U16 arg2;
};
-
-#define ANYOF_BITMAP_SIZE 32 /* 256 b/(8 b/B) */
+/* 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 {
/* has runtime (locale) \d, \w, ..., [:posix:] classes */
struct regnode_charclass_class {
- U8 flags; /* ANYOF_POSIXL 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 */
+ char bitmap[ANYOF_BITMAP_SIZE]; /* both compile-time ... */
U32 classflags; /* and run-time */
};
-/* like above, but also has folds that are used only if the runtime locale is
- * UTF-8. */
-struct regnode_charclass_posixl_fold {
- U8 flags; /* ANYOF_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 */
- SV* utf8_locale_list; /* list of code points matched by folds
- in a UTF-8 locale */
-};
-
-/* A synthetic start class; 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. */
+/* 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_POSIXL 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 */
+ char bitmap[ANYOF_BITMAP_SIZE]; /* both compile-time ... */
U32 classflags; /* and run-time */
- SV* utf8_locale_list; /* list of code points matched by folds
- in a UTF-8 locale */
- SV* invlist; /* list of code points matched */
+
+ /* 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
#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, with one
- * currently available. The ABOVE_LATIN1_ALL bit could be freed up
+#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
* only for /d, so there are no combinatorial issues. The LOC flag to use is
* probably the POSIXL one.
* Several flags are not used in synthetic start class (SSC) nodes, so could be
- * shared should new flags be needed for SSCs, like ANYOF_EMPTY_STRING now. */
+ * shared should new flags be needed for SSCs, like SSC_MATCHES_EMPTY_STRING
+ * now. */
/* regexec.c is expecting this to be in the low bit */
-#define ANYOF_INVERT 0x01
+#define ANYOF_INVERT 0x01
/* For the SSC node only, which cannot be inverted, so is shared with that bit.
- * This means "Does this SSC match an empty string?" This is used only during
- * regex compilation. */
-#define ANYOF_EMPTY_STRING ANYOF_INVERT
+ * This is used only during regex compilation. */
+#define SSC_MATCHES_EMPTY_STRING ANYOF_INVERT
-/* spare 0x02 */
+/* 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
+#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_POSIXL 0x08
-#define ANYOF_CLASS ANYOF_POSIXL
-#define ANYOF_LARGE ANYOF_POSIXL
+#define ANYOF_MATCHES_POSIXL 0x08
/* Should we raise a warning if matching against an above-Unicode code point?
* */
-#define ANYOF_WARN_SUPER 0x10
+#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_ABOVE_LATIN1_ALL 0x40
-#define ANYOF_UNICODE_ALL ANYOF_ABOVE_LATIN1_ALL
+/* 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_NON_ASCII_ALL 0x80
+#define ANYOF_MATCHES_ALL_NON_UTF8_NON_ASCII 0x80
-#define ANYOF_FLAGS_ALL (0xf5)
+#define ANYOF_FLAGS_ALL (0xff)
-#define ANYOF_LOCALE_FLAGS (ANYOF_LOC_FOLD | ANYOF_POSIXL)
+#define ANYOF_LOCALE_FLAGS (ANYOF_LOC_FOLD | ANYOF_MATCHES_POSIXL)
/* 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)
+#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() */
#define ANYOF_SIZE (sizeof(struct regnode_charclass))
#define ANYOF_POSIXL_SIZE (sizeof(regnode_charclass_posixl))
#define ANYOF_CLASS_SIZE ANYOF_POSIXL_SIZE
-#define ANYOF_POSIXL_FOLD_SIZE (sizeof(regnode_charclass_posixl_fold))
#define ANYOF_FLAGS(p) ((p)->flags)
-#define ANYOF_BIT(c) (1 << ((c) & 7))
+#define ANYOF_BIT(c) (1U << ((c) & 7))
#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_SETALL(ret) ANYOF_POSIXL_SETALL(ret)
#define ANYOF_POSIXL_TEST_ANY_SET(p) \
- ((ANYOF_FLAGS(p) & ANYOF_POSIXL) \
+ ((ANYOF_FLAGS(p) & ANYOF_MATCHES_POSIXL) \
&& (((regnode_charclass_posixl*)(p))->classflags))
#define ANYOF_CLASS_TEST_ANY_SET(p) ANYOF_POSIXL_TEST_ANY_SET(p)
== ((1U << ((ANYOF_POSIXL_MAX) - 1))) - 1)
#define ANYOF_POSIXL_TEST_ALL_SET(p) \
- ((ANYOF_FLAGS(p) & ANYOF_POSIXL) \
+ ((ANYOF_FLAGS(p) & ANYOF_MATCHES_POSIXL) \
&& ((regnode_charclass_posixl*) (p))->classflags \
== ((1U << ((ANYOF_POSIXL_MAX) - 1))) - 1)
#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_POSIXL_SKIP ((ANYOF_POSIXL_SIZE - 1)/sizeof(regnode))
-#define ANYOF_POSIXL_FOLD_SKIP ((ANYOF_POSIXL_FOLD_SIZE - 1)/sizeof(regnode))
#define ANYOF_CLASS_SKIP ANYOF_POSIXL_SKIP
-#define ANYOF_UTF8_LOCALE_INVLIST(node) (((regnode_charclass_posixl_fold*) (node))->utf8_locale_list)
-
/*
* Utility definitions.
*/