X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/96ac09754b53d56f88a3effe7c9c82d363d89ee5..50646f6f0beccf4783bfef8cb75375f49c727140:/handy.h

diff --git a/handy.h b/handy.h
index ee4d668..b75eafb 100644
--- a/handy.h
+++ b/handy.h
@@ -470,7 +470,7 @@ There are three variants for all the functions in this section.  The base ones
 operate using the character set of the platform Perl is running on.  The ones
 with an C<_A> suffix operate on the ASCII character set, and the ones with an
 C<_L1> suffix operate on the full Latin1 character set.  All are unaffected by
-locale
+locale.
 
 For ASCII platforms, the base function with no suffix and the one with the
 C<_A> suffix are identical.  The function with the C<_L1> suffix imposes the
@@ -640,6 +640,8 @@ EXTCONST U32 PL_charclass[];
 #   define isUPPER_A(c)  cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_UPPER_A))
 #   define isWORDCHAR_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_WORDCHAR_A))
 #   define isXDIGIT_A(c)  cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_XDIGIT_A))
+    /* Either participates in a fold with a character above 255, or is a
+     * multi-char fold */
 #   define _HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(c) ((! cBOOL(FITS_IN_8_BITS(c))) || (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_NONLATIN1_FOLD))
 #else   /* No perl.h. */
 #   define isOCTAL_A(c)  ((c) >= '0' && (c) <= '9')
@@ -881,16 +883,13 @@ EXTCONST U32 PL_charclass[];
 #define isBLANK_LC_uni(c)	isBLANK(c) /* could be wrong */
 
 #define isALNUM_utf8(p)		is_utf8_alnum(p)
-/* The ID_Start of Unicode was originally quite limiting: it assumed an
- * L-class character (meaning that you could not have, say, a CJK charac-
- * ter). So, instead, perl has for a long time allowed ID_Continue but
- * not digits.
- * We still preserve that for backward compatibility. But we also make sure
- * that it is alphanumeric, so S_scan_word in toke.c will not hang. See
- *    http://rt.perl.org/rt3/Ticket/Display.html?id=74022
- * for more detail than you ever wanted to know about. */
-#define isIDFIRST_utf8(p) \
-    (is_utf8_idcont(p) && !is_utf8_digit(p) && is_utf8_alnum(p))
+/* To prevent S_scan_word in toke.c from hanging, we have to make sure that
+ * IDFIRST is an alnum.  See
+ * http://rt.perl.org/rt3/Ticket/Display.html?id=74022
+ * for more detail than you ever wanted to know about.  This used to be not the
+ * XID version, but we decided to go with the more modern Unicode definition */
+#define isIDFIRST_utf8(p)	(is_utf8_xidfirst(p) && is_utf8_alnum(p))
+#define isIDCONT_utf8(p)	is_utf8_xidcont(p)
 #define isALPHA_utf8(p)		is_utf8_alpha(p)
 #define isSPACE_utf8(p)		is_utf8_space(p)
 #define isDIGIT_utf8(p)		is_utf8_digit(p)
@@ -955,7 +954,7 @@ The XSUB-writer's interface to the C C<malloc> function.
 In 5.9.3, Newx() and friends replace the older New() API, and drops
 the first parameter, I<x>, a debug aid which allowed callers to identify
 themselves.  This aid has been superseded by a new build option,
-PERL_MEM_LOG (see L<perlhack/PERL_MEM_LOG>).  The older API is still
+PERL_MEM_LOG (see L<perlhacktips/PERL_MEM_LOG>).  The older API is still
 there for use in XS modules supporting older perls.
 
 =for apidoc Am|void|Newxc|void* ptr|int nitems|type|cast