Re: \N{...} in regular expression [PATCH]

Message-ID: <9b18b3110609181637m796d6c16o1b2741edc5f09eb2@mail.gmail.com>

p4raw-id: //depot/perl@28868

MANIFEST

@@ -3177,6 +3177,7 @@ t/io/through.t			See if pipe passes data intact
 t/io/utf8.t			See if file seeking works
 t/japh/abigail.t		Obscure tests
 t/lib/1_compile.t		See if the various libraries and extensions compile
+t/lib/Cname.pm			Test charnames in regexes (op/pat.t)
 t/lib/common.pl			Helper for lib/{warnings,feature}.t
 t/lib/commonsense.t		See if configuration meets basic needs
 t/lib/compmod.pl		Helper for 1_compile.t

embed.fnc

@@ -1308,6 +1308,7 @@ Es	|regnode*|regclass	|NN struct RExC_state_t *state|U32 depth
 ERsn	|I32	|regcurly	|NN const char *
 Es	|regnode*|reg_node	|NN struct RExC_state_t *state|U8 op
 Es	|regnode*|regpiece	|NN struct RExC_state_t *state|NN I32 *flagp|U32 depth
+Es	|regnode*|reg_namedseq	|NN struct RExC_state_t *state|NULLOK UV *valuep
 Es	|void	|reginsert	|NN struct RExC_state_t *state|U8 op|NN regnode *opnd
 Es	|void	|regtail	|NN struct RExC_state_t *state|NN regnode *p|NN const regnode *val|U32 depth
 Es	|U32	|join_exact	|NN struct RExC_state_t *state|NN regnode *scan|NN I32 *min|U32 flags|NULLOK regnode *val|U32 depth

embed.h

@@ -1312,6 +1312,7 @@
 #define regcurly		S_regcurly
 #define reg_node		S_reg_node
 #define regpiece		S_regpiece
+#define reg_namedseq		S_reg_namedseq
 #define reginsert		S_reginsert
 #define regtail			S_regtail
 #define join_exact		S_join_exact
@@ -3500,6 +3501,7 @@
 #define regcurly		S_regcurly
 #define reg_node(a,b)		S_reg_node(aTHX_ a,b)
 #define regpiece(a,b,c)		S_regpiece(aTHX_ a,b,c)
+#define reg_namedseq(a,b)	S_reg_namedseq(aTHX_ a,b)
 #define reginsert(a,b,c)	S_reginsert(aTHX_ a,b,c)
 #define regtail(a,b,c,d)	S_regtail(aTHX_ a,b,c,d)
 #define join_exact(a,b,c,d,e,f)	S_join_exact(aTHX_ a,b,c,d,e,f)

pod/perldiag.pod

@@ -1334,6 +1334,14 @@ specified in the C<\N{...}> escape.  Perhaps you forgot to load the
 corresponding C<overload> or C<charnames> pragma?  See L<charnames> and
 L<overload>.
 
+=item Constant(%s)%s: %s in regex; marked by <-- HERE in m/%s/
+
+(F) The parser found inconsistencies while attempting to find 
+the character name specified in the C<\N{...}> escape.  Perhaps you 
+forgot to load the corresponding C<charnames> pragma?  
+See L<charnames>.
+
+
 =item Constant is not %s reference
 
 (F) A constant value (perhaps declared using the C<use constant> pragma)
@@ -1841,6 +1849,13 @@ about 250 characters for simple names, and somewhat more for compound
 names (like C<$A::B>).  You've exceeded Perl's limits.  Future versions
 of Perl are likely to eliminate these arbitrary limitations.
 
+=item Ignoring %s in character class in regex; marked by <-- HERE in m/%s/
+
+(W) Named unicode character escapes (\N{...}) may return multi-char
+or zero length sequences. When such an escape is used in a character class
+its behaviour is not well defined. Check that the correct escape has 
+been used, and the correct charname handler is in scope.
+
 =item Illegal binary digit %s
 
 (F) You used a digit other than 0 or 1 in a binary number.

proto.h

@@ -3565,6 +3565,9 @@ STATIC regnode*	S_regpiece(pTHX_ struct RExC_state_t *state, I32 *flagp, U32 dep
 			__attribute__nonnull__(pTHX_1)
 			__attribute__nonnull__(pTHX_2);
 
+STATIC regnode*	S_reg_namedseq(pTHX_ struct RExC_state_t *state, UV *valuep)
+			__attribute__nonnull__(pTHX_1);
+
 STATIC void	S_reginsert(pTHX_ struct RExC_state_t *state, U8 op, regnode *opnd)
 			__attribute__nonnull__(pTHX_1)
 			__attribute__nonnull__(pTHX_3);