*/
#include "regcharclass.h"
-typedef OP OP_4tree; /* Will be redefined later. */
-
-
/* Convert branch sequences to more efficient trie ops? */
#define PERL_ENABLE_TRIE_OPTIMISATION 1
#define RE_TRACK_PATTERN_OFFSETS
#endif
-/* 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
* 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.]
+ * boundary, and reads the offset directly as a short.]
*/
/* This is the stuff that used to live in regexp.h that was truly
*
* See regexp.h for flags used externally to the regexp engine
*/
+#define RXp_INTFLAGS(rx) ((rx)->intflags)
+#define RX_INTFLAGS(prog) RXp_INTFLAGS(ReANY(prog))
+
#define PREGf_SKIP 0x00000001
#define PREGf_IMPLICIT 0x00000002 /* Converted .* to ^.* */
#define PREGf_NAUGHTY 0x00000004 /* how exponential is this pattern? */
#define PREGf_VERBARG_SEEN 0x00000008
#define PREGf_CUTGROUP_SEEN 0x00000010
+#define PREGf_USE_RE_EVAL 0x00000020 /* compiled with "use re 'eval'" */
+/* these used to be extflags, but are now intflags */
+#define PREGf_NOSCAN 0x00000040
+#define PREGf_CANY_SEEN 0x00000080
+#define PREGf_GPOS_SEEN 0x00000100
+#define PREGf_GPOS_FLOAT 0x00000200
+
+#define PREGf_ANCH_MBOL 0x00000400
+#define PREGf_ANCH_SBOL 0x00000800
+#define PREGf_ANCH_GPOS 0x00001000
+#define PREGf_ANCH \
+ ( PREGf_ANCH_SBOL | PREGf_ANCH_GPOS | PREGf_ANCH_MBOL )
/* this is where the old regcomp.h started */
U16 arg2;
};
-
-#define ANYOF_BITMAP_SIZE 32 /* 256 b/(8 b/B) */
-#define ANYOF_CLASSBITMAP_SIZE 4 /* up to 32 (8*4) named classes */
+/* This give the number of code points that can be in the bitmap of an ANYOF
+ * node. The shift number must currently be one of: 8..12. It can't be less
+ * than 8 (256) because some code relies on it being at least that. Above 12
+ * (4096), and you start running into warnings that some data structure widths
+ * have been exceeded, though the test suite as of this writing still passes
+ * for up through 16, which is as high as anyone would ever want to go,
+ * encompassing all of the Unicode BMP, and thus including all the economically
+ * important world scripts. At 12 most of them are: including Arabic,
+ * Cyrillic, Greek, Hebrew, Indian subcontinent, Latin, and Thai; but not Han,
+ * Japanese, nor Korean. (The regarglen structure in regnodes.h is a U8, and
+ * the trie types TRIEC and AHOCORASICKC are larger than U8 for shift values
+ * below above 12.) Be sure to benchmark before changing, as larger sizes do
+ * significantly slow down the test suite */
+#define NUM_ANYOF_CODE_POINTS (1 << 8)
+
+#define ANYOF_BITMAP_SIZE (NUM_ANYOF_CODE_POINTS / 8) /* 8 bits/Byte */
+
+/* Note that these form structs which are supersets of the next smaller one, by
+ * appending fields. Alignment problems can occur if one of those optional
+ * fields requires stricter alignment than the base struct. And formal
+ * parameters that can really be two or more of the structs should be
+ * declared as the smallest one it could be. See commit message for
+ * 7dcac5f6a5195002b55c935ee1d67f67e1df280b. Regnode allocation is done
+ * without regard to alignment, and changing it to would also require changing
+ * the code that inserts and deletes regnodes. The basic single-argument
+ * regnode has a U32, which is what reganode() allocates as a unit. Therefore
+ * no field can require stricter alignment than U32. */
/* also used by trie */
struct regnode_charclass {
U8 flags;
U8 type;
U16 next_off;
- U32 arg1; /* used as ptr in S_regclass */
+ U32 arg1; /* set by set_ANYOF_arg() */
char bitmap[ANYOF_BITMAP_SIZE]; /* only compile-time */
};
/* has runtime (locale) \d, \w, ..., [:posix:] classes */
struct regnode_charclass_class {
- U8 flags; /* ANYOF_CLASS bit must go here */
+ U8 flags; /* ANYOF_MATCHES_POSIXL bit must go here */
+ U8 type;
+ U16 next_off;
+ U32 arg1;
+ char bitmap[ANYOF_BITMAP_SIZE]; /* both compile-time ... */
+ U32 classflags; /* and run-time */
+};
+
+/* A synthetic start class (SSC); is a regnode_charclass_posixl_fold, plus an
+ * extra SV*, used only during its construction and which is not used by
+ * regexec.c. Note that the 'next_off' field is unused, as the SSC stands
+ * alone, so there is never a next node. Also, there is no alignment issue,
+ * becase these are declared or allocated as a complete unit so the compiler
+ * takes care of alignment. This is unlike the other regnodes which are
+ * allocated in terms of multiples of a single-argument regnode. SSC nodes can
+ * have a pointer field because there is no alignment issue, and because it is
+ * set to NULL after construction, before any cloning of the pattern */
+struct regnode_ssc {
+ U8 flags; /* ANYOF_MATCHES_POSIXL bit must go here */
U8 type;
U16 next_off;
- U32 arg1; /* used as ptr in S_regclass */
- char bitmap[ANYOF_BITMAP_SIZE]; /* both compile-time */
- char classflags[ANYOF_CLASSBITMAP_SIZE]; /* and run-time */
+ U32 arg1;
+ char bitmap[ANYOF_BITMAP_SIZE]; /* both compile-time ... */
+ U32 classflags; /* and run-time */
+
+ /* Auxiliary, only used during construction; NULL afterwards: list of code
+ * points matched */
+ SV* invlist;
};
+/* We take advantage of 'next_off' not otherwise being used in the SSC by
+ * actually using it: by setting it to 1. This allows us to test and
+ * distinguish between an SSC and other ANYOF node types, as 'next_off' cannot
+ * otherwise be 1, because it is the offset to the next regnode expressed in
+ * units of regnodes. Since an ANYOF node contains extra fields, it adds up
+ * to 12 regnode units on 32-bit systems, (hence the minimum this can be (if
+ * not 0) is 11 there. Even if things get tightly packed on a 64-bit system,
+ * it still would be more than 1. */
+#define set_ANYOF_SYNTHETIC(n) STMT_START{ OP(n) = ANYOF; \
+ NEXT_OFF(n) = 1; \
+ } STMT_END
+#define is_ANYOF_SYNTHETIC(n) (OP(n) == ANYOF && NEXT_OFF(n) == 1)
+
/* XXX fix this description.
Impose a limit of REG_INFTY on various pattern matching operations
to limit stack growth and to avoid "infinite" recursions.
#define NEXT_OFF(p) ((p)->next_off)
#define NODE_ALIGN(node)
-#define NODE_ALIGN_FILL(node) ((node)->flags = 0xde) /* deadbeef */
+/* the following define was set to 0xde in 075abff3
+ * as part of some linting logic. I have set it to 0
+ * as otherwise in every place where we /might/ set flags
+ * we have to set it 0 explicitly, which duplicates
+ * assignments and IMO adds an unacceptable level of
+ * surprise to working in the regex engine. If this
+ * is changed from 0 then at the very least make sure
+ * that SBOL for /^/ sets the flags to 0 explicitly.
+ * -- Yves */
+#define NODE_ALIGN_FILL(node) ((node)->flags = 0)
#define SIZE_ALIGN NODE_ALIGN
(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 FILL_ADVANCE_NODE_2L_ARG(ptr, op, arg1, arg2) \
+ STMT_START { \
+ ARG_SET(ptr, arg1); \
+ ARG2L_SET(ptr, arg2); \
+ FILL_ADVANCE_NODE(ptr, op); \
+ (ptr) += 2; \
+ } STMT_END
#define REG_MAGIC 0234
-#define SIZE_ONLY (RExC_emit == &PL_regdummy)
+#define SIZE_ONLY (RExC_emit == (regnode *) & RExC_emit_dummy)
+#define PASS1 SIZE_ONLY
+#define PASS2 (! SIZE_ONLY)
-/* If the bitmap doesn't fully represent what this ANYOF node can match, the
+/* If the bitmap fully represents what this ANYOF node can match, the
* ARG is set to this special value (since 0, 1, ... are legal, but will never
* reach this high). */
-#define ANYOF_NONBITMAP_EMPTY ((U32) -1)
-
-/* The information used to be stored as as combination of the ANYOF_UTF8 and
- * ANYOF_NONBITMAP_NON_UTF8 bits in the flags field, but was moved out of there
- * to free up a bit for other uses. This tries to hide the change from
- * existing code as much as possible. Now, the data structure that goes in ARG
- * is not allocated unless it is needed, and that is what is used to determine
- * if there is something outside the bitmap. The code now assumes that if
- * that structure exists, that any UTF-8 encoded string should be tried against
- * it, but a non-UTF8-encoded string will be tried only if the
- * ANYOF_NONBITMAP_NON_UTF8 bit is also set. */
-#define ANYOF_NONBITMAP(node) (ARG(node) != ANYOF_NONBITMAP_EMPTY)
-
-/* Flags for node->flags of ANYOF. These are in short supply, so some games
- * are done to share them, as described below. If necessary, the ANYOF_LOCALE
- * and ANYOF_CLASS bits could be shared with a space penalty for locale nodes,
- * but this isn't quite so easy, as the optimizer also uses ANYOF_CLASS.
- * Another option would be to push them into new nodes. E.g. there could be an
- * ANYOF_LOCALE node that would be in place of the flag of the same name.
- * Once the planned change to compile all the above-latin1 code points is done,
- * then the UNICODE_ALL bit can be freed up, with a small performance penalty.
- * If flags need to be added that are applicable to the synthetic start class
- * only, with some work, they could be put in the next-node field, or in an
- * unused bit of the classflags field. */
-
-#define ANYOF_LOCALE 0x01 /* /l modifier */
+#define ANYOF_ONLY_HAS_BITMAP ((U32) -1)
+
+/* Flags for node->flags of ANYOF. These are in short supply, with none
+ * currently available. The ABOVE_BITMAP_ALL bit could be freed up
+ * by resorting to creating a swash containing everything above 255. This
+ * introduces a performance penalty. An option that wouldn't slow things down
+ * would be to split one of the two LOC flags out into a separate
+ * node, like what was done with ANYOF_NON_UTF8_NON_ASCII_ALL in commit
+ * 34fdef848b1687b91892ba55e9e0c3430e0770f6 (but which was reverted because it
+ * wasn't the best option available at the time), and using a LOC flag is
+ * probably better than that commit anyway. But it could be reinstated if we
+ * need a bit. The LOC flags are only for /l nodes; the reverted commit was
+ * only for /d, so there are no combinatorial issues. The LOC flag to use is
+ * probably the POSIXL one. Now that there is an ANYOFL (locale) node, another
+ * option would be to make all of those include the POSIXL data structure,
+ * which would get rid of needing a separate POSIXL flag. But it would
+ * increase the size of all such nodes, so it's probably not as atractive as
+ * having an ANYOF_POSIXL node type. But if we did do it, note that not all 32
+ * bits of that extra space are used, one bit of that could be set aside for
+ * the LOC_FOLD flag, yielding yet another bit. This would require extra code
+ * for masking, so again not the most attractive solution.
+ *
+ * Several flags are not used in synthetic start class (SSC) nodes, so could be
+ * shared should new flags be needed for SSCs, like SSC_MATCHES_EMPTY_STRING
+ * now. */
-/* The fold is calculated and stored in the bitmap where possible at compile
- * time. However there are two cases where it isn't possible. These share
- * this bit: 1) under locale, where the actual folding varies depending on
- * what the locale is at the time of execution; and 2) where the folding is
- * specified in a swash, not the bitmap, such as characters which aren't
- * specified in the bitmap, or properties that aren't looked at at compile time
- */
-#define ANYOF_LOC_NONBITMAP_FOLD 0x02
+/* regexec.c is expecting this to be in the low bit */
+#define ANYOF_INVERT 0x01
-#define ANYOF_INVERT 0x04
+/* For the SSC node only, which cannot be inverted, so is shared with that bit.
+ * This is used only during regex compilation. */
+#define SSC_MATCHES_EMPTY_STRING ANYOF_INVERT
-/* Set if this is a struct regnode_charclass_class vs a regnode_charclass. This
+/* Are there things outside the bitmap that will match only if the target
+ * string is encoded in UTF-8? (This is not set if ANYOF_ABOVE_BITMAP_ALL is
+ * set) */
+#define ANYOF_HAS_UTF8_NONBITMAP_MATCHES 0x02
+
+/* The fold is calculated and stored in the bitmap where possible at compile
+ * time. However under locale, the actual folding varies depending on
+ * what the locale is at the time of execution, so it has to be deferred until
+ * then */
+#define ANYOF_LOC_FOLD 0x04
+
+/* Set if this is a regnode_charclass_posixl vs a regnode_charclass. This
* is used for runtime \d, \w, [:posix:], ..., which are used only in locale
* and the optimizer's synthetic start class. Non-locale \d, etc are resolved
* at compile-time */
-#define ANYOF_CLASS 0x08
-#define ANYOF_LARGE ANYOF_CLASS /* Same; name retained for back compat */
-
-/* EOS, meaning that it can match an empty string too, is used for the
- * synthetic start class only. */
-#define ANYOF_EOS 0x10
-
-/* ? Is this node the synthetic start class (ssc). This bit is shared with
- * ANYOF_EOS, as the latter is used only for the ssc, and then not used by
- * regexec.c. And, the code is structured so that if it is set, the ssc is
- * not used, so it is guaranteed to be 0 for the ssc by the time regexec.c
- * gets executed, and 0 for a non-ssc ANYOF node, as it only ever gets set for
- * a potential ssc candidate. Thus setting it to 1 after it has been
- * determined that the ssc will be used is not ambiguous */
-#define ANYOF_IS_SYNTHETIC ANYOF_EOS
+#define ANYOF_MATCHES_POSIXL 0x08
+
+/* Should we raise a warning if matching against an above-Unicode code point?
+ * */
+#define ANYOF_WARN_SUPER 0x10
/* Can match something outside the bitmap that isn't in utf8 */
-#define ANYOF_NONBITMAP_NON_UTF8 0x20
+#define ANYOF_HAS_NONBITMAP_NON_UTF8_MATCHES 0x20
-/* Matches every code point 0x100 and above*/
-#define ANYOF_UNICODE_ALL 0x40
+/* Matches every code point NUM_ANYOF_CODE_POINTS and above*/
+#define ANYOF_MATCHES_ALL_ABOVE_BITMAP 0x40
/* Match all Latin1 characters that aren't ASCII when the target string is not
* in utf8. */
-#define ANYOF_NON_UTF8_LATIN1_ALL 0x80
+#define ANYOF_MATCHES_ALL_NON_UTF8_NON_ASCII 0x80
+
+#define ANYOF_FLAGS_ALL (0xff)
-#define ANYOF_FLAGS_ALL 0xff
+#define ANYOF_LOCALE_FLAGS (ANYOF_LOC_FOLD | ANYOF_MATCHES_POSIXL)
-/* These are the flags that ANYOF_INVERT being set or not doesn't affect
- * whether they are operative or not. e.g., the node still has LOCALE
- * regardless of being inverted; whereas ANYOF_UNICODE_ALL means something
- * different if inverted */
-#define INVERSION_UNAFFECTED_FLAGS (ANYOF_LOCALE \
- |ANYOF_LOC_NONBITMAP_FOLD \
- |ANYOF_CLASS \
- |ANYOF_EOS \
- |ANYOF_NONBITMAP_NON_UTF8)
+/* These are the flags that apply to both regular ANYOF nodes and synthetic
+ * start class nodes during construction of the SSC. During finalization of
+ * the SSC, other of the flags could be added to it */
+#define ANYOF_COMMON_FLAGS (ANYOF_WARN_SUPER|ANYOF_HAS_UTF8_NONBITMAP_MATCHES)
/* 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 /* \d */
-#define ANYOF_NDIGIT 5
-#define ANYOF_ALNUMC 6 /* [[:alnum:]] 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
-
-/* pseudo classes, not stored in the class bitmap, but used as flags
+/* 2n should be the normal one, paired with its complement at 2n+1 */
+
+#define ANYOF_ALPHA ((_CC_ALPHA) * 2)
+#define ANYOF_NALPHA ((ANYOF_ALPHA) + 1)
+#define ANYOF_ALPHANUMERIC ((_CC_ALPHANUMERIC) * 2) /* [[:alnum:]] isalnum(3), utf8::IsAlnum */
+#define ANYOF_NALPHANUMERIC ((ANYOF_ALPHANUMERIC) + 1)
+#define ANYOF_ASCII ((_CC_ASCII) * 2)
+#define ANYOF_NASCII ((ANYOF_ASCII) + 1)
+#define ANYOF_BLANK ((_CC_BLANK) * 2) /* GNU extension: space and tab: non-vertical space */
+#define ANYOF_NBLANK ((ANYOF_BLANK) + 1)
+#define ANYOF_CASED ((_CC_CASED) * 2) /* Pseudo class for [:lower:] or
+ [:upper:] under /i */
+#define ANYOF_NCASED ((ANYOF_CASED) + 1)
+#define ANYOF_CNTRL ((_CC_CNTRL) * 2)
+#define ANYOF_NCNTRL ((ANYOF_CNTRL) + 1)
+#define ANYOF_DIGIT ((_CC_DIGIT) * 2) /* \d */
+#define ANYOF_NDIGIT ((ANYOF_DIGIT) + 1)
+#define ANYOF_GRAPH ((_CC_GRAPH) * 2)
+#define ANYOF_NGRAPH ((ANYOF_GRAPH) + 1)
+#define ANYOF_LOWER ((_CC_LOWER) * 2)
+#define ANYOF_NLOWER ((ANYOF_LOWER) + 1)
+#define ANYOF_PRINT ((_CC_PRINT) * 2)
+#define ANYOF_NPRINT ((ANYOF_PRINT) + 1)
+#define ANYOF_PSXSPC ((_CC_PSXSPC) * 2) /* POSIX space: \s plus the vertical tab */
+#define ANYOF_NPSXSPC ((ANYOF_PSXSPC) + 1)
+#define ANYOF_PUNCT ((_CC_PUNCT) * 2)
+#define ANYOF_NPUNCT ((ANYOF_PUNCT) + 1)
+#define ANYOF_SPACE ((_CC_SPACE) * 2) /* \s */
+#define ANYOF_NSPACE ((ANYOF_SPACE) + 1)
+#define ANYOF_UPPER ((_CC_UPPER) * 2)
+#define ANYOF_NUPPER ((ANYOF_UPPER) + 1)
+#define ANYOF_WORDCHAR ((_CC_WORDCHAR) * 2) /* \w, PL_utf8_alnum, utf8::IsWord, ALNUM */
+#define ANYOF_NWORDCHAR ((ANYOF_WORDCHAR) + 1)
+#define ANYOF_XDIGIT ((_CC_XDIGIT) * 2)
+#define ANYOF_NXDIGIT ((ANYOF_XDIGIT) + 1)
+
+/* pseudo classes below this, not stored in the class bitmap, but used as flags
during compilation of char classes */
-#define ANYOF_VERTWS (ANYOF_MAX+1)
-#define ANYOF_NVERTWS (ANYOF_MAX+2)
-#define ANYOF_HORIZWS (ANYOF_MAX+3)
-#define ANYOF_NHORIZWS (ANYOF_MAX+4)
+#define ANYOF_VERTWS ((_CC_VERTSPACE) * 2)
+#define ANYOF_NVERTWS ((ANYOF_VERTWS)+1)
+
+/* It is best if this is the last one, as all above it are stored as bits in a
+ * bitmap, and it isn't part of that bitmap */
+#if _CC_VERTSPACE != _HIGHEST_REGCOMP_DOT_H_SYNC
+# error Problem with handy.h _HIGHEST_REGCOMP_DOT_H_SYNC #define
+#endif
+
+#define ANYOF_POSIXL_MAX (ANYOF_VERTWS) /* So upper loop limit is written:
+ * '< ANYOF_MAX'
+ * Hence doesn't include VERTWS, as that
+ * is a pseudo class */
+#define ANYOF_MAX ANYOF_POSIXL_MAX
+
+#if (ANYOF_POSIXL_MAX > 32) /* Must fit in 32-bit word */
+# error Problem with handy.h _CC_foo #defines
+#endif
+
+#define ANYOF_HORIZWS ((ANYOF_POSIXL_MAX)+2) /* = (ANYOF_NVERTWS + 1) */
+#define ANYOF_NHORIZWS ((ANYOF_POSIXL_MAX)+3)
+
+#define ANYOF_UNIPROP ((ANYOF_POSIXL_MAX)+4) /* Used to indicate a Unicode
+ property: \p{} or \P{} */
/* Backward source code compatibility. */
#define ANYOF_NALNUML ANYOF_NALNUM
#define ANYOF_SPACEL ANYOF_SPACE
#define ANYOF_NSPACEL ANYOF_NSPACE
+#define ANYOF_ALNUM ANYOF_WORDCHAR
+#define ANYOF_NALNUM ANYOF_NWORDCHAR
/* 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_BIT(c) (1U << ((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_POSIXL_SET(p, c) (((regnode_charclass_posixl*) (p))->classflags |= (1U << (c)))
+#define ANYOF_CLASS_SET(p, c) ANYOF_POSIXL_SET((p), (c))
-#define ANYOF_CLASS_ZERO(ret) Zero(((struct regnode_charclass_class*)(ret))->classflags, ANYOF_CLASSBITMAP_SIZE, char)
-#define ANYOF_CLASS_SETALL(ret) \
- memset (((struct regnode_charclass_class*)(ret))->classflags, 255, ANYOF_CLASSBITMAP_SIZE)
-#define ANYOF_BITMAP_ZERO(ret) Zero(((struct regnode_charclass*)(ret))->bitmap, ANYOF_BITMAP_SIZE, char)
+#define ANYOF_POSIXL_CLEAR(p, c) (((regnode_charclass_posixl*) (p))->classflags &= ~ (1U <<(c)))
+#define ANYOF_CLASS_CLEAR(p, c) ANYOF_POSIXL_CLEAR((p), (c))
+
+#define ANYOF_POSIXL_TEST(p, c) (((regnode_charclass_posixl*) (p))->classflags & (1U << (c)))
+#define ANYOF_CLASS_TEST(p, c) ANYOF_POSIXL_TEST((p), (c))
+
+#define ANYOF_POSIXL_ZERO(ret) STMT_START { ((regnode_charclass_posixl*) (ret))->classflags = 0; } STMT_END
+#define ANYOF_CLASS_ZERO(ret) ANYOF_POSIXL_ZERO(ret)
+
+/* Shifts a bit to get, eg. 0x4000_0000, then subtracts 1 to get 0x3FFF_FFFF */
+#define ANYOF_POSIXL_SETALL(ret) STMT_START { ((regnode_charclass_posixl*) (ret))->classflags = ((1U << ((ANYOF_POSIXL_MAX) - 1))) - 1; } STMT_END
+#define ANYOF_CLASS_SETALL(ret) ANYOF_POSIXL_SETALL(ret)
+#define ANYOF_POSIXL_TEST_ANY_SET(p) \
+ ((ANYOF_FLAGS(p) & ANYOF_MATCHES_POSIXL) \
+ && (((regnode_charclass_posixl*)(p))->classflags))
+#define ANYOF_CLASS_TEST_ANY_SET(p) ANYOF_POSIXL_TEST_ANY_SET(p)
+
+/* Since an SSC always has this field, we don't have to test for that; nor do
+ * we want to because the bit isn't set for SSC during its construction */
+#define ANYOF_POSIXL_SSC_TEST_ANY_SET(p) \
+ cBOOL(((regnode_ssc*)(p))->classflags)
+#define ANYOF_POSIXL_SSC_TEST_ALL_SET(p) /* Are all bits set? */ \
+ (((regnode_ssc*) (p))->classflags \
+ == ((1U << ((ANYOF_POSIXL_MAX) - 1))) - 1)
+
+#define ANYOF_POSIXL_TEST_ALL_SET(p) \
+ ((ANYOF_FLAGS(p) & ANYOF_MATCHES_POSIXL) \
+ && ((regnode_charclass_posixl*) (p))->classflags \
+ == ((1U << ((ANYOF_POSIXL_MAX) - 1))) - 1)
+
+#define ANYOF_POSIXL_OR(source, dest) STMT_START { (dest)->classflags |= (source)->classflags ; } STMT_END
+#define ANYOF_CLASS_OR(source, dest) ANYOF_POSIXL_OR((source), (dest))
+
+#define ANYOF_POSIXL_AND(source, dest) STMT_START { (dest)->classflags &= (source)->classflags ; } STMT_END
+
+#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_TEST(p, c) cBOOL(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) /* Assumes sizeof(p) == 32 */ \
- 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))
-
-#if ANYOF_CLASSBITMAP_SIZE != 4
-# error ANYOF_CLASSBITMAP_SIZE is expected to be 4
-#endif
-#define ANYOF_CLASS_TEST_ANY_SET(p) ((ANYOF_FLAGS(p) & ANYOF_CLASS) \
- && memNE (((struct regnode_charclass_class*)(p))->classflags, \
- "\0\0\0\0", ANYOF_CLASSBITMAP_SIZE))
-/*#define ANYOF_CLASS_ADD_SKIP (ANYOF_CLASS_SKIP - ANYOF_SKIP)
- * */
+#define ANYOF_SKIP EXTRA_SIZE(struct regnode_charclass)
+#define ANYOF_POSIXL_SKIP EXTRA_SIZE(regnode_charclass_posixl)
/*
* Utility definitions.
#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_ZERO_LEN_SEEN 0x00000001
+#define REG_LOOKBEHIND_SEEN 0x00000002
+#define REG_GPOS_SEEN 0x00000004
/* spare */
-#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
-#define REG_SEEN_VERBARG 0x00000080
-#define REG_SEEN_CUTGROUP 0x00000100
-#define REG_SEEN_RUN_ON_COMMENT 0x00000200
-#define REG_SEEN_EXACTF_SHARP_S 0x00000400
+#define REG_CANY_SEEN 0x00000010
+#define REG_RECURSE_SEEN 0x00000020
+#define REG_TOP_LEVEL_BRANCHES_SEEN 0x00000040
+#define REG_VERBARG_SEEN 0x00000080
+#define REG_CUTGROUP_SEEN 0x00000100
+#define REG_RUN_ON_COMMENT_SEEN 0x00000200
+#define REG_UNFOLDED_MULTI_SEEN 0x00000400
+#define REG_GOSTART_SEEN 0x00000800
+#define REG_UNBOUNDED_QUANTIFIER_SEEN 0x00001000
+
START_EXTERN_C
#define check_offset_max substrs->data[2].max_offset
#define check_end_shift substrs->data[2].end_shift
-#define RX_ANCHORED_SUBSTR(rx) (((struct regexp *)SvANY(rx))->anchored_substr)
-#define RX_ANCHORED_UTF8(rx) (((struct regexp *)SvANY(rx))->anchored_utf8)
-#define RX_FLOAT_SUBSTR(rx) (((struct regexp *)SvANY(rx))->float_substr)
-#define RX_FLOAT_UTF8(rx) (((struct regexp *)SvANY(rx))->float_utf8)
+#define RX_ANCHORED_SUBSTR(rx) (ReANY(rx)->anchored_substr)
+#define RX_ANCHORED_UTF8(rx) (ReANY(rx)->anchored_utf8)
+#define RX_FLOAT_SUBSTR(rx) (ReANY(rx)->float_substr)
+#define RX_FLOAT_UTF8(rx) (ReANY(rx)->float_utf8)
/* trie related stuff */
#define RE_DEBUG_COMPILE_TRIE 0x000004
#define RE_DEBUG_COMPILE_DUMP 0x000008
#define RE_DEBUG_COMPILE_FLAGS 0x000010
+#define RE_DEBUG_COMPILE_TEST 0x000020
/* Execute */
#define RE_DEBUG_EXECUTE_MASK 0x00FF00
if (re_debug_flags & RE_DEBUG_COMPILE_TRIE) x )
#define DEBUG_FLAGS_r(x) DEBUG_r( \
if (re_debug_flags & RE_DEBUG_COMPILE_FLAGS) x )
+#define DEBUG_TEST_r(x) DEBUG_r( \
+ if (re_debug_flags & RE_DEBUG_COMPILE_TEST) x )
/* Execute */
#define DEBUG_EXECUTE_r(x) DEBUG_r( \
if (re_debug_flags & RE_DEBUG_EXECUTE_MASK) x )
if (re_debug_flags & RE_DEBUG_EXTRA_GPOS) x )
/* initialization */
-/* get_sv() can return NULL during global destruction. re_debug_flags can get
- * clobbered by a longjmp, so must be initialized */
+/* get_sv() can return NULL during global destruction. */
#define GET_RE_DEBUG_FLAGS DEBUG_r({ \
SV * re_debug_flags_sv = NULL; \
- re_debug_flags = 0; \
re_debug_flags_sv = get_sv(RE_DEBUG_FLAGS, 1); \
if (re_debug_flags_sv) { \
if (!SvIOK(re_debug_flags_sv)) \
#ifdef DEBUGGING
-#define GET_RE_DEBUG_FLAGS_DECL VOL IV re_debug_flags \
- PERL_UNUSED_DECL = 0; GET_RE_DEBUG_FLAGS;
+#define GET_RE_DEBUG_FLAGS_DECL VOL IV re_debug_flags = 0; \
+ PERL_UNUSED_VAR(re_debug_flags); GET_RE_DEBUG_FLAGS;
#define RE_PV_COLOR_DECL(rpv,rlen,isuni,dsv,pv,l,m,c1,c2) \
const char * const rpv = \
https://perl5.git.perl.org / perl5.git / blobdiff