the published Unicode rules; otherwise, it uses the C library function that
gives the named classification. For example, C<isDIGIT_LC()> when not in a
UTF-8 locale returns the result of calling C<isdigit()>. FALSE is always
-returned if the input won't fit into an octet.
+returned if the input won't fit into an octet. On some platforms where the C
+library function is known to be defective, Perl changes its result to follow
+the POSIX standard's rules.
Variant C<isFOO_LC_uvchr> is like C<isFOO_LC>, but is defined on any UV. It
returns the same as C<isFOO_LC> for input code points less than 256, and
alphanumeric.
See the L<top of this section|/Character classification> for an explanation of
variants
-C<isWORDCHAR_A>, C<isWORDCHAR_L1>, C<isWORDCHAR_uni>, C<isWORDCHAR_utf8>,
-C<isWORDCHAR_LC>, C<isWORDCHAR_LC_uvchr>, and C<isWORDCHAR_LC_utf8>.
+C<isWORDCHAR_A>, C<isWORDCHAR_L1>, C<isWORDCHAR_uni>, and C<isWORDCHAR_utf8>.
+C<isWORDCHAR_LC>, C<isWORDCHAR_LC_uvchr>, and C<isWORDCHAR_LC_utf8> are also as
+described there, but additionally include the platform's native underscore.
=for apidoc Am|bool|isXDIGIT|char ch
Returns a boolean indicating whether the specified character is a hexadecimal
* digits */
#define isOCTAL_A(c) cBOOL(FITS_IN_8_BITS(c) && (0xF8 & (c)) == '0')
-/* ASCII range only */
#ifdef H_PERL /* If have access to perl.h, lookup in its table */
/* Character class numbers. For internal core Perl use only. The ones less
# define _CC_QUOTEMETA 21
# define _CC_NON_FINAL_FOLD 22
# define _CC_IS_IN_SOME_FOLD 23
-# define _CC_BACKSLASH_FOO_LBRACE_IS_META 31 /* temp, see mk_PL_charclass.pl */
-/* Unused: 24-30
+# define _CC_MNEMONIC_CNTRL 24
+/* Unused: 25-31
* If more bits are needed, one could add a second word for non-64bit
* QUAD_IS_INT systems, using some #ifdefs to distinguish between having a 2nd
* word or not. The IS_IN_SOME_FOLD bit is the most easily expendable, as it
# define isALPHANUMERIC_A(c) _generic_isCC_A(c, _CC_ALPHANUMERIC)
# define isBLANK_A(c) _generic_isCC_A(c, _CC_BLANK)
# define isCNTRL_A(c) _generic_isCC_A(c, _CC_CNTRL)
-# define isDIGIT_A(c) _generic_isCC(c, _CC_DIGIT)
+# define isDIGIT_A(c) _generic_isCC(c, _CC_DIGIT) /* No non-ASCII digits */
# define isGRAPH_A(c) _generic_isCC_A(c, _CC_GRAPH)
# define isLOWER_A(c) _generic_isCC_A(c, _CC_LOWER)
# define isPRINT_A(c) _generic_isCC_A(c, _CC_PRINT)
# define isSPACE_A(c) _generic_isCC_A(c, _CC_SPACE)
# define isUPPER_A(c) _generic_isCC_A(c, _CC_UPPER)
# define isWORDCHAR_A(c) _generic_isCC_A(c, _CC_WORDCHAR)
-# define isXDIGIT_A(c) _generic_isCC(c, _CC_XDIGIT)
+# define isXDIGIT_A(c) _generic_isCC(c, _CC_XDIGIT) /* No non-ASCII xdigits */
# define isIDFIRST_A(c) _generic_isCC_A(c, _CC_IDFIRST)
# define isALPHA_L1(c) _generic_isCC(c, _CC_ALPHA)
# define isALPHANUMERIC_L1(c) _generic_isCC(c, _CC_ALPHANUMERIC)
_generic_isCC(c, _CC_NON_FINAL_FOLD)
# define _IS_IN_SOME_FOLD_ONLY_FOR_USE_BY_REGCOMP_DOT_C(c) \
_generic_isCC(c, _CC_IS_IN_SOME_FOLD)
-#else /* else we don't have perl.h */
+# define _IS_MNEMONIC_CNTRL_ONLY_FOR_USE_BY_REGCOMP_DOT_C(c) \
+ _generic_isCC(c, _CC_MNEMONIC_CNTRL)
+#else /* else we don't have perl.h H_PERL */
/* If we don't have perl.h, we are compiling a utility program. Below we
* hard-code various macro definitions that wouldn't otherwise be available
|| ((c) <= 'F' && (c) >= 'A'))
/* The _L1 macros may be unnecessary for the utilities; I (khw) added them
- * during debugging, and it seems best to keep them. */
+ * during debugging, and it seems best to keep them. We may be called
+ * without NATIVE_TO_LATIN1 being defined. On ASCII platforms, it doesn't
+ * do anything anyway, so make it not a problem */
+# if ! defined(EBCDIC) && ! defined(NATIVE_TO_LATIN1)
+# define NATIVE_TO_LATIN1(ch) (ch)
+# endif
# define isPSXSPC_A(c) isSPACE_A(c) /* XXX Assumes SPACE matches '\v' */
# define isALPHA_L1(c) (isUPPER_L1(c) || isLOWER_L1(c))
# define isALPHANUMERIC_L1(c) (isALPHA_L1(c) || isDIGIT_A(c))
/* And these aren't accurate at all. They are useful only for above
* Latin1, which utilities and bootstrapping don't deal with */
# define _IS_NON_FINAL_FOLD_ONLY_FOR_USE_BY_REGCOMP_DOT_C(c) 0
+# define _HAS_NONLATIN1_SIMPLE_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(c) 0
# define _HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(c) 0
/* Many of the macros later in this file are defined in terms of these. By
(FITS_IN_8_BITS(c) && S_bootstrap_ctype((U8) (c), (classnum), TRUE))
# define _generic_isCC_A(c, classnum) \
(FITS_IN_8_BITS(c) && S_bootstrap_ctype((U8) (c), (classnum), FALSE))
-#endif /* End of no perl.h */
+#endif /* End of no perl.h H_PERL */
#define isALPHANUMERIC(c) isALPHANUMERIC_A(c)
#define isALPHA(c) isALPHA_A(c)
* both ASCII and EBCDIC the last 3 bits of the octal digits range from 0-7. */
#define OCTAL_VALUE(c) (__ASSERT_(isOCTAL(c)) (7 & (c)))
+/* Efficiently returns a boolean as to if two native characters are equivalent
+ * case-insenstively. At least one of the characters must be one of [A-Za-z];
+ * the ALPHA in the name is to remind you of that. This is asserted() in
+ * DEBUGGING builds. Because [A-Za-z] are invariant under UTF-8, this macro
+ * works (on valid input) for both non- and UTF-8-encoded bytes.
+ *
+ * When one of the inputs is a compile-time constant and gets folded by the
+ * compiler, this reduces to an AND and a TEST. On both EBCDIC and ASCII
+ * machines, 'A' and 'a' differ by a single bit; the same with the upper and
+ * lower case of all other ASCII-range alphabetics. On ASCII platforms, they
+ * are 32 apart; on EBCDIC, they are 64. At compile time, this uses an
+ * exclusive 'or' to find that bit and then inverts it to form a mask, with
+ * just a single 0, in the bit position where the upper- and lowercase differ.
+ * */
+#define isALPHA_FOLD_EQ(c1, c2) \
+ (__ASSERT_(isALPHA_A(c1) || isALPHA_A(c2)) \
+ ((c1) & ~('A' ^ 'a')) == ((c2) & ~('A' ^ 'a')))
+#define isALPHA_FOLD_NE(c1, c2) (! isALPHA_FOLD_EQ((c1), (c2)))
+
/*
=head1 Memory Management
https://perl5.git.perl.org / perl5.git / blobdiff