[perl5.git] / regcomp.h

/*    regcomp.h
 */

typedef OP OP_4tree;			/* Will be redefined later. */

/*
 * 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
 * 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]
 * regmlen	length of regmust string
 *  [regmlen not used currently]
 *
 * Regstart and reganch permit very fast decisions on suitable starting points
 * for a match, cutting down the work a lot.  Regmust permits fast rejection
 * of lines that cannot possibly match.  The regmust tests are costly enough
 * that pregcomp() supplies a regmust only if the r.e. contains something
 * potentially expensive (at present, the only such thing detected is * or +
 * at the start of the r.e., which can involve a lot of backup).  Regmlen is
 * supplied because the test in pregexec() needs it and pregcomp() is computing
 * it anyway.
 * [regmust is now supplied always.  The tests that use regmust have a
 * heuristic that disables the test if it usually matches.]
 *
 * [In fact, we now use regmust in many cases to locate where the search
 * starts in the string, so if regback is >= 0, the regmust search is never
 * wasted effort.  The regback variable says how many characters back from
 * where regmust matched is the earliest possible start of the match.
 * For instance, /[a-z].foo/ has a regmust of 'foo' and a regback of 2.]
 */

/*
 * Structure for regexp "program".  This is essentially a linear encoding
 * of a nondeterministic finite-state machine (aka syntax charts or
 * "railroad normal form" in parsing technology).  Each node is an opcode
 * plus a "next" pointer, possibly plus an operand.  "Next" pointers of
 * all nodes except BRANCH implement concatenation; a "next" pointer with
 * a BRANCH on both ends of it is connecting two alternatives.  (Here we
 * have one of the subtle syntax dependencies:  an individual BRANCH (as
 * opposed to a collection of them) is never concatenated with anything
 * because of operator precedence.)  The operand of some types of node is
 * 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:
 */

/*
 * 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.
 *
 * [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.]
 */

struct regnode_string {
    U8	flags;
    U8  type;
    U16 next_off;
    U8 string[1];
};

struct regnode_1 {
    U8	flags;
    U8  type;
    U16 next_off;
    U32 arg1;
};

struct regnode_2 {
    U8	flags;
    U8  type;
    U16 next_off;
    U16 arg1;
    U16 arg2;
};

/* 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 REG_INFTY
#  define REG_INFTY I16_MAX
#endif

#define ARG_VALUE(arg) (arg)
#define ARG__SET(arg,val) ((arg) = (val))

#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 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))

#ifndef lint
#  define NEXT_OFF(p) ((p)->next_off)
#  define NODE_ALIGN(node)
#  define NODE_ALIGN_FILL(node) ((node)->flags = 0xde) /* deadbeef */
#else /* lint */
#  define NEXT_OFF(p) 0
#  define NODE_ALIGN(node)
#  define NODE_ALIGN_FILL(node)
#endif /* lint */

#define SIZE_ALIGN NODE_ALIGN

#define	OP(p)		((p)->type)
#define	OPERAND(p)	(((struct regnode_string *)p)->string)
#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 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 (PL_regcode == &PL_regdummy)

/* 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

/* Utility macros for bitmap of ANYOF */
#define ANYOF_BYTE(p,c)     (p)[1 + (((c) >> 3) & 31)]
#define ANYOF_BIT(c)        (1 << ((c) & 7))
#define ANYOF_SET(p,c)      (ANYOF_BYTE(p,c) |=  ANYOF_BIT(c))
#define ANYOF_CLEAR(p,c)    (ANYOF_BYTE(p,c) &= ~ANYOF_BIT(c))
#define ANYOF_TEST(p,c)     (ANYOF_BYTE(p,c) &   ANYOF_BIT(c))

#define ANY_SKIP ((33 - 1)/sizeof(regnode) + 1)

/*
 * Utility definitions.
 */
#ifndef lint
#ifndef CHARMASK
#define	UCHARAT(p)	((int)*(unsigned char *)(p))
#else
#define	UCHARAT(p)	((int)*(p)&CHARMASK)
#endif
#else /* lint */
#define UCHARAT(p)	PL_regdummy
#endif /* lint */

#define	FAIL(m) \
    STMT_START {							\
	if (!SIZE_ONLY)							\
	    SAVEDESTRUCTOR(S_clear_re,(void*)PL_regcomp_rx);		\
	Perl_croak(aTHX_ "/%.127s/: %s",  PL_regprecomp,m);		\
    } STMT_END

#define	FAIL2(pat,m) \
    STMT_START {							\
	if (!SIZE_ONLY)							\
	    SAVEDESTRUCTOR(S_clear_re,(void*)PL_regcomp_rx);		\
	S_re_croak2(aTHX_ "/%.127s/: ",pat,PL_regprecomp,m);		\
    } STMT_END

#define EXTRA_SIZE(guy) ((sizeof(guy)-1)/sizeof(struct regnode))

#define REG_SEEN_ZERO_LEN	1
#define REG_SEEN_LOOKBEHIND	2
#define REG_SEEN_GPOS		4
#define REG_SEEN_EVAL		8

START_EXTERN_C

#include "regnodes.h"

/* The following have no fixed length. char* since we do strchr on it. */
#ifndef DOINIT
EXTCONST char PL_varies[];
#else
EXTCONST char 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. char* since we do strchr on it. */
/* (Note that lenght 1 means "one character" under UTF8, not "one octet".) */
#ifndef DOINIT
EXTCONST char PL_simple[];
#else
EXTCONST char PL_simple[] = {
    REG_ANY, ANYUTF8, SANY, SANYUTF8, ANYOF, ANYOFUTF8,
    ALNUM, ALNUMUTF8, ALNUML, ALNUMLUTF8,
    NALNUM, NALNUMUTF8, NALNUML, NALNUMLUTF8,
    SPACE, SPACEUTF8, SPACEL, SPACELUTF8,
    NSPACE, NSPACEUTF8, NSPACEL, NSPACELUTF8,
    DIGIT, DIGITUTF8, NDIGIT, NDIGITUTF8, 0
};
#endif

END_EXTERN_C

typedef struct re_scream_pos_data_s
{
    char **scream_olds;		/* match pos */
    I32 *scream_pos;		/* Internal iterator of scream. */
} re_scream_pos_data;

struct reg_data {
    U32 count;
    U8 *what;
    void* data[1];
};

struct reg_substr_datum {
    I32 min_offset;
    I32 max_offset;
    SV *substr;
};

struct reg_substr_data {
    struct reg_substr_datum data[3];	/* Actual array */
};

#define anchored_substr substrs->data[0].substr
#define anchored_offset substrs->data[0].min_offset
#define float_substr substrs->data[1].substr
#define float_min_offset substrs->data[1].min_offset
#define float_max_offset substrs->data[1].max_offset
#define check_substr substrs->data[2].substr
#define check_offset_min substrs->data[2].min_offset
#define check_offset_max substrs->data[2].max_offset
Commit	Line	Data
a0d0e21e	1	/* regcomp.h
a687059c LW	2	*/
a687059c LW	3
c277df42 IZ	4	typedef OP OP_4tree; /* Will be redefined later. */
c277df42 IZ	5
a687059c LW	6	/*
	7	* The "internal use only" fields in regexp.h are present to pass info from
	8	* compile to execute that permits the execute phase to run lots faster on
	9	* simple cases. They are:
	10	*
79072805	11	* regstart sv that must begin a match; Nullch if none obvious
a687059c LW	12	* reganch is the match anchored (at beginning-of-line only)?
a687059c LW	13	* regmust string (pointer into program) that match must include, or NULL
79072805	14	* [regmust changed to SV* for bminstr()--law]
a687059c LW	15	* regmlen length of regmust string
	16	* [regmlen not used currently]
	17	*
	18	* Regstart and reganch permit very fast decisions on suitable starting points
	19	* for a match, cutting down the work a lot. Regmust permits fast rejection
	20	* of lines that cannot possibly match. The regmust tests are costly enough
e50aee73	21	* that pregcomp() supplies a regmust only if the r.e. contains something
a687059c LW	22	* potentially expensive (at present, the only such thing detected is * or +
a687059c LW	23	* at the start of the r.e., which can involve a lot of backup). Regmlen is
e50aee73	24	* supplied because the test in pregexec() needs it and pregcomp() is computing
a687059c LW	25	* it anyway.
	26	* [regmust is now supplied always. The tests that use regmust have a
	27	* heuristic that disables the test if it usually matches.]
	28	*
	29	* [In fact, we now use regmust in many cases to locate where the search
	30	* starts in the string, so if regback is >= 0, the regmust search is never
	31	* wasted effort. The regback variable says how many characters back from
	32	* where regmust matched is the earliest possible start of the match.
	33	* For instance, /[a-z].foo/ has a regmust of 'foo' and a regback of 2.]
	34	*/
	35
	36	/*
	37	* Structure for regexp "program". This is essentially a linear encoding
	38	* of a nondeterministic finite-state machine (aka syntax charts or
	39	* "railroad normal form" in parsing technology). Each node is an opcode
	40	* plus a "next" pointer, possibly plus an operand. "Next" pointers of
	41	* all nodes except BRANCH implement concatenation; a "next" pointer with
	42	* a BRANCH on both ends of it is connecting two alternatives. (Here we
	43	* have one of the subtle syntax dependencies: an individual BRANCH (as
	44	* opposed to a collection of them) is never concatenated with anything
	45	* because of operator precedence.) The operand of some types of node is
	46	* a literal string; for others, it is a node leading into a sub-FSM. In
	47	* particular, the operand of a BRANCH node is the first node of the branch.
	48	* (NB this is not a tree structure: the tail of the branch connects
	49	* to the thing following the set of BRANCHes.) The opcodes are:
	50	*/
	51
a687059c LW	52	/*
	53	* A node is one char of opcode followed by two chars of "next" pointer.
	54	* "Next" pointers are stored as two 8-bit pieces, high order first. The
	55	* value is a positive offset from the opcode of the node containing it.
	56	* An operand, if any, simply follows the node. (Note that much of the
	57	* code generation knows about this implicit relationship.)
	58	*
	59	* Using two bytes for the "next" pointer is vast overkill for most things,
	60	* but allows patterns to get big without disasters.
	61	*
e91177ed	62	* [The "next" pointer is always aligned on an even
a687059c LW	63	* boundary, and reads the offset directly as a short. Also, there is no
	64	* special test to reverse the sign of BACK pointers since the offset is
	65	* stored negative.]
	66	*/
	67
c277df42 IZ	68	struct regnode_string {
	69	U8 flags;
	70	U8 type;
	71	U16 next_off;
	72	U8 string[1];
	73	};
a687059c	74
c277df42 IZ	75	struct regnode_1 {
	76	U8 flags;
	77	U8 type;
	78	U16 next_off;
	79	U32 arg1;
	80	};
	81
	82	struct regnode_2 {
	83	U8 flags;
	84	U8 type;
	85	U16 next_off;
	86	U16 arg1;
	87	U16 arg2;
	88	};
	89
9c7e81e8 AD	90	/* XXX fix this description.
	91	Impose a limit of REG_INFTY on various pattern matching operations
	92	to limit stack growth and to avoid "infinite" recursions.
	93	*/
	94	/* The default size for REG_INFTY is I16_MAX, which is the same as
	95	SHORT_MAX (see perl.h). Unfortunately I16 isn't necessarily 16 bits
	96	(see handy.h). On the Cray C90, sizeof(short)==4 and hence I16_MAX is
	97	((1<<31)-1), while on the Cray T90, sizeof(short)==8 and I16_MAX is
	98	((1<<63)-1). To limit stack growth to reasonable sizes, supply a
	99	smaller default.
	100	--Andy Dougherty 11 June 1998
	101	*/
	102	#if SHORTSIZE > 2
	103	# ifndef REG_INFTY
	104	# define REG_INFTY ((1<<15)-1)
	105	# endif
	106	#endif
	107
	108	#ifndef REG_INFTY
83cfe48c IZ	109	# define REG_INFTY I16_MAX
83cfe48c IZ	110	#endif
a687059c	111
e91177ed IZ	112	#define ARG_VALUE(arg) (arg)
e91177ed IZ	113	#define ARG__SET(arg,val) ((arg) = (val))
c277df42 IZ	114
	115	#define ARG(p) ARG_VALUE(ARG_LOC(p))
	116	#define ARG1(p) ARG_VALUE(ARG1_LOC(p))
	117	#define ARG2(p) ARG_VALUE(ARG2_LOC(p))
	118	#define ARG_SET(p, val) ARG__SET(ARG_LOC(p), (val))
	119	#define ARG1_SET(p, val) ARG__SET(ARG1_LOC(p), (val))
	120	#define ARG2_SET(p, val) ARG__SET(ARG2_LOC(p), (val))
	121
	122	#ifndef lint
e91177ed IZ	123	# define NEXT_OFF(p) ((p)->next_off)
	124	# define NODE_ALIGN(node)
	125	# define NODE_ALIGN_FILL(node) ((node)->flags = 0xde) /* deadbeef */
a687059c	126	#else /* lint */
c277df42	127	# define NEXT_OFF(p) 0
e91177ed IZ	128	# define NODE_ALIGN(node)
e91177ed IZ	129	# define NODE_ALIGN_FILL(node)
a687059c LW	130	#endif /* lint */
a687059c LW	131
c277df42 IZ	132	#define SIZE_ALIGN NODE_ALIGN
c277df42 IZ	133
e91177ed IZ	134	#define OP(p) ((p)->type)
	135	#define OPERAND(p) (((struct regnode_string *)p)->string)
	136	#define NODE_ALIGN(node)
	137	#define ARG_LOC(p) (((struct regnode_1 *)p)->arg1)
	138	#define ARG1_LOC(p) (((struct regnode_2 *)p)->arg1)
	139	#define ARG2_LOC(p) (((struct regnode_2 *)p)->arg2)
	140	#define NODE_STEP_REGNODE 1 /* sizeof(regnode)/sizeof(regnode) */
	141	#define EXTRA_STEP_2ARGS EXTRA_SIZE(struct regnode_2)
	142
	143	#define NODE_STEP_B 4
c277df42 IZ	144
	145	#define NEXTOPER(p) ((p) + NODE_STEP_REGNODE)
	146	#define PREVOPER(p) ((p) - NODE_STEP_REGNODE)
	147
e91177ed	148	#define FILL_ADVANCE_NODE(ptr, op) STMT_START { \
c277df42	149	(ptr)->type = op; (ptr)->next_off = 0; (ptr)++; } STMT_END
e91177ed	150	#define FILL_ADVANCE_NODE_ARG(ptr, op, arg) STMT_START { \
c277df42	151	ARG_SET(ptr, arg); FILL_ADVANCE_NODE(ptr, op); (ptr) += 1; } STMT_END
a687059c	152
22c35a8c	153	#define REG_MAGIC 0234
a687059c	154
3280af22	155	#define SIZE_ONLY (PL_regcode == &PL_regdummy)
c277df42	156
bbce6d69	157	/* Flags for first parameter byte of ANYOF */
	158	#define ANYOF_INVERT 0x40
	159	#define ANYOF_FOLD 0x20
	160	#define ANYOF_LOCALE 0x10
	161	#define ANYOF_ISA 0x0F
	162	#define ANYOF_ALNUML 0x08
	163	#define ANYOF_NALNUML 0x04
	164	#define ANYOF_SPACEL 0x02
	165	#define ANYOF_NSPACEL 0x01
	166
ae5c130c GS	167	/* Utility macros for bitmap of ANYOF */
	168	#define ANYOF_BYTE(p,c) (p)[1 + (((c) >> 3) & 31)]
	169	#define ANYOF_BIT(c) (1 << ((c) & 7))
	170	#define ANYOF_SET(p,c) (ANYOF_BYTE(p,c) \|= ANYOF_BIT(c))
	171	#define ANYOF_CLEAR(p,c) (ANYOF_BYTE(p,c) &= ~ANYOF_BIT(c))
	172	#define ANYOF_TEST(p,c) (ANYOF_BYTE(p,c) & ANYOF_BIT(c))
	173
c277df42	174	#define ANY_SKIP ((33 - 1)/sizeof(regnode) + 1)
c277df42	175
a687059c LW	176	/*
	177	* Utility definitions.
	178	*/
	179	#ifndef lint
2304df62	180	#ifndef CHARMASK
a687059c LW	181	#define UCHARAT(p) ((int)(unsigned char )(p))
a687059c LW	182	#else
2304df62	183	#define UCHARAT(p) ((int)*(p)&CHARMASK)
a687059c LW	184	#endif
a687059c LW	185	#else /* lint */
6b88bc9c	186	#define UCHARAT(p) PL_regdummy
a687059c LW	187	#endif /* lint */
a687059c LW	188
34184a49	189	#define FAIL(m) \
cea2e8a9 GS	190	STMT_START { \
	191	if (!SIZE_ONLY) \
	192	SAVEDESTRUCTOR(S_clear_re,(void*)PL_regcomp_rx); \
	193	Perl_croak(aTHX_ "/%.127s/: %s", PL_regprecomp,m); \
34184a49 GS	194	} STMT_END
	195
	196	#define FAIL2(pat,m) \
cea2e8a9 GS	197	STMT_START { \
	198	if (!SIZE_ONLY) \
	199	SAVEDESTRUCTOR(S_clear_re,(void*)PL_regcomp_rx); \
	200	S_re_croak2(aTHX_ "/%.127s/: ",pat,PL_regprecomp,m); \
34184a49	201	} STMT_END
c277df42 IZ	202
	203	#define EXTRA_SIZE(guy) ((sizeof(guy)-1)/sizeof(struct regnode))
	204
d09b2d29 IZ	205	#define REG_SEEN_ZERO_LEN 1
	206	#define REG_SEEN_LOOKBEHIND 2
	207	#define REG_SEEN_GPOS 4
	208	#define REG_SEEN_EVAL 8
	209
73c4f7a1 GS	210	START_EXTERN_C
73c4f7a1 GS	211
d09b2d29 IZ	212	#include "regnodes.h"
	213
	214	/* The following have no fixed length. char* since we do strchr on it. */
	215	#ifndef DOINIT
22c35a8c	216	EXTCONST char PL_varies[];
d09b2d29	217	#else
22c35a8c	218	EXTCONST char PL_varies[] = {
d09b2d29	219	BRANCH, BACK, STAR, PLUS, CURLY, CURLYX, REF, REFF, REFFL,
a0ed51b3	220	WHILEM, CURLYM, CURLYN, BRANCHJ, IFTHEN, SUSPEND, CLUMP, 0
c277df42	221	};
d09b2d29	222	#endif
c277df42	223
d09b2d29	224	/* The following always have a length of 1. char* since we do strchr on it. */
a0ed51b3	225	/* (Note that lenght 1 means "one character" under UTF8, not "one octet".) */
d09b2d29	226	#ifndef DOINIT
22c35a8c	227	EXTCONST char PL_simple[];
d09b2d29	228	#else
22c35a8c GS	229	EXTCONST char PL_simple[] = {
22c35a8c GS	230	REG_ANY, ANYUTF8, SANY, SANYUTF8, ANYOF, ANYOFUTF8,
a0ed51b3 LW	231	ALNUM, ALNUMUTF8, ALNUML, ALNUMLUTF8,
	232	NALNUM, NALNUMUTF8, NALNUML, NALNUMLUTF8,
	233	SPACE, SPACEUTF8, SPACEL, SPACELUTF8,
	234	NSPACE, NSPACEUTF8, NSPACEL, NSPACELUTF8,
	235	DIGIT, DIGITUTF8, NDIGIT, NDIGITUTF8, 0
c277df42 IZ	236	};
	237	#endif
	238
73c4f7a1	239	END_EXTERN_C
f722798b IZ	240
	241	typedef struct re_scream_pos_data_s
	242	{
	243	char *scream_olds; / match pos */
	244	I32 scream_pos; / Internal iterator of scream. */
	245	} re_scream_pos_data;
	246
	247	struct reg_data {
	248	U32 count;
	249	U8 *what;
	250	void* data[1];
	251	};
	252
	253	struct reg_substr_datum {
	254	I32 min_offset;
	255	I32 max_offset;
	256	SV *substr;
	257	};
	258
	259	struct reg_substr_data {
	260	struct reg_substr_datum data[3]; /* Actual array */
	261	};
	262
	263	#define anchored_substr substrs->data[0].substr
	264	#define anchored_offset substrs->data[0].min_offset
	265	#define float_substr substrs->data[1].substr
	266	#define float_min_offset substrs->data[1].min_offset
	267	#define float_max_offset substrs->data[1].max_offset
	268	#define check_substr substrs->data[2].substr
	269	#define check_offset_min substrs->data[2].min_offset
	270	#define check_offset_max substrs->data[2].max_offset