regcomp.c: Add some static functions

This commit adds some functions that are currently unused, but will be
used in a future commit.  This commit is essentially to make the
differences smaller in that commit, as 'diff' is getting confused and
not outputting the logical differences.  The functions are added in a
block at the beginning of the file to avoid the 'diff' issues.  A later
white-space only commit will move them to more appropriate positions.

5 files changed, 305 insertions, 0 deletions

diff --git a/embed.fnc b/embed.fnc
index bb0805f6c4..ec203f9c1f 100644
--- a/embed.fnc
+++ b/embed.fnc
@@ -2059,12 +2059,29 @@ Esn	|void	|ssc_anything	|NN const RExC_state_t *pRExC_state \
 EsRn	|int	|ssc_is_anything|NN const regnode_ssc *ssc
 Esn	|void	|ssc_init	|NN const RExC_state_t *pRExC_state \
 				|NN regnode_ssc *ssc
+EsR	|int	|ssc_is_cp_posixl_init|NN const RExC_state_t *pRExC_state \
+				|NN const regnode_ssc *ssc
 Es	|void	|ssc_and	|NN const RExC_state_t *pRExC_state \
 				|NN regnode_ssc *ssc                \
 				|NN const regnode_ssc *and_with
+Esn	|void	|ssc_flags_and	|NN regnode_ssc *ssc|const U8 and_with
 Esn	|void	|ssc_or		|NN const RExC_state_t *pRExC_state \
 				|NN regnode_ssc *ssc \
 				|NN const regnode_ssc *or_with
+Es	|SV*	|get_ANYOF_cp_list_for_ssc                                 \
+				|NN const RExC_state_t *pRExC_state \
+				|NN const regnode_charclass_posixl* const node
+Ei	|void	|ssc_intersection|NN regnode_ssc *ssc \
+				|NN SV* const invlist|const bool invert_2nd
+Ei	|void	|ssc_union	|NN regnode_ssc *ssc \
+				|NN SV* const invlist|const bool invert_2nd
+Ei	|void	|ssc_add_range	|NN regnode_ssc *ssc \
+				|UV const start|UV const end
+Ei	|void	|ssc_cp_and	|NN regnode_ssc *ssc \
+				|UV const cp
+Ei	|void	|ssc_clear_locale|NN regnode_ssc *ssc
+Es	|void	|ssc_finalize	|NN RExC_state_t *pRExC_state \
+				|NN regnode_ssc *ssc
 Es	|SSize_t|study_chunk	|NN RExC_state_t *pRExC_state \
 				|NN regnode **scanp|NN SSize_t *minlenp \
 				|NN SSize_t *deltap|NN regnode *last \
diff --git a/embed.h b/embed.h
index 45d9f89767..fca8736feb 100644
--- a/embed.h
+++ b/embed.h
@@ -903,6 +903,7 @@
 #define alloc_maybe_populate_EXACT(a,b,c,d,e)	S_alloc_maybe_populate_EXACT(aTHX_ a,b,c,d,e)
 #define compute_EXACTish(a)	S_compute_EXACTish(aTHX_ a)
 #define could_it_be_a_POSIX_class(a)	S_could_it_be_a_POSIX_class(aTHX_ a)
+#define get_ANYOF_cp_list_for_ssc(a,b)	S_get_ANYOF_cp_list_for_ssc(aTHX_ a,b)
 #define get_invlist_iter_addr(a)	S_get_invlist_iter_addr(aTHX_ a)
 #define get_invlist_previous_index_addr(a)	S_get_invlist_previous_index_addr(aTHX_ a)
 #define grok_bslash_N(a,b,c,d,e,f,g)	S_grok_bslash_N(aTHX_ a,b,c,d,e,f,g)
@@ -944,11 +945,19 @@
 #define regwhite		S_regwhite
 #define scan_commit(a,b,c,d)	S_scan_commit(aTHX_ a,b,c,d)
 #define set_ANYOF_arg(a,b,c,d,e,f)	S_set_ANYOF_arg(aTHX_ a,b,c,d,e,f)
+#define ssc_add_range(a,b,c)	S_ssc_add_range(aTHX_ a,b,c)
 #define ssc_and(a,b,c)		S_ssc_and(aTHX_ a,b,c)
 #define ssc_anything		S_ssc_anything
+#define ssc_clear_locale(a)	S_ssc_clear_locale(aTHX_ a)
+#define ssc_cp_and(a,b)		S_ssc_cp_and(aTHX_ a,b)
+#define ssc_finalize(a,b)	S_ssc_finalize(aTHX_ a,b)
+#define ssc_flags_and		S_ssc_flags_and
 #define ssc_init		S_ssc_init
+#define ssc_intersection(a,b,c)	S_ssc_intersection(aTHX_ a,b,c)
 #define ssc_is_anything		S_ssc_is_anything
+#define ssc_is_cp_posixl_init(a,b)	S_ssc_is_cp_posixl_init(aTHX_ a,b)
 #define ssc_or			S_ssc_or
+#define ssc_union(a,b,c)	S_ssc_union(aTHX_ a,b,c)
 #define study_chunk(a,b,c,d,e,f,g,h,i,j,k)	S_study_chunk(aTHX_ a,b,c,d,e,f,g,h,i,j,k)
 #  endif
 #  if defined(PERL_IN_REGCOMP_C) || defined (PERL_IN_DUMP_C)
diff --git a/proto.h b/proto.h
index 91949484fb..568cdf733c 100644
--- a/proto.h
+++ b/proto.h
@@ -6557,6 +6557,12 @@ STATIC bool	S_could_it_be_a_POSIX_class(pTHX_ RExC_state_t *pRExC_state)
 #define PERL_ARGS_ASSERT_COULD_IT_BE_A_POSIX_CLASS	\
 	assert(pRExC_state)
 
+STATIC SV*	S_get_ANYOF_cp_list_for_ssc(pTHX_ const RExC_state_t *pRExC_state, const regnode_charclass_posixl* const node)
+			__attribute__nonnull__(pTHX_1)
+			__attribute__nonnull__(pTHX_2);
+#define PERL_ARGS_ASSERT_GET_ANYOF_CP_LIST_FOR_SSC	\
+	assert(pRExC_state); assert(node)
+
 PERL_STATIC_INLINE STRLEN*	S_get_invlist_iter_addr(pTHX_ SV* invlist)
 			__attribute__warn_unused_result__
 			__attribute__nonnull__(pTHX_1);
@@ -6809,6 +6815,11 @@ STATIC void	S_set_ANYOF_arg(pTHX_ RExC_state_t* const pRExC_state, regnode* cons
 #define PERL_ARGS_ASSERT_SET_ANYOF_ARG	\
 	assert(pRExC_state); assert(node)
 
+PERL_STATIC_INLINE void	S_ssc_add_range(pTHX_ regnode_ssc *ssc, UV const start, UV const end)
+			__attribute__nonnull__(pTHX_1);
+#define PERL_ARGS_ASSERT_SSC_ADD_RANGE	\
+	assert(ssc)
+
 STATIC void	S_ssc_and(pTHX_ const RExC_state_t *pRExC_state, regnode_ssc *ssc, const regnode_ssc *and_with)
 			__attribute__nonnull__(pTHX_1)
 			__attribute__nonnull__(pTHX_2)
@@ -6822,18 +6833,52 @@ STATIC void	S_ssc_anything(const RExC_state_t *pRExC_state, regnode_ssc *ssc)
 #define PERL_ARGS_ASSERT_SSC_ANYTHING	\
 	assert(pRExC_state); assert(ssc)
 
+PERL_STATIC_INLINE void	S_ssc_clear_locale(pTHX_ regnode_ssc *ssc)
+			__attribute__nonnull__(pTHX_1);
+#define PERL_ARGS_ASSERT_SSC_CLEAR_LOCALE	\
+	assert(ssc)
+
+PERL_STATIC_INLINE void	S_ssc_cp_and(pTHX_ regnode_ssc *ssc, UV const cp)
+			__attribute__nonnull__(pTHX_1);
+#define PERL_ARGS_ASSERT_SSC_CP_AND	\
+	assert(ssc)
+
+STATIC void	S_ssc_finalize(pTHX_ RExC_state_t *pRExC_state, regnode_ssc *ssc)
+			__attribute__nonnull__(pTHX_1)
+			__attribute__nonnull__(pTHX_2);
+#define PERL_ARGS_ASSERT_SSC_FINALIZE	\
+	assert(pRExC_state); assert(ssc)
+
+STATIC void	S_ssc_flags_and(regnode_ssc *ssc, const U8 and_with)
+			__attribute__nonnull__(1);
+#define PERL_ARGS_ASSERT_SSC_FLAGS_AND	\
+	assert(ssc)
+
 STATIC void	S_ssc_init(const RExC_state_t *pRExC_state, regnode_ssc *ssc)
 			__attribute__nonnull__(1)
 			__attribute__nonnull__(2);
 #define PERL_ARGS_ASSERT_SSC_INIT	\
 	assert(pRExC_state); assert(ssc)
 
+PERL_STATIC_INLINE void	S_ssc_intersection(pTHX_ regnode_ssc *ssc, SV* const invlist, const bool invert_2nd)
+			__attribute__nonnull__(pTHX_1)
+			__attribute__nonnull__(pTHX_2);
+#define PERL_ARGS_ASSERT_SSC_INTERSECTION	\
+	assert(ssc); assert(invlist)
+
 STATIC int	S_ssc_is_anything(const regnode_ssc *ssc)
 			__attribute__warn_unused_result__
 			__attribute__nonnull__(1);
 #define PERL_ARGS_ASSERT_SSC_IS_ANYTHING	\
 	assert(ssc)
 
+STATIC int	S_ssc_is_cp_posixl_init(pTHX_ const RExC_state_t *pRExC_state, const regnode_ssc *ssc)
+			__attribute__warn_unused_result__
+			__attribute__nonnull__(pTHX_1)
+			__attribute__nonnull__(pTHX_2);
+#define PERL_ARGS_ASSERT_SSC_IS_CP_POSIXL_INIT	\
+	assert(pRExC_state); assert(ssc)
+
 STATIC void	S_ssc_or(const RExC_state_t *pRExC_state, regnode_ssc *ssc, const regnode_ssc *or_with)
 			__attribute__nonnull__(1)
 			__attribute__nonnull__(2)
@@ -6841,6 +6886,12 @@ STATIC void	S_ssc_or(const RExC_state_t *pRExC_state, regnode_ssc *ssc, const re
 #define PERL_ARGS_ASSERT_SSC_OR	\
 	assert(pRExC_state); assert(ssc); assert(or_with)
 
+PERL_STATIC_INLINE void	S_ssc_union(pTHX_ regnode_ssc *ssc, SV* const invlist, const bool invert_2nd)
+			__attribute__nonnull__(pTHX_1)
+			__attribute__nonnull__(pTHX_2);
+#define PERL_ARGS_ASSERT_SSC_UNION	\
+	assert(ssc); assert(invlist)
+
 STATIC SSize_t	S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp, SSize_t *minlenp, SSize_t *deltap, regnode *last, struct scan_data_t *data, I32 stopparen, U8* recursed, regnode_ssc *and_withp, U32 flags, U32 depth)
 			__attribute__nonnull__(pTHX_1)
 			__attribute__nonnull__(pTHX_2)
diff --git a/regcomp.c b/regcomp.c
index cade9624ea..ec24583f1f 100644
--- a/regcomp.c
+++ b/regcomp.c
@@ -259,6 +259,227 @@ struct RExC_state_t {
 #define _invlist_intersection_complement_2nd(a, b, output) \
                  _invlist_intersection_maybe_complement_2nd(a, b, TRUE, output)
 
+STATIC void
+S_ssc_flags_and(regnode_ssc *ssc, const U8 and_with)
+{
+    /* Take the flags 'and_with' and accumulate them anded into the flags for
+     * the SSC 'ssc'.  The non-SSC related flags in 'and_with' are ignored. */
+
+    const U8 ssc_only_flags = ANYOF_FLAGS(ssc) & ~ANYOF_LOCALE_FLAGS;
+
+    PERL_ARGS_ASSERT_SSC_FLAGS_AND;
+
+    /* Use just the SSC-related flags from 'and_with' */
+    ANYOF_FLAGS(ssc) &= (and_with & ANYOF_LOCALE_FLAGS);
+    ANYOF_FLAGS(ssc) |= ssc_only_flags;
+}
+
+STATIC int
+S_ssc_is_cp_posixl_init(pTHX_ const RExC_state_t *pRExC_state,
+                              const regnode_ssc *ssc)
+{
+    /* Returns TRUE if the SSC 'ssc' is in its initial state with regard only
+     * to the list of code points matched, and locale posix classes; hence does
+     * not check its flags) */
+
+    UV start, end;
+    bool ret;
+
+    PERL_ARGS_ASSERT_SSC_IS_CP_POSIXL_INIT;
+
+    assert(OP(ssc) == ANYOF_SYNTHETIC);
+
+    invlist_iterinit(ssc->invlist);
+    ret = invlist_iternext(ssc->invlist, &start, &end)
+          && start == 0
+          && end == UV_MAX;
+
+    invlist_iterfinish(ssc->invlist);
+
+    if (! ret) {
+        return FALSE;
+    }
+
+    if (RExC_contains_locale) {
+        if (! (ANYOF_FLAGS(ssc) & ANYOF_LOCALE)
+            || ! (ANYOF_FLAGS(ssc) & ANYOF_POSIXL)
+            || ! ANYOF_POSIXL_TEST_ALL_SET(ssc))
+        {
+            return FALSE;
+        }
+    }
+
+    return TRUE;
+}
+
+STATIC SV*
+S_get_ANYOF_cp_list_for_ssc(pTHX_ const RExC_state_t *pRExC_state,
+                                  const regnode_charclass_posixl* const node)
+{
+    /* Returns a mortal inversion list defining which code points are matched
+     * by 'node', which is of type ANYOF.  Handles complementing the result if
+     * appropriate.  If some code points aren't knowable at this time, the
+     * returned list must, and will, contain every possible code point. */
+
+    SV* invlist = sv_2mortal(_new_invlist(0));
+    unsigned int i;
+    const U32 n = ARG(node);
+
+    PERL_ARGS_ASSERT_GET_ANYOF_CP_LIST_FOR_SSC;
+
+    /* Look at the data structure created by S_set_ANYOF_arg() */
+    if (n != ANYOF_NONBITMAP_EMPTY) {
+        SV * const rv = MUTABLE_SV(RExC_rxi->data->data[n]);
+        AV * const av = MUTABLE_AV(SvRV(rv));
+        SV **const ary = AvARRAY(av);
+        assert(RExC_rxi->data->what[n] == 's');
+
+        if (ary[1] && ary[1] != &PL_sv_undef) { /* Has compile-time swash */
+            invlist = sv_2mortal(invlist_clone(_get_swash_invlist(ary[1])));
+        }
+        else if (ary[0] && ary[0] != &PL_sv_undef) {
+
+            /* Here, no compile-time swash, and there are things that won't be
+             * known until runtime -- we have to assume it could be anything */
+            return _add_range_to_invlist(invlist, 0, UV_MAX);
+        }
+        else {
+
+            /* Here no compile-time swash, and no run-time only data.  Use the
+             * node's inversion list */
+            invlist = sv_2mortal(invlist_clone(ary[2]));
+        }
+    }
+
+    /* An ANYOF node contains a bitmap for the first 256 code points, and an
+     * inversion list for the others, but if there are code points that should
+     * match only conditionally on the target string being UTF-8, those are
+     * placed in the inversion list, and not the bitmap.  Since there are
+     * circumstances under which they could match, they are included in the
+     * SSC.  But if the ANYOF node is to be inverted, we have to exclude them
+     * here, so that when we invert below, the end result actually does include
+     * them.  (Think about "\xe0" =~ /[^\xc0]/di;).  We have to do this here
+     * before we add the unconditionally matched code points */
+    if (ANYOF_FLAGS(node) & ANYOF_INVERT) {
+        _invlist_intersection_complement_2nd(invlist,
+                                             PL_UpperLatin1,
+                                             &invlist);
+    }
+
+    /* Add in the points from the bit map */
+    for (i = 0; i < 256; i++) {
+        if (ANYOF_BITMAP_TEST(node, i)) {
+            invlist = add_cp_to_invlist(invlist, i);
+        }
+    }
+
+    /* If this can match all upper Latin1 code points, have to add them
+     * as well */
+    if (ANYOF_FLAGS(node) & ANYOF_NON_UTF8_LATIN1_ALL) {
+        _invlist_union(invlist, PL_UpperLatin1, &invlist);
+    }
+
+    /* Similarly for these */
+    if (ANYOF_FLAGS(node) & ANYOF_ABOVE_LATIN1_ALL) {
+        invlist = _add_range_to_invlist(invlist, 256, UV_MAX);
+    }
+
+    if (ANYOF_FLAGS(node) & ANYOF_INVERT) {
+        _invlist_invert(invlist);
+    }
+
+    return invlist;
+}
+
+PERL_STATIC_INLINE void
+S_ssc_union(pTHX_ regnode_ssc *ssc, SV* const invlist, const bool invert2nd)
+{
+    PERL_ARGS_ASSERT_SSC_UNION;
+
+    assert(OP(ssc) == ANYOF_SYNTHETIC);
+
+    _invlist_union_maybe_complement_2nd(ssc->invlist,
+                                        invlist,
+                                        invert2nd,
+                                        &ssc->invlist);
+}
+
+PERL_STATIC_INLINE void
+S_ssc_intersection(pTHX_ regnode_ssc *ssc,
+                         SV* const invlist,
+                         const bool invert2nd)
+{
+    PERL_ARGS_ASSERT_SSC_INTERSECTION;
+
+    assert(OP(ssc) == ANYOF_SYNTHETIC);
+
+    _invlist_intersection_maybe_complement_2nd(ssc->invlist,
+                                               invlist,
+                                               invert2nd,
+                                               &ssc->invlist);
+}
+
+PERL_STATIC_INLINE void
+S_ssc_add_range(pTHX_ regnode_ssc *ssc, const UV start, const UV end)
+{
+    PERL_ARGS_ASSERT_SSC_ADD_RANGE;
+
+    assert(OP(ssc) == ANYOF_SYNTHETIC);
+
+    ssc->invlist = _add_range_to_invlist(ssc->invlist, start, end);
+}
+
+PERL_STATIC_INLINE void
+S_ssc_cp_and(pTHX_ regnode_ssc *ssc, const UV cp)
+{
+    /* AND just the single code point 'cp' into the SSC 'ssc' */
+
+    SV* cp_list = _new_invlist(2);
+
+    PERL_ARGS_ASSERT_SSC_CP_AND;
+
+    assert(OP(ssc) == ANYOF_SYNTHETIC);
+
+    cp_list = add_cp_to_invlist(cp_list, cp);
+    ssc_intersection(ssc, cp_list,
+                     FALSE /* Not inverted */
+                     );
+    SvREFCNT_dec_NN(cp_list);
+}
+
+PERL_STATIC_INLINE void
+S_ssc_clear_locale(pTHX_ regnode_ssc *ssc)
+{
+    /* Set the SSC 'ssc' to not match any locale things */
+
+    PERL_ARGS_ASSERT_SSC_CLEAR_LOCALE;
+
+    assert(OP(ssc) == ANYOF_SYNTHETIC);
+
+    ANYOF_POSIXL_ZERO(ssc);
+    ANYOF_FLAGS(ssc) &= ~ANYOF_LOCALE_FLAGS;
+}
+
+STATIC void
+S_ssc_finalize(pTHX_ RExC_state_t *pRExC_state, regnode_ssc *ssc)
+{
+    /* The inversion list in the SSC is marked mortal; now we need a more
+     * permanent copy, which is stored the same way that is done in a regular
+     * ANYOF node, with the first 256 code points in a bit map */
+
+    SV* invlist = invlist_clone(ssc->invlist);
+
+    PERL_ARGS_ASSERT_SSC_FINALIZE;
+
+    assert(OP(ssc) == ANYOF_SYNTHETIC);
+
+    populate_ANYOF_from_invlist( (regnode *) ssc, &invlist);
+
+    set_ANYOF_arg(pRExC_state, (regnode *) ssc, invlist, NULL, NULL, FALSE);
+
+    assert(! (ANYOF_FLAGS(ssc) & ANYOF_LOCALE) || RExC_contains_locale);
+}
+
 /* About scan_data_t.
 
   During optimisation we recurse through the regexp program performing
diff --git a/regcomp.h b/regcomp.h
index 0eb989fa7a..f0153fc12c 100644
--- a/regcomp.h
+++ b/regcomp.h
@@ -365,6 +365,9 @@ struct regnode_ssc {
 	                           |ANYOF_LOC_FOLD                      \
 	                           |ANYOF_POSIXL                         \
 	                           |ANYOF_NONBITMAP_NON_UTF8)
+#define ANYOF_LOCALE_FLAGS (ANYOF_LOCALE                        \
+                           |ANYOF_LOC_FOLD                      \
+                           |ANYOF_POSIXL)
 
 /* Character classes for node->classflags of ANYOF */
 /* Should be synchronized with a table in regprop() */
@@ -472,6 +475,10 @@ struct regnode_ssc {
 	 && (((regnode_charclass_posixl*)(p))->classflags))
 #define ANYOF_CLASS_TEST_ANY_SET(p) ANYOF_POSIXL_TEST_ANY_SET(p)
 
+#define ANYOF_POSIXL_TEST_ALL_SET(p)                               \
+        ((ANYOF_FLAGS(p) & ANYOF_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))