/* Convert branch sequences to more efficient trie ops? */
#define PERL_ENABLE_TRIE_OPTIMISATION 1
-/* Be really agressive about optimising patterns with trie sequences? */
+/* Be really aggressive about optimising patterns with trie sequences? */
#define PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION 1
-/* Use old style unicode mappings for perl and posix character classes
- *
- * NOTE: Enabling this essentially breaks character class matching against unicode
- * strings, so that POSIX char classes match when they shouldn't, and \d matches
- * way more than 10 characters, and sometimes a charclass and its complement either
- * both match or neither match.
- * NOTE: Disabling this will cause various backwards compatibility issues to rear
- * their head, and tests to fail. However it will make the charclass behaviour
- * consistant regardless of internal string type, and make character class inversions
- * consistant. The tests that fail in the regex engine are basically broken tests.
- *
- * Personally I think 5.12 should disable this for sure. Its a bit more debatable for
- * 5.10, so for now im leaving it enabled.
- * XXX: It is now enabled for 5.11/5.12
- *
- * -demerphq
- */
-#define PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS 1
-
/* Should the optimiser take positive assertions into account? */
#define PERL_ENABLE_POSITIVE_ASSERTION_STUDY 0
#undef STRING
#define OP(p) ((p)->type)
-#define FLAGS(p) ((p)->flags) /* Caution: Doesn't apply to all \
- regnode types */
+#define FLAGS(p) ((p)->flags) /* Caution: Doesn't apply to all \
+ regnode types. For some, it's the \
+ character set of the regnode */
#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 SIZE_ONLY (RExC_emit == &PL_regdummy)
-/* Flags for node->flags of several of the node types */
-#define USE_UNI 0x01
-
-/* Flags for node->flags of ANYOF */
-
-#define ANYOF_LOCALE 0x01
-#define ANYOF_FOLD 0x02
-#define ANYOF_INVERT 0x04
-
-/* CLASS is never set unless LOCALE is too: has runtime \d, \w, [:posix:], ... */
-#define ANYOF_CLASS 0x08
+/* If the bitmap doesn't fully represent 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
+ * (and the code at the time this comment was written, is written so that all
+ * that is necessary to make the change would be to redefine the ANYOF_CLASS
+ * define). Once the planned change to compile all the above-latin1 code points
+ * is done, then the UNICODE_ALL bit can be freed up. 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
+
+/* 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
+
+#define ANYOF_INVERT 0x04
+
+/* EOS, meaning that it can match an empty string too, is used for the
+ * synthetic start class (ssc) only. It looks like it could share the INVERT
+ * bit, as the ssc is never inverted. But doing that caused this reges to
+ * not match:
+ * 'foo/file.fob' =~ m,^(?=[^\.])[^/]* /(?=[^\.])[^/]*\.fo[^/]$,;
+ * (except the space between the * and the / above shouldn't be there; it was
+ * inserted to make this comment continue on.)
+ * Rather than try to figure out what was going on in the optimizer, I (khw)
+ * found a way to save a different bit. But my original line of reasoning was
+ * "The bit just needs to be turned off before regexec.c gets a hold of it so
+ * that regexec.c doesn't think it's inverted, but this happens automatically,
+ * as if the ssc can match an EOS, the ssc is discarded, and never passed to
+ * regexec.c" */
+#define ANYOF_EOS 0x10
+
+/* CLASS is never set unless LOCALE is too: has runtime \d, \w, [:posix:], ...
+ * The non-locale ones are resolved at compile-time */
+#define ANYOF_CLASS 0x08
#define ANYOF_LARGE ANYOF_CLASS /* Same; name retained for back compat */
-/* EOS used for regstclass only */
-#define ANYOF_EOS 0x10 /* Can match an empty string too */
+/* Can match something outside the bitmap that isn't in utf8 */
+#define ANYOF_NONBITMAP_NON_UTF8 0x20
+
+/* Matches every code point 0x100 and above*/
+#define ANYOF_UNICODE_ALL 0x40
-#define ANYOF_UNICODE 0x20 /* Matches >= one thing past 0xff */
-#define ANYOF_UNICODE_ALL 0x40 /* Matches 0x100 - infinity */
+/* 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_FLAGS_ALL 0xff
#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)
+
+/* The class bit can be set to the locale one if necessary to save bits at the
+ * expense of having locale ANYOF nodes always have a class bit map, and hence
+ * take up extra space. This allows convenient changing it as development
+ * proceeds on this */
+#if ANYOF_CLASS == ANYOF_LOCALE
+# undef ANYOF_CLASS_ADD_SKIP
+# define ANYOF_ADD_LOC_SKIP (ANYOF_CLASS_SKIP - ANYOF_SKIP)
+
+ /* Quicker way to see if there are actually any tests. This is because
+ * currently the set of tests can be empty even when the class bitmap is
+ * allocated */
+# if ANYOF_CLASSBITMAP_SIZE != 4
+# error ANYOF_CLASSBITMAP_SIZE is expected to be 4
+# endif
+# define ANYOF_CLASS_TEST_ANY_SET(p) /* assumes sizeof(p) = 4 */ \
+ memNE (((struct regnode_charclass_class*)(p))->classflags, \
+ "\0\0\0\0", ANYOF_CLASSBITMAP_SIZE)
+#else
+# define ANYOF_CLASS_ADD_SKIP (ANYOF_CLASS_SKIP - ANYOF_SKIP)
+# undef ANYOF_ADD_LOC_SKIP
+# define ANYOF_CLASS_TEST_ANY_SET(p) (ANYOF_FLAGS(p) & ANYOF_CLASS)
+#endif
/*
};
typedef struct _reg_ac_data reg_ac_data;
-/* ANY_BIT doesnt use the structure, so we can borrow it here.
+/* ANY_BIT doesn't 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... */
const char * const rpv = \
pv_pretty((dsv), (pv), (l), (m), \
PL_colors[(c1)],PL_colors[(c2)], \
- PERL_PV_ESCAPE_RE |((isuni) ? PERL_PV_ESCAPE_UNI : 0) ); \
+ PERL_PV_ESCAPE_RE|PERL_PV_ESCAPE_NONASCII |((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)], \
- PERL_PV_ESCAPE_RE |((isuni) ? PERL_PV_ESCAPE_UNI : 0) )
+ PERL_PV_ESCAPE_RE|PERL_PV_ESCAPE_NONASCII |((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_ESCAPE_RE | PERL_PV_PRETTY_ELLIPSES | \
+ ( PERL_PV_PRETTY_QUOTE | PERL_PV_ESCAPE_RE | PERL_PV_ESCAPE_NONASCII | PERL_PV_PRETTY_ELLIPSES | \
((isuni) ? PERL_PV_ESCAPE_UNI : 0)) \
)
https://perl5.git.perl.org / perl5.git / blobdiff