(c)) >> 3) & 31])
+#define ANYOF_BITMAP_SET(p, c) (ANYOF_BITMAP_BYTE(p, c) |= ANYOF_BIT(c))
+#define ANYOF_BITMAP_CLEAR(p,c) (ANYOF_BITMAP_BYTE(p, c) &= ~ANYOF_BIT(c))
+#define ANYOF_BITMAP_TEST(p, c) (ANYOF_BITMAP_BYTE(p, c) & ANYOF_BIT(c))
+
+#define ANYOF_BITMAP_SETALL(p) \
+ memset (ANYOF_BITMAP(p), 255, ANYOF_BITMAP_SIZE)
+#define ANYOF_BITMAP_CLEARALL(p) \
+ Zero (ANYOF_BITMAP(p), ANYOF_BITMAP_SIZE)
+/* Check that all 256 bits are all set. Used in S_cl_is_anything() */
+#define ANYOF_BITMAP_TESTALLSET(p) \
+ memEQ (ANYOF_BITMAP(p), "\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377\377", ANYOF_BITMAP_SIZE)
+
+#define ANYOF_SKIP ((ANYOF_SIZE - 1)/sizeof(regnode))
+#define ANYOF_CLASS_SKIP ((ANYOF_CLASS_SIZE - 1)/sizeof(regnode))
+#define ANYOF_CLASS_ADD_SKIP (ANYOF_CLASS_SKIP - ANYOF_SKIP)
+
+
+/*
+ * Utility definitions.
+ */
+#ifndef CHARMASK
+# define UCHARAT(p) ((int)*(const U8*)(p))
#else
-EXT char regkind[] = {
- END,
- BOL,
- BOL,
- BOL,
- EOL,
- EOL,
- EOL,
- ANY,
- ANY,
- ANYOF,
- CURLY,
- CURLY,
- BRANCH,
- BACK,
- EXACT,
- EXACT,
- EXACT,
- NOTHING,
- STAR,
- PLUS,
- BOUND,
- BOUND,
- NBOUND,
- NBOUND,
- REF,
- REF,
- REF,
- OPEN,
- CLOSE,
- MINMOD,
- GPOS,
- BRANCH,
- BRANCH,
- END,
- WHILEM,
- ALNUM,
- ALNUM,
- NALNUM,
- NALNUM,
- SPACE,
- SPACE,
- NSPACE,
- NSPACE,
- DIGIT,
- NDIGIT,
-};
+# define UCHARAT(p) ((int)*(p)&CHARMASK)
+#endif
+
+#define EXTRA_SIZE(guy) ((sizeof(guy)-1)/sizeof(struct regnode))
+
+#define REG_SEEN_ZERO_LEN 0x00000001
+#define REG_SEEN_LOOKBEHIND 0x00000002
+#define REG_SEEN_GPOS 0x00000004
+#define REG_SEEN_EVAL 0x00000008
+#define REG_SEEN_CANY 0x00000010
+#define REG_SEEN_SANY REG_SEEN_CANY /* src bckwrd cmpt */
+#define REG_SEEN_RECURSE 0x00000020
+#define REG_TOP_LEVEL_BRANCHES 0x00000040
+
+START_EXTERN_C
+
+#ifdef PLUGGABLE_RE_EXTENSION
+#include "re_nodes.h"
+#else
+#include "regnodes.h"
#endif
-/* The following have no fixed length. */
+/* The following have no fixed length. U8 so we can do strchr() on it. */
#ifndef DOINIT
-EXT char varies[];
+EXTCONST U8 PL_varies[];
#else
-EXT char varies[] = {
- BRANCH, BACK, STAR, PLUS, CURLY, CURLYX, REF, REFF, REFFL, WHILEM, 0
+EXTCONST U8 PL_varies[] = {
+ BRANCH, BACK, STAR, PLUS, CURLY, CURLYX, REF, REFF, REFFL,
+ WHILEM, CURLYM, CURLYN, BRANCHJ, IFTHEN, SUSPEND, CLUMP, 0
};
#endif
-/* The following always have a length of 1. */
+/* The following always have a length of 1. U8 we can do strchr() on it. */
+/* (Note that length 1 means "one character" under UTF8, not "one octet".) */
#ifndef DOINIT
-EXT char simple[];
+EXTCONST U8 PL_simple[];
#else
-EXT char simple[] = {
- ANY, SANY, ANYOF,
- ALNUM, ALNUML, NALNUM, NALNUML,
- SPACE, SPACEL, NSPACE, NSPACEL,
- DIGIT, NDIGIT, 0
+EXTCONST U8 PL_simple[] = {
+ REG_ANY, SANY, CANY,
+ ANYOF,
+ ALNUM, ALNUML,
+ NALNUM, NALNUML,
+ SPACE, SPACEL,
+ NSPACE, NSPACEL,
+ DIGIT, NDIGIT,
+ 0
};
#endif
-EXT char regdummy;
+#ifndef PLUGGABLE_RE_EXTENSION
+#ifndef DOINIT
+EXTCONST regexp_engine PL_core_reg_engine;
+#else /* DOINIT */
+EXTCONST regexp_engine PL_core_reg_engine = {
+ Perl_pregcomp,
+ Perl_regexec_flags,
+ Perl_re_intuit_start,
+ Perl_re_intuit_string,
+ Perl_pregfree,
+#if defined(USE_ITHREADS)
+ Perl_regdupe
+#endif
+};
+#endif /* DOINIT */
+#endif /* PLUGGABLE_RE_EXTENSION */
-/*
- * A node is one char of opcode followed by two chars of "next" pointer.
- * "Next" pointers are stored as two 8-bit pieces, high order first. The
- * value is a positive offset from the opcode of the node containing it.
- * An operand, if any, simply follows the node. (Note that much of the
- * code generation knows about this implicit relationship.)
- *
- * Using two bytes for the "next" pointer is vast overkill for most things,
- * but allows patterns to get big without disasters.
- *
- * [If REGALIGN is defined, the "next" pointer is always aligned on an even
- * boundary, and reads the offset directly as a short. Also, there is no
- * special test to reverse the sign of BACK pointers since the offset is
- * stored negative.]
+
+END_EXTERN_C
+
+
+/* .what is a character array with one character for each member of .data
+ * The character describes the function of the corresponding .data item:
+ * f - start-class data for regstclass optimization
+ * n - Root of op tree for (?{EVAL}) item
+ * o - Start op for (?{EVAL}) item
+ * p - Pad for (?{EVAL}) item
+ * s - swash for unicode-style character class, and the multicharacter
+ * strings resulting from casefolding the single-character entries
+ * in the character class
+ * t - trie struct
+ * T - aho-trie struct
+ * S - sv for named capture lookup
+ * 20010712 mjd@plover.com
+ * (Remember to update re_dup() and pregfree() if you add any items.)
*/
+struct reg_data {
+ U32 count;
+ U8 *what;
+ void* data[1];
+};
-#ifndef gould
-#ifndef cray
-#ifndef eta10
-#define REGALIGN
-#endif
-#endif
-#endif
+struct reg_substr_datum {
+ I32 min_offset;
+ I32 max_offset;
+ SV *substr; /* non-utf8 variant */
+ SV *utf8_substr; /* utf8 variant */
+ I32 end_shift;
+};
-#define OP(p) (*(p))
+struct reg_substr_data {
+ struct reg_substr_datum data[3]; /* Actual array */
+};
-#ifndef lint
-#ifdef REGALIGN
-#define NEXT(p) (*(short*)(p+1))
-#define ARG1(p) (*(unsigned short*)(p+3))
-#define ARG2(p) (*(unsigned short*)(p+5))
-#else
-#define NEXT(p) (((*((p)+1)&0377)<<8) + (*((p)+2)&0377))
-#define ARG1(p) (((*((p)+3)&0377)<<8) + (*((p)+4)&0377))
-#define ARG2(p) (((*((p)+5)&0377)<<8) + (*((p)+6)&0377))
-#endif
-#else /* lint */
-#define NEXT(p) 0
-#endif /* lint */
+#define anchored_substr substrs->data[0].substr
+#define anchored_utf8 substrs->data[0].utf8_substr
+#define anchored_offset substrs->data[0].min_offset
+#define anchored_end_shift substrs->data[0].end_shift
-#define OPERAND(p) ((p) + 3)
+#define float_substr substrs->data[1].substr
+#define float_utf8 substrs->data[1].utf8_substr
+#define float_min_offset substrs->data[1].min_offset
+#define float_max_offset substrs->data[1].max_offset
+#define float_end_shift substrs->data[1].end_shift
-#ifdef REGALIGN
-#define NEXTOPER(p) ((p) + 4)
-#define PREVOPER(p) ((p) - 4)
-#else
-#define NEXTOPER(p) ((p) + 3)
-#define PREVOPER(p) ((p) - 3)
+#define check_substr substrs->data[2].substr
+#define check_utf8 substrs->data[2].utf8_substr
+#define check_offset_min substrs->data[2].min_offset
+#define check_offset_max substrs->data[2].max_offset
+#define check_end_shift substrs->data[2].end_shift
+
+
+
+/* trie related stuff */
+
+/* a transition record for the state machine. the
+ check field determines which state "owns" the
+ transition. the char the transition is for is
+ determined by offset from the owning states base
+ field. the next field determines which state
+ is to be transitioned to if any.
+*/
+struct _reg_trie_trans {
+ U32 next;
+ U32 check;
+};
+
+/* a transition list element for the list based representation */
+struct _reg_trie_trans_list_elem {
+ U16 forid;
+ U32 newstate;
+};
+typedef struct _reg_trie_trans_list_elem reg_trie_trans_le;
+
+/* a state for compressed nodes. base is an offset
+ into an array of reg_trie_trans array. If wordnum is
+ nonzero the state is accepting. if base is zero then
+ the state has no children (and will be accepting)
+*/
+struct _reg_trie_state {
+ U16 wordnum;
+ union {
+ U32 base;
+ reg_trie_trans_le* list;
+ } trans;
+};
+
+
+
+typedef struct _reg_trie_state reg_trie_state;
+typedef struct _reg_trie_trans reg_trie_trans;
+
+
+/* anything in here that needs to be freed later
+ should be dealt with in pregfree */
+struct _reg_trie_data {
+ U16 uniquecharcount; /* unique chars in trie (width of trans table) */
+ U32 lasttrans; /* last valid transition element */
+ U16 *charmap; /* byte to charid lookup array */
+ HV *widecharmap; /* code points > 255 to charid */
+ reg_trie_state *states; /* state data */
+ reg_trie_trans *trans; /* array of transition elements */
+ char *bitmap; /* stclass bitmap */
+ U32 refcount; /* number of times this trie is referenced */
+ U32 startstate; /* initial state - used for common prefix optimisation */
+ STRLEN minlen; /* minimum length of words in trie - build/opt only? */
+ STRLEN maxlen; /* maximum length of words in trie - build/opt only? */
+ U32 *wordlen; /* array of lengths of words */
+ U16 *jump; /* optional 1 indexed array of offsets before tail
+ for the node following a given word. */
+ U16 *nextword; /* optional 1 indexed array to support linked list
+ of duplicate wordnums */
+ U32 statecount; /* Build only - number of states in the states array
+ (including the unused zero state) */
+ U32 wordcount; /* Build only */
+#ifdef DEBUGGING
+ STRLEN charcount; /* Build only */
+ AV *words; /* Array of words contained in trie, for dumping */
+ AV *revcharmap; /* Map of each charid back to its character representation */
#endif
+};
+typedef struct _reg_trie_data reg_trie_data;
-#define MAGIC 0234
+struct _reg_ac_data {
+ U32 *fail;
+ reg_trie_state *states;
+ reg_trie_data *trie;
+ U32 refcount;
+};
+typedef struct _reg_ac_data reg_ac_data;
-/* Flags for first parameter byte of ANYOF */
-#define ANYOF_INVERT 0x40
-#define ANYOF_FOLD 0x20
-#define ANYOF_LOCALE 0x10
-#define ANYOF_ISA 0x0F
-#define ANYOF_ALNUML 0x08
-#define ANYOF_NALNUML 0x04
-#define ANYOF_SPACEL 0x02
-#define ANYOF_NSPACEL 0x01
+/* ANY_BIT doesnt use the structure, so we can borrow it here.
+ This is simpler than refactoring all of it as wed end up with
+ three different sets... */
-/*
- * Utility definitions.
- */
-#ifndef lint
-#ifndef CHARMASK
-#define UCHARAT(p) ((int)*(unsigned char *)(p))
+#define TRIE_BITMAP(p) (((reg_trie_data *)(p))->bitmap)
+#define TRIE_BITMAP_BYTE(p, c) (TRIE_BITMAP(p)[(((U8)(c)) >> 3) & 31])
+#define TRIE_BITMAP_SET(p, c) (TRIE_BITMAP_BYTE(p, c) |= ANYOF_BIT((U8)c))
+#define TRIE_BITMAP_CLEAR(p,c) (TRIE_BITMAP_BYTE(p, c) &= ~ANYOF_BIT((U8)c))
+#define TRIE_BITMAP_TEST(p, c) (TRIE_BITMAP_BYTE(p, c) & ANYOF_BIT((U8)c))
+
+#define IS_ANYOF_TRIE(op) ((op)==TRIEC || (op)==AHOCORASICKC)
+#define IS_TRIE_AC(op) ((op)>=AHOCORASICK)
+
+
+#define BITMAP_BYTE(p, c) (((U8*)p)[(((U8)(c)) >> 3) & 31])
+#define BITMAP_TEST(p, c) (BITMAP_BYTE(p, c) & ANYOF_BIT((U8)c))
+
+/* these defines assume uniquecharcount is the correct variable, and state may be evaluated twice */
+#define TRIE_NODENUM(state) (((state)-1)/(trie->uniquecharcount)+1)
+#define SAFE_TRIE_NODENUM(state) ((state) ? (((state)-1)/(trie->uniquecharcount)+1) : (state))
+#define TRIE_NODEIDX(state) ((state) ? (((state)-1)*(trie->uniquecharcount)+1) : (state))
+
+#ifdef DEBUGGING
+#define TRIE_CHARCOUNT(trie) ((trie)->charcount)
+#define TRIE_REVCHARMAP(trie) ((trie)->revcharmap)
#else
-#define UCHARAT(p) ((int)*(p)&CHARMASK)
+#define TRIE_CHARCOUNT(trie) (trie_charcount)
+#define TRIE_REVCHARMAP(trie) (trie_revcharmap)
#endif
-#else /* lint */
-#define UCHARAT(p) regdummy
-#endif /* lint */
-#define FAIL(m) croak("/%.127s/: %s",regprecomp,m)
+#define RE_TRIE_MAXBUF_INIT 65536
+#define RE_TRIE_MAXBUF_NAME "\022E_TRIE_MAXBUF"
+#define RE_DEBUG_FLAGS "\022E_DEBUG_FLAGS"
+
+/*
+
+RE_DEBUG_FLAGS is used to control what debug output is emitted
+its divided into three groups of options, some of which interact.
+The three groups are: Compile, Execute, Extra. There is room for a
+further group, as currently only the low three bytes are used.
+
+ Compile Options:
+
+ PARSE
+ PEEP
+ TRIE
+ PROGRAM
+ OFFSETS
+
+ Execute Options:
+
+ INTUIT
+ MATCH
+ TRIE
+
+ Extra Options
+
+ TRIE
+ OFFSETS
+
+If you modify any of these make sure you make corresponding changes to
+re.pm, especially to the documentation.
+
+*/
+
+
+/* Compile */
+#define RE_DEBUG_COMPILE_MASK 0x0000FF
+#define RE_DEBUG_COMPILE_PARSE 0x000001
+#define RE_DEBUG_COMPILE_OPTIMISE 0x000002
+#define RE_DEBUG_COMPILE_TRIE 0x000004
+#define RE_DEBUG_COMPILE_DUMP 0x000008
+
+/* Execute */
+#define RE_DEBUG_EXECUTE_MASK 0x00FF00
+#define RE_DEBUG_EXECUTE_INTUIT 0x000100
+#define RE_DEBUG_EXECUTE_MATCH 0x000200
+#define RE_DEBUG_EXECUTE_TRIE 0x000400
+
+/* Extra */
+#define RE_DEBUG_EXTRA_MASK 0xFF0000
+#define RE_DEBUG_EXTRA_TRIE 0x010000
+#define RE_DEBUG_EXTRA_OFFSETS 0x020000
+#define RE_DEBUG_EXTRA_OFFDEBUG 0x040000
+#define RE_DEBUG_EXTRA_STATE 0x080000
+#define RE_DEBUG_EXTRA_OPTIMISE 0x100000
+
+#define RE_DEBUG_FLAG(x) (re_debug_flags & x)
+/* Compile */
+#define DEBUG_COMPILE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_COMPILE_MASK) x )
+#define DEBUG_PARSE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_COMPILE_PARSE) x )
+#define DEBUG_OPTIMISE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_COMPILE_OPTIMISE) x )
+#define DEBUG_PARSE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_COMPILE_PARSE) x )
+#define DEBUG_DUMP_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_COMPILE_DUMP) x )
+#define DEBUG_TRIE_COMPILE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_COMPILE_TRIE) x )
+
+/* Execute */
+#define DEBUG_EXECUTE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXECUTE_MASK) x )
+#define DEBUG_INTUIT_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXECUTE_INTUIT) x )
+#define DEBUG_MATCH_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXECUTE_MATCH) x )
+#define DEBUG_TRIE_EXECUTE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXECUTE_TRIE) x )
+
+/* Extra */
+#define DEBUG_EXTRA_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXTRA_MASK) x )
+#define DEBUG_OFFSETS_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXTRA_OFFSETS) x )
+#define DEBUG_STATE_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXTRA_STATE) x )
+#define DEBUG_OPTIMISE_MORE_r(x) DEBUG_r( \
+ if ((RE_DEBUG_EXTRA_OPTIMISE|RE_DEBUG_COMPILE_OPTIMISE) == \
+ (re_debug_flags & (RE_DEBUG_EXTRA_OPTIMISE|RE_DEBUG_COMPILE_OPTIMISE)) ) x )
+#define MJD_OFFSET_DEBUG(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_EXTRA_OFFDEBUG) \
+ Perl_warn_nocontext x )
+#define DEBUG_TRIE_COMPILE_MORE_r(x) DEBUG_TRIE_COMPILE_r( \
+ if (re_debug_flags & RE_DEBUG_EXTRA_TRIE) x )
+#define DEBUG_TRIE_EXECUTE_MORE_r(x) DEBUG_TRIE_EXECUTE_r( \
+ if (re_debug_flags & RE_DEBUG_EXTRA_TRIE) x )
+
+#define DEBUG_TRIE_r(x) DEBUG_r( \
+ if (re_debug_flags & (RE_DEBUG_COMPILE_TRIE \
+ | RE_DEBUG_EXECUTE_TRIE )) x )
+
+/* initialization */
+/* get_sv() can return NULL during global destruction. */
+#define GET_RE_DEBUG_FLAGS DEBUG_r({ \
+ SV * re_debug_flags_sv = NULL; \
+ re_debug_flags_sv = get_sv(RE_DEBUG_FLAGS, 1); \
+ if (re_debug_flags_sv) { \
+ if (!SvIOK(re_debug_flags_sv)) \
+ sv_setuv(re_debug_flags_sv, RE_DEBUG_COMPILE_DUMP | RE_DEBUG_EXECUTE_MASK ); \
+ re_debug_flags=SvIV(re_debug_flags_sv); \
+ }\
+})
+
+#ifdef DEBUGGING
+
+#define GET_RE_DEBUG_FLAGS_DECL IV re_debug_flags = 0; GET_RE_DEBUG_FLAGS;
+
+#define RE_PV_COLOR_DECL(rpv,rlen,isuni,dsv,pv,l,m,c1,c2) \
+ const char * const rpv = \
+ pv_pretty((dsv), (pv), (l), (m), \
+ PL_colors[(c1)],PL_colors[(c2)], \
+ ((isuni) ? PERL_PV_ESCAPE_UNI : 0) ); \
+ const int rlen = SvCUR(dsv)
+
+#define RE_SV_ESCAPE(rpv,isuni,dsv,sv,m) \
+ const char * const rpv = \
+ pv_pretty((dsv), (SvPV_nolen_const(sv)), (SvCUR(sv)), (m), \
+ PL_colors[(c1)],PL_colors[(c2)], \
+ ((isuni) ? PERL_PV_ESCAPE_UNI : 0) )
+
+#define RE_PV_QUOTED_DECL(rpv,isuni,dsv,pv,l,m) \
+ const char * const rpv = \
+ pv_pretty((dsv), (pv), (l), (m), \
+ PL_colors[0], PL_colors[1], \
+ ( PERL_PV_PRETTY_QUOTE | PERL_PV_PRETTY_ELIPSES | \
+ ((isuni) ? PERL_PV_ESCAPE_UNI : 0)) \
+ )
+
+#define RE_SV_DUMPLEN(ItEm) (SvCUR(ItEm) - (SvTAIL(ItEm)!=0))
+#define RE_SV_TAIL(ItEm) (SvTAIL(ItEm) ? "$" : "")
+
+#else /* if not DEBUGGING */
+
+#define GET_RE_DEBUG_FLAGS_DECL
+#define RE_PV_COLOR_DECL(rpv,rlen,isuni,dsv,pv,l,m,c1,c2)
+#define RE_SV_ESCAPE(rpv,isuni,dsv,sv,m)
+#define RE_PV_QUOTED_DECL(rpv,isuni,dsv,pv,l,m)
+#define RE_SV_DUMPLEN(ItEm)
+#define RE_SV_TAIL(ItEm)
+
+#endif /* DEBUG RELATED DEFINES */
+
https://perl5.git.perl.org / perl5.git / blobdiff