X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/e2e755386e52b4bdb22a5c9618390859ed5f7323..40e43b7872ae49f9f7db93781a2041b2d19b256c:/handy.h diff --git a/handy.h b/handy.h index 216d0ea..abfc2c2 100644 --- a/handy.h +++ b/handy.h @@ -1,7 +1,7 @@ /* handy.h * * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1999, 2000, - * 2001, 2002, 2004, 2005, 2006, 2007, 2008 by Larry Wall and others + * 2001, 2002, 2004, 2005, 2006, 2007, 2008, 2012 by Larry Wall and others * * You may distribute under the terms of either the GNU General Public * License or the Artistic License, as specified in the README file. @@ -70,10 +70,13 @@ Null SV pointer. (No longer available when C is defined.) #define MUTABLE_IO(p) ((IO *)MUTABLE_PTR(p)) #define MUTABLE_SV(p) ((SV *)MUTABLE_PTR(p)) -/* XXX Configure ought to have a test for a boolean type, if I can - just figure out all the headers such a test needs. - Andy Dougherty August 1996 -*/ +#ifdef I_STDBOOL +# include +# ifndef HAS_BOOL +# define HAS_BOOL 1 +# endif +#endif + /* bool is built-in for g++-2.6.3 and later, which might be used for extensions. <_G_config.h> defines _G_HAVE_BOOL, but we can't be sure _G_config.h will be included before this file. _G_config.h @@ -112,9 +115,10 @@ Null SV pointer. (No longer available when C is defined.) /* a simple (bool) cast may not do the right thing: if bool is defined * as char for example, then the cast from int is implementation-defined + * (bool)!!(cbool) in a ternary triggers a bug in xlc on AIX */ -#define cBOOL(cbool) ((bool)!!(cbool)) +#define cBOOL(cbool) ((cbool) ? (bool)1 : (bool)0) /* Try to figure out __func__ or __FUNCTION__ equivalent, if any. * XXX Should really be a Configure probe, with HAS__FUNCTION__ @@ -208,19 +212,6 @@ typedef U64TYPE U64; # endif #endif -/* HMB H.Merijn Brand - a placeholder for preparing Configure patches: - * - * USE_DTRACE HAS_PSEUDOFORK HAS_TIMEGM LOCALTIME_R_NEEDS_TZSET - * GMTIME_MAX GMTIME_MIN LOCALTIME_MAX LOCALTIME_MIN - * HAS_CTIME64 HAS_LOCALTIME64 HAS_GMTIME64 HAS_DIFFTIME64 - * HAS_MKTIME64 HAS_ASCTIME64 HAS_GETADDRINFO HAS_GETNAMEINFO - * HAS_INETNTOP HAS_INETPTON CHARBITS HAS_PRCTL - * Not (yet) used at top level, but mention them for metaconfig - */ - -/* Mention I8SIZE, U8SIZE, I16SIZE, U16SIZE, I32SIZE, U32SIZE, - I64SIZE, and U64SIZE here so that metaconfig pulls them in. */ - #if defined(UINT8_MAX) && defined(INT16_MAX) && defined(INT32_MAX) /* I8_MAX and I8_MIN constants are not defined, as I8 is an ambiguous type. @@ -483,7 +474,7 @@ 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 and by C. 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 @@ -572,27 +563,101 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc */ -/* FITS_IN_8_BITS(c) returns true if c occupies no more than 8 bits. It is - * designed to be hopefully bomb-proof, making sure that no bits of - * information are lost even on a 64-bit machine, but to get the compiler to - * optimize it out if possible. This is because Configure makes sure that the - * machine has an 8-bit byte, so if c is stored in a byte, the sizeof() - * guarantees that this evaluates to a constant true at compile time. The use - * of the mask instead of '< 256' keeps gcc from complaining that it is alway - * true, when c's storage class is a byte. Use U64TYPE because U64 is known - * only in the perl core, and this macro can be called from outside that */ +/* Specify the widest unsigned type on the platform. Use U64TYPE because U64 + * is known only in the perl core, and this macro can be called from outside + * that */ #ifdef HAS_QUAD -# define FITS_IN_8_BITS(c) ((sizeof(c) == 1) || (((U64TYPE)(c) & 0xFF) == (U64TYPE)(c))) +# define WIDEST_UTYPE U64TYPE #else -# define FITS_IN_8_BITS(c) ((sizeof(c) == 1) || (((U32)(c) & 0xFF) == (U32)(c))) +# define WIDEST_UTYPE U32 +#endif + +/* FITS_IN_8_BITS(c) returns true if c doesn't have a bit set other than in + * the lower 8. It is designed to be hopefully bomb-proof, making sure that no + * bits of information are lost even on a 64-bit machine, but to get the + * compiler to optimize it out if possible. This is because Configure makes + * sure that the machine has an 8-bit byte, so if c is stored in a byte, the + * sizeof() guarantees that this evaluates to a constant true at compile time. + */ +#define FITS_IN_8_BITS(c) ((sizeof(c) == 1) || !(((WIDEST_UTYPE)(c)) & ~0xFF)) + +#ifdef EBCDIC +# define isASCII(c) (FITS_IN_8_BITS(c) && (NATIVE_TO_UNI((U8) (c)) < 128)) +#else +# define isASCII(c) ((WIDEST_UTYPE)(c) < 128) #endif -#define isASCII(c) (FITS_IN_8_BITS(c) ? NATIVE_TO_UNI((U8) c) <= 127 : 0) #define isASCII_A(c) isASCII(c) +#define isASCII_L1(c) isASCII(c) + +/* The lower 3 bits in both the ASCII and EBCDIC representations of '0' are 0, + * and the 8 possible permutations of those bits exactly comprise the 8 octal + * 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 */ -# include "l1_char_class_tab.h" +/* Bits for PL_charclass[]. These use names used in l1_char_class_tab.h but + * their actual definitions are here. If that has a name not used here, it + * won't compile. */ +# define _CC_ALNUMC_A (1<<0) +# define _CC_ALNUMC_L1 (1<<1) +# define _CC_ALPHA_A (1<<2) +# define _CC_ALPHA_L1 (1<<3) +# define _CC_BLANK_A (1<<4) +# define _CC_BLANK_L1 (1<<5) +# define _CC_CHARNAME_CONT (1<<6) +# define _CC_CNTRL_A (1<<7) +# define _CC_CNTRL_L1 (1<<8) +# define _CC_DIGIT_A (1<<9) +# define _CC_GRAPH_A (1<<10) +# define _CC_GRAPH_L1 (1<<11) +# define _CC_IDFIRST_A (1<<12) +# define _CC_IDFIRST_L1 (1<<13) +# define _CC_LOWER_A (1<<14) +# define _CC_LOWER_L1 (1<<15) +# define _CC_PRINT_A (1<<17) +# define _CC_PRINT_L1 (1<<18) +# define _CC_PSXSPC_A (1<<19) +# define _CC_PSXSPC_L1 (1<<20) +# define _CC_PUNCT_A (1<<21) +# define _CC_PUNCT_L1 (1<<22) +# define _CC_SPACE_A (1<<23) +# define _CC_SPACE_L1 (1<<24) +# define _CC_UPPER_A (1<<25) +# define _CC_UPPER_L1 (1<<26) +# define _CC_WORDCHAR_A (1<<27) +# define _CC_WORDCHAR_L1 (1<<28) +# define _CC_XDIGIT_A (1<<29) +# define _CC_NONLATIN1_FOLD (1<<30) +# define _CC_QUOTEMETA (1U<<31) /* 1U keeps Solaris from griping */ +/* Unused: (1<<16) + * If more are needed, can give up some of the above. The first ones to go + * would be those that require just two tests to verify; either there are two + * code points, like BLANK_A, or it occupies a single range like DIGIT_A, + * UPPER_A, and LOWER_A. Also consider the ones that can be replaced with two + * tests and an additional mask, so + * + * #define isCNTRL_A cBOOL(FITS_IN_8_BITS(c) \ + * && (( ! (~0x1F & NATIVE_TO_UNI(c)])) \ + * || UNLIKELY(NATIVE_TO_UNI(c) == 0x7f))) + * + * This takes advantage of the contiguous block of these with the first one's + * representation having the lower order bits all zero;, except the DELETE must + * be tested specially. A similar pattern can be used for for isCNTRL_L1, + * isPRINT_A, and isPRINT_L1 + */ + +# ifdef DOINIT +EXTCONST U32 PL_charclass[] = { +# include "l1_char_class_tab.h" +}; + +# else /* ! DOINIT */ +EXTCONST U32 PL_charclass[]; +# endif + # define isALNUMC_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_ALNUMC_A)) # define isALPHA_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_ALPHA_A)) # define isBLANK_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_BLANK_A)) @@ -601,7 +666,6 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define isGRAPH_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_GRAPH_A)) # define isIDFIRST_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_IDFIRST_A)) # define isLOWER_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_LOWER_A)) -# define isOCTAL_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_OCTAL_A)) # define isPRINT_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_PRINT_A)) # define isPSXSPC_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_PSXSPC_A)) # define isPUNCT_A(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_PUNCT_A)) @@ -609,9 +673,11 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # 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)) +# define _isQUOTEMETA(c) cBOOL(FITS_IN_8_BITS(c) && (PL_charclass[(U8) NATIVE_TO_UNI(c)] & _CC_QUOTEMETA)) #else /* No perl.h. */ -# define isOCTAL_A(c) ((c) >= '0' && (c) <= '9') # ifdef EBCDIC # define isALNUMC_A(c) (isASCII(c) && isALNUMC(c)) # define isALPHA_A(c) (isASCII(c) && isALPHA(c)) @@ -632,8 +698,8 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define isALNUMC_A(c) (isALPHA_A(c) || isDIGIT_A(c)) # define isALPHA_A(c) (isUPPER_A(c) || isLOWER_A(c)) # define isBLANK_A(c) ((c) == ' ' || (c) == '\t') -# define isCNTRL_A(c) (FITS_IN_8_BITS(c) ? ((U8) (c) < ' ' || (c) == 127) : 0) -# define isDIGIT_A(c) ((c) >= '0' && (c) <= '9') +# define isCNTRL_A(c) (FITS_IN_8_BITS(c) && ((U8) (c) < ' ' || (c) == 127)) +# define isDIGIT_A(c) ((c) <= '9' && (c) >= '0') # define isGRAPH_A(c) (isWORDCHAR_A(c) || isPUNCT_A(c)) # define isIDFIRST_A(c) (isALPHA_A(c) || (c) == '_') # define isLOWER_A(c) ((c) >= 'a' && (c) <= 'z') @@ -641,9 +707,9 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define isPSXSPC_A(c) (isSPACE_A(c) || (c) == '\v') # define isPUNCT_A(c) (((c) >= 33 && (c) <= 47) || ((c) >= 58 && (c) <= 64) || ((c) >= 91 && (c) <= 96) || ((c) >= 123 && (c) <= 126)) # define isSPACE_A(c) ((c) == ' ' || (c) == '\t' || (c) == '\n' || (c) =='\r' || (c) == '\f') -# define isUPPER_A(c) ((c) >= 'A' && (c) <= 'Z') +# define isUPPER_A(c) ((c) <= 'Z' && (c) >= 'A') # define isWORDCHAR_A(c) (isALPHA_A(c) || isDIGIT_A(c) || (c) == '_') -# define isXDIGIT_A(c) (isDIGIT_A(c) || ((c) >= 'a' && (c) <= 'f') || ((c) >= 'A' && (c) <= 'F')) +# define isXDIGIT_A(c) (isDIGIT_A(c) || ((c) >= 'a' && (c) <= 'f') || ((c) <= 'F' && (c) >= 'A')) # endif #endif /* ASCII range definitions */ @@ -712,7 +778,7 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define toLOWER(c) tolower(c) # define toUPPER(c) toupper(c) #else /* Not EBCDIC: ASCII-only matching */ -# define isALNUMC(c) isALNUMC_A(c) +# define isALNUMC(c) isALNUMC_A(c) /* Mnemonic: "C's alnum" = alpha + digit */ # define isALPHA(c) isALPHA_A(c) # define isBLANK(c) isBLANK_A(c) # define isCNTRL(c) isCNTRL_A(c) @@ -759,6 +825,8 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define isIDFIRST_LC(c) \ (NXIsAlpha((unsigned int)(c)) || (char)(c) == '_') # define isALPHA_LC(c) NXIsAlpha((unsigned int)(c)) +# define isASCII_LC(c) isASCII((unsigned int)(c)) +# define isBLANK_LC(c) isBLANK((unsigned int)(c)) # define isSPACE_LC(c) NXIsSpace((unsigned int)(c)) # define isDIGIT_LC(c) NXIsDigit((unsigned int)(c)) # define isUPPER_LC(c) NXIsUpper((unsigned int)(c)) @@ -775,9 +843,23 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # if defined(CTYPE256) || (!defined(isascii) && !defined(HAS_ISASCII)) +/* Note that the foo_LC() macros in this case generally are defined only on + * code points 0-256, and give undefined, unwarned results if called with + * values outside that range */ + # define isALNUM_LC(c) (isalnum((unsigned char)(c)) || (char)(c) == '_') # define isIDFIRST_LC(c) (isalpha((unsigned char)(c)) || (char)(c) == '_') # define isALPHA_LC(c) isalpha((unsigned char)(c)) +# ifdef HAS_ISASCII +# define isASCII_LC(c) isascii((unsigned char)(c)) +# else +# define isASCII_LC(c) isASCII((unsigned char)(c)) +# endif +# ifdef HAS_ISBLANK +# define isBLANK_LC(c) isblank((unsigned char)(c)) +# else +# define isBLANK_LC(c) isBLANK((unsigned char)(c)) +# endif # define isSPACE_LC(c) isspace((unsigned char)(c)) # define isDIGIT_LC(c) isdigit((unsigned char)(c)) # define isUPPER_LC(c) isupper((unsigned char)(c)) @@ -795,6 +877,12 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define isALNUM_LC(c) (isascii(c) && (isalnum(c) || (c) == '_')) # define isIDFIRST_LC(c) (isascii(c) && (isalpha(c) || (c) == '_')) # define isALPHA_LC(c) (isascii(c) && isalpha(c)) +# define isASCII_LC(c) isascii(c) +# ifdef HAS_ISBLANK +# define isBLANK_LC(c) (isascii(c) && isblank(c)) +# else +# define isBLANK_LC(c) isBLANK(c) +# endif # define isSPACE_LC(c) (isascii(c) && isspace(c)) # define isDIGIT_LC(c) (isascii(c) && isdigit(c)) # define isUPPER_LC(c) (isascii(c) && isupper(c)) @@ -811,27 +899,39 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc #endif /* USE_NEXT_CTYPE */ #define isPSXSPC_LC(c) (isSPACE_LC(c) || (c) == '\v') -#define isBLANK_LC(c) isBLANK(c) /* could be wrong */ - -#define isALNUM_uni(c) is_uni_alnum(c) -#define isIDFIRST_uni(c) is_uni_idfirst(c) -#define isALPHA_uni(c) is_uni_alpha(c) -#define isSPACE_uni(c) is_uni_space(c) -#define isDIGIT_uni(c) is_uni_digit(c) -#define isUPPER_uni(c) is_uni_upper(c) -#define isLOWER_uni(c) is_uni_lower(c) -#define isASCII_uni(c) is_uni_ascii(c) -#define isCNTRL_uni(c) is_uni_cntrl(c) -#define isGRAPH_uni(c) is_uni_graph(c) -#define isPRINT_uni(c) is_uni_print(c) -#define isPUNCT_uni(c) is_uni_punct(c) -#define isXDIGIT_uni(c) is_uni_xdigit(c) + +/* For use in the macros just below. If the input is Latin1, use the Latin1 + * (_L1) version of the macro; otherwise use the function. Won't compile if + * 'c' isn't unsigned, as won't match function prototype. The macros do bounds + * checking, so have duplicate checks here, so could create versions of the + * macros that don't, but experiments show that gcc optimizes them out anyway. + */ +#define generic_uni(macro, function, c) ((c) < 256 \ + ? CAT2(macro, _L1)(c) \ + : function(c)) +/* Note that all ignore 'use bytes' */ + +#define isALNUM_uni(c) generic_uni(isWORDCHAR, is_uni_alnum, c) +#define isIDFIRST_uni(c) generic_uni(isIDFIRST, is_uni_idfirst, c) +#define isALPHA_uni(c) generic_uni(isALPHA, is_uni_alpha, c) +#define isSPACE_uni(c) generic_uni(isSPACE, is_uni_space, c) +#define isDIGIT_uni(c) generic_uni(isDIGIT, is_uni_digit, c) +#define isUPPER_uni(c) generic_uni(isUPPER, is_uni_upper, c) +#define isLOWER_uni(c) generic_uni(isLOWER, is_uni_lower, c) +#define isASCII_uni(c) isASCII(c) +/* All controls are in Latin1 */ +#define isCNTRL_uni(c) ((c) < 256 && isCNTRL_L1(c)) +#define isGRAPH_uni(c) generic_uni(isGRAPH, is_uni_graph, c) +#define isPRINT_uni(c) generic_uni(isPRINT, is_uni_print, c) +#define isPUNCT_uni(c) generic_uni(isPUNCT, is_uni_punct, c) +#define isXDIGIT_uni(c) generic_uni(isXDIGIT, is_uni_xdigit, c) #define toUPPER_uni(c,s,l) to_uni_upper(c,s,l) #define toTITLE_uni(c,s,l) to_uni_title(c,s,l) #define toLOWER_uni(c,s,l) to_uni_lower(c,s,l) #define toFOLD_uni(c,s,l) to_uni_fold(c,s,l) -#define isPSXSPC_uni(c) (isSPACE_uni(c) ||(c) == '\f') +/* Posix and regular space differ only in U+000B, which is in Latin1 */ +#define isPSXSPC_uni(c) ((c) < 256 ? isPSXSPC_L1(c) : isSPACE_uni(c)) #define isBLANK_uni(c) isBLANK(c) /* could be wrong */ #define isALNUM_LC_uvchr(c) (c < 256 ? isALNUM_LC(c) : is_uni_alnum_lc(c)) @@ -849,47 +949,75 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc #define isPSXSPC_LC_uni(c) (isSPACE_LC_uni(c) ||(c) == '\f') #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)) -#define isALPHA_utf8(p) is_utf8_alpha(p) -#define isSPACE_utf8(p) is_utf8_space(p) -#define isDIGIT_utf8(p) is_utf8_digit(p) -#define isUPPER_utf8(p) is_utf8_upper(p) -#define isLOWER_utf8(p) is_utf8_lower(p) -#define isASCII_utf8(p) is_utf8_ascii(p) -#define isCNTRL_utf8(p) is_utf8_cntrl(p) -#define isGRAPH_utf8(p) is_utf8_graph(p) -#define isPRINT_utf8(p) is_utf8_print(p) -#define isPUNCT_utf8(p) is_utf8_punct(p) -#define isXDIGIT_utf8(p) is_utf8_xdigit(p) +/* For use in the macros just below. If the input is ASCII, use the ASCII (_A) + * version of the macro; if the input is in the upper Latin1 range, use the + * Latin1 (_L1) version of the macro, after converting from utf8; otherwise use + * the function. This relies on the fact that ASCII characters have the same + * representation whether utf8 or not */ +#define generic_utf8(macro, function, p) (isASCII(*(p)) \ + ? CAT2(CAT2(macro,_),A)(*(p)) \ + : (UTF8_IS_DOWNGRADEABLE_START(*(p))) \ + ? CAT2(macro, _L1) \ + (TWO_BYTE_UTF8_TO_UNI(*(p), \ + *((p)+1))) \ + : function(p)) + +/* Note that all assume that the utf8 has been validated, and ignore 'use + * bytes' */ + +#define isALNUM_utf8(p) generic_utf8(isWORDCHAR, 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. XXX It is unclear if this should extend to + * isIDFIRST_uni() which it hasn't so far. (In the ASCII range, there isn't a + * difference.) This used to be not the XID version, but we decided to go with + * the more modern Unicode definition */ +#define isIDFIRST_utf8(p) (isASCII(*(p)) \ + ? isIDFIRST_A(*(p)) \ + : (UTF8_IS_DOWNGRADEABLE_START(*(p))) \ + ? isIDFIRST_L1(TWO_BYTE_UTF8_TO_UNI(*(p), \ + *((p)+1)))\ + : Perl__is_utf8__perl_idstart(aTHX_ p)) +#define isIDCONT_utf8(p) generic_utf8(isWORDCHAR, is_utf8_xidcont, p) +#define isALPHA_utf8(p) generic_utf8(isALPHA, is_utf8_alpha, p) +#define isSPACE_utf8(p) generic_utf8(isSPACE, is_utf8_space, p) +#define isDIGIT_utf8(p) generic_utf8(isDIGIT, is_utf8_digit, p) +#define isUPPER_utf8(p) generic_utf8(isUPPER, is_utf8_upper, p) +#define isLOWER_utf8(p) generic_utf8(isLOWER, is_utf8_lower, p) +/* Because ASCII is invariant under utf8, the non-utf8 macro works */ +#define isASCII_utf8(p) isASCII(p) +#define isCNTRL_utf8(p) generic_utf8(isCNTRL, is_utf8_cntrl, p) +#define isGRAPH_utf8(p) generic_utf8(isGRAPH, is_utf8_graph, p) +#define isPRINT_utf8(p) generic_utf8(isPRINT, is_utf8_print, p) +#define isPUNCT_utf8(p) generic_utf8(isPUNCT, is_utf8_punct, p) +#define isXDIGIT_utf8(p) generic_utf8(isXDIGIT, is_utf8_xdigit, p) #define toUPPER_utf8(p,s,l) to_utf8_upper(p,s,l) #define toTITLE_utf8(p,s,l) to_utf8_title(p,s,l) #define toLOWER_utf8(p,s,l) to_utf8_lower(p,s,l) -#define isPSXSPC_utf8(c) (isSPACE_utf8(c) ||(c) == '\f') +/* Posix and regular space differ only in U+000B, which is in ASCII (and hence + * Latin1 */ +#define isPSXSPC_utf8(p) ((isASCII(*(p))) \ + ? isPSXSPC_A(*(p)) \ + : (UTF8_IS_DOWNGRADEABLE_START(*(p)) \ + ? isPSXSPC_L1(TWO_BYTE_UTF8_TO_UNI(*(p), \ + *((p)+1)))\ + : isSPACE_utf8(p))) #define isBLANK_utf8(c) isBLANK(c) /* could be wrong */ -#define isALNUM_LC_utf8(p) isALNUM_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isIDFIRST_LC_utf8(p) isIDFIRST_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isALPHA_LC_utf8(p) isALPHA_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isSPACE_LC_utf8(p) isSPACE_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isDIGIT_LC_utf8(p) isDIGIT_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isUPPER_LC_utf8(p) isUPPER_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isLOWER_LC_utf8(p) isLOWER_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isALNUMC_LC_utf8(p) isALNUMC_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isCNTRL_LC_utf8(p) isCNTRL_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isGRAPH_LC_utf8(p) isGRAPH_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isPRINT_LC_utf8(p) isPRINT_LC_uvchr(utf8_to_uvchr(p, 0)) -#define isPUNCT_LC_utf8(p) isPUNCT_LC_uvchr(utf8_to_uvchr(p, 0)) +#define isALNUM_LC_utf8(p) isALNUM_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isIDFIRST_LC_utf8(p) isIDFIRST_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isALPHA_LC_utf8(p) isALPHA_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isSPACE_LC_utf8(p) isSPACE_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isDIGIT_LC_utf8(p) isDIGIT_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isUPPER_LC_utf8(p) isUPPER_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isLOWER_LC_utf8(p) isLOWER_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isALNUMC_LC_utf8(p) isALNUMC_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isCNTRL_LC_utf8(p) isCNTRL_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isGRAPH_LC_utf8(p) isGRAPH_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isPRINT_LC_utf8(p) isPRINT_LC_uvchr(valid_utf8_to_uvchr(p, 0)) +#define isPUNCT_LC_utf8(p) isPUNCT_LC_uvchr(valid_utf8_to_uvchr(p, 0)) #define isPSXSPC_LC_utf8(c) (isSPACE_LC_utf8(c) ||(c) == '\f') #define isBLANK_LC_utf8(c) isBLANK(c) /* could be wrong */ @@ -924,7 +1052,7 @@ The XSUB-writer's interface to the C C function. In 5.9.3, Newx() and friends replace the older New() API, and drops the first parameter, I, a debug aid which allowed callers to identify themselves. This aid has been superseded by a new build option, -PERL_MEM_LOG (see L). The older API is still +PERL_MEM_LOG (see L). 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 @@ -974,7 +1102,7 @@ destination, C is the number of items, and C is the type. Like C but returns dest. Useful for encouraging compilers to tail-call optimise. -=for apidoc Am|void|StructCopy|type src|type dest|type +=for apidoc Am|void|StructCopy|type *src|type *dest|type This is an architecture-independent macro to copy one structure to another. =for apidoc Am|void|PoisonWith|void* dest|int nitems|type|U8 byte @@ -1188,8 +1316,8 @@ void Perl_mem_log_del_sv(const SV *sv, const char *filename, const int linenumbe * Local variables: * c-indentation-style: bsd * c-basic-offset: 4 - * indent-tabs-mode: t + * indent-tabs-mode: nil * End: * - * ex: set ts=8 sts=4 sw=4 noet: + * ex: set ts=8 sts=4 sw=4 et: */