X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/dcccc8ffbf2d700aa062eb88abe7b8cffc41ba48..ff25fcfb6e53726e91f2e25cef66d2957b596760:/handy.h diff --git a/handy.h b/handy.h index bcdf650..3e6fd52 100644 --- a/handy.h +++ b/handy.h @@ -96,19 +96,6 @@ Null SV pointer. (No longer available when C is defined.) # endif #endif -/* The NeXT dynamic loader headers will not build with the bool macro - So declare them now to clear confusion. -*/ -#if defined(NeXT) || defined(__NeXT__) -# undef FALSE -# undef TRUE - typedef enum bool { FALSE = 0, TRUE = 1 } bool; -# define ENUM_BOOL 1 -# ifndef HAS_BOOL -# define HAS_BOOL 1 -# endif /* !HAS_BOOL */ -#endif /* NeXT || __NeXT__ */ - #ifndef HAS_BOOL # ifdef bool # undef bool @@ -288,6 +275,7 @@ typedef U64TYPE U64; #define TYPE_DIGITS(T) BIT_DIGITS(sizeof(T) * 8) #define TYPE_CHARS(T) (TYPE_DIGITS(T) + 2) /* sign, NUL */ +/* Unused by core; should be deprecated */ #define Ctl(ch) ((ch) & 037) /* This is a helper macro to avoid preprocessor issues, replaced by nothing @@ -461,6 +449,16 @@ Test two strings to see if they are equal. The C parameter indicates the number of bytes to compare. Returns true or false. (A wrapper for C). +=for apidoc Am|bool|memEQ|char* s1|char* s2|STRLEN len +Test two buffers (which may contain embedded C characters, to see if they +are equal. The C parameter indicates the number of bytes to compare. +Returns zero if equal, or non-zero if non-equal. + +=for apidoc Am|bool|memNE|char* s1|char* s2|STRLEN len +Test two buffers (which may contain embedded C characters, to see if they +are not equal. The C parameter indicates the number of bytes to compare. +Returns zero if non-equal, or non-zero if equal. + =cut */ @@ -548,7 +546,9 @@ for. If Perl can determine that the current locale is a UTF-8 locale, it uses the published Unicode rules; otherwise, it uses the C library function that gives the named classification. For example, C when not in a UTF-8 locale returns the result of calling C. 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 is like C, but is defined on any UV. It returns the same as C for input code points less than 256, and @@ -661,9 +661,7 @@ C, and C. Returns a boolean indicating whether the specified character is a whitespace character. This is analogous to what C matches in a regular expression. Starting in Perl 5.18 -(experimentally), this also matches what C does. -("Experimentally" means that this change may be backed out in 5.22 if -field experience indicates that it was unwise.) Prior to 5.18, only the +this also matches what C does. Prior to 5.18, only the locale forms of this macro (the ones with C in their names) matched precisely what C does. In those releases, the only difference, in the non-locale variants, was that C did not match a vertical tab. @@ -675,10 +673,8 @@ C, and C. =for apidoc Am|bool|isPSXSPC|char ch (short for Posix Space) -Starting in 5.18, this is identical (experimentally) in all its forms to the -corresponding C macros. ("Experimentally" means that this change -may be backed out in 5.22 if field experience indicates that it -was unwise.) +Starting in 5.18, this is identical in all its forms to the +corresponding C macros. The locale forms of this macro are identical to their corresponding C forms in all Perl releases. In releases prior to 5.18, the non-locale forms differ from their C forms only in that the @@ -717,8 +713,9 @@ word character includes more than the standard C language meaning of alphanumeric. See the L for an explanation of variants -C, C, C, C, -C, C, and C. +C, C, C, and C. +C, C, and C 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 @@ -842,10 +839,6 @@ The first code point of the lowercased version is returned The input character at C

is assumed to be well-formed. -=for apidoc Am|U8|toLOWER_LC|U8 ch -Converts the specified character to lowercase using the current locale's rules, -if possible; otherwise returns the input character itself. - =for apidoc Am|U8|toTITLE|U8 ch Converts the specified character to titlecase. If the input is anything but an ASCII lowercase character, that input character itself is returned. Variant @@ -903,27 +896,22 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc #ifdef EBCDIC # ifndef _ALL_SOURCE - /* This returns the wrong results on at least z/OS unless this is - * defined. */ + /* The native libc isascii() et.al. functions return the wrong results + * on at least z/OS unless this is defined. */ # error _ALL_SOURCE should probably be defined # endif - - /* We could be called without perl.h, in which case NATIVE_TO_ASCII() is - * likely not defined, and so we use the native function */ -# define isASCII(c) cBOOL(isascii(c)) #else + /* There is a simple definition of ASCII for ASCII platforms. But the + * EBCDIC one isn't so simple, so is defined using table look-up like the + * other macros below */ # define isASCII(c) ((WIDEST_UTYPE)(c) < 128) #endif -#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 */ /* Character class numbers. For internal core Perl use only. The ones less @@ -957,27 +945,26 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc * useful to group these which have no members that match above Latin1, (or * above ASCII in the latter case) */ -# define _CC_SPACE 10 /* \s */ +# define _CC_SPACE 10 /* \s, [:space:] */ # define _CC_BLANK 11 /* [:blank:] */ # define _CC_XDIGIT 12 /* [:xdigit:] */ -# define _CC_PSXSPC 13 /* [:space:] */ -# define _CC_CNTRL 14 /* [:cntrl:] */ -# define _CC_ASCII 15 /* [:ascii:] */ -# define _CC_VERTSPACE 16 /* \v */ +# define _CC_CNTRL 13 /* [:cntrl:] */ +# define _CC_ASCII 14 /* [:ascii:] */ +# define _CC_VERTSPACE 15 /* \v */ # define _HIGHEST_REGCOMP_DOT_H_SYNC _CC_VERTSPACE /* The members of the third group below do not need to be coordinated with data * structures in regcomp.[ch] and regexec.c. */ -# define _CC_IDFIRST 17 -# define _CC_CHARNAME_CONT 18 -# define _CC_NONLATIN1_FOLD 19 -# define _CC_NONLATIN1_SIMPLE_FOLD 20 -# 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_IDFIRST 16 +# define _CC_CHARNAME_CONT 17 +# define _CC_NONLATIN1_FOLD 18 +# define _CC_NONLATIN1_SIMPLE_FOLD 19 +# define _CC_QUOTEMETA 20 +# define _CC_NON_FINAL_FOLD 21 +# define _CC_IS_IN_SOME_FOLD 22 +# define _CC_MNEMONIC_CNTRL 23 +/* Unused: 24-31 * 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 @@ -1000,7 +987,6 @@ typedef enum { _CC_ENUM_GRAPH = _CC_GRAPH, _CC_ENUM_LOWER = _CC_LOWER, _CC_ENUM_PRINT = _CC_PRINT, - _CC_ENUM_PSXSPC = _CC_PSXSPC, _CC_ENUM_PUNCT = _CC_PUNCT, _CC_ENUM_SPACE = _CC_SPACE, _CC_ENUM_UPPER = _CC_UPPER, @@ -1038,6 +1024,7 @@ static const char* const swash_property_names[] = { }; #endif +START_EXTERN_C # ifdef DOINIT EXTCONST U32 PL_charclass[] = { # include "l1_char_class_tab.h" @@ -1046,6 +1033,7 @@ EXTCONST U32 PL_charclass[] = { # else /* ! DOINIT */ EXTCONST U32 PL_charclass[]; # endif +END_EXTERN_C /* The 1U keeps Solaris from griping when shifting sets the uppermost bit */ # define _CC_mask(classnum) (1U << (classnum)) @@ -1070,16 +1058,15 @@ EXTCONST U32 PL_charclass[]; # 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 isPSXSPC_A(c) _generic_isCC_A(c, _CC_PSXSPC) # define isPUNCT_A(c) _generic_isCC_A(c, _CC_PUNCT) # 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) @@ -1092,13 +1079,17 @@ EXTCONST U32 PL_charclass[]; # define isGRAPH_L1(c) _generic_isCC(c, _CC_GRAPH) # define isLOWER_L1(c) _generic_isCC(c, _CC_LOWER) # define isPRINT_L1(c) _generic_isCC(c, _CC_PRINT) -# define isPSXSPC_L1(c) _generic_isCC(c, _CC_PSXSPC) +# define isPSXSPC_L1(c) isSPACE_L1(c) # define isPUNCT_L1(c) _generic_isCC(c, _CC_PUNCT) # define isSPACE_L1(c) _generic_isCC(c, _CC_SPACE) # define isUPPER_L1(c) _generic_isCC(c, _CC_UPPER) # define isWORDCHAR_L1(c) _generic_isCC(c, _CC_WORDCHAR) # define isIDFIRST_L1(c) _generic_isCC(c, _CC_IDFIRST) +# ifdef EBCDIC +# define isASCII(c) _generic_isCC(c, _CC_ASCII) +# endif + /* Participates in a single-character fold with a character above 255 */ # define _HAS_NONLATIN1_SIMPLE_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(c) ((! cBOOL(FITS_IN_8_BITS(c))) || (PL_charclass[(U8) (c)] & _CC_mask(_CC_NONLATIN1_SIMPLE_FOLD))) @@ -1110,44 +1101,97 @@ EXTCONST U32 PL_charclass[]; _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 - * to it. */ -# ifdef EBCDIC - /* Use the native functions. They likely will return false for all - * non-ASCII values, but this makes sure */ -# define isLOWER_A(c) (isASCII(c) && islower(c)) -# define isPRINT_A(c) (isASCII(c) && isprint(c)) -# define isUPPER_A(c) (isASCII(c) && isupper(c)) -# else /* ASCII platform. These are coded based on first principals */ + * to it. Most are coded based on first principals. First some ones common + * to both ASCII and EBCDIC */ +# define isDIGIT_A(c) ((c) <= '9' && (c) >= '0') +# define isBLANK_A(c) ((c) == ' ' || (c) == '\t') +# define isSPACE_A(c) (isBLANK_A(c) \ + || (c) == '\n' \ + || (c) == '\r' \ + || (c) == '\v' \ + || (c) == '\f') +# ifdef EBCDIC /* There are gaps between 'i' and 'j'; 'r' and 's'. Same + for uppercase. This is ordered to exclude most things + early */ +# define isLOWER_A(c) ((c) >= 'a' && (c) <= 'z' \ + && ((c) <= 'i' \ + || ((c) >= 'j' && (c) <= 'r') \ + || (c) >= 's')) +# define isUPPER_A(c) ((c) >= 'A' && (c) <= 'Z' \ + && ((c) <= 'I' \ + || ((c) >= 'J' && (c) <= 'R') \ + || (c) >= 'S')) +# else /* ASCII platform. */ # define isLOWER_A(c) ((c) >= 'a' && (c) <= 'z') -# define isPRINT_A(c) (((c) >= 32 && (c) < 127)) # define isUPPER_A(c) ((c) <= 'Z' && (c) >= 'A') -# endif /* Below are common definitions for ASCII and non-ASCII */ +# endif + + /* Some more ASCII, non-ASCII common definitions */ # define isALPHA_A(c) (isUPPER_A(c) || isLOWER_A(c)) # define isALPHANUMERIC_A(c) (isALPHA_A(c) || isDIGIT_A(c)) -# define isBLANK_A(c) ((c) == ' ' || (c) == '\t') -# define isCNTRL_A(c) (isASCII(c) && (! isPRINT_A(c))) -# define isDIGIT_A(c) ((c) <= '9' && (c) >= '0') -# define isGRAPH_A(c) (isPRINT_A(c) && (c) != ' ') -# define isIDFIRST_A(c) (isALPHA_A(c) || (c) == '_') -# define isPUNCT_A(c) (isGRAPH_A(c) && (! isALPHANUMERIC_A(c))) -# define isSPACE_A(c) ((c) == ' ' \ - || (c) == '\t' \ - || (c) == '\n' \ - || (c) == '\r' \ - || (c) == '\v' \ - || (c) == '\f') # define isWORDCHAR_A(c) (isALPHANUMERIC_A(c) || (c) == '_') -# define isXDIGIT_A(c) (isDIGIT_A(c) \ - || ((c) >= 'a' && (c) <= 'f') \ +# define isIDFIRST_A(c) (isALPHA_A(c) || (c) == '_') +# define isXDIGIT_A(c) (isDIGIT_A(c) \ + || ((c) >= 'a' && (c) <= 'f') \ || ((c) <= 'F' && (c) >= 'A')) +# ifdef EBCDIC +# define isPUNCT_A(c) ((c) == '-' || (c) == '!' || (c) == '"' \ + || (c) == '#' || (c) == '$' || (c) == '%' \ + || (c) == '&' || (c) == '\'' || (c) == '(' \ + || (c) == ')' || (c) == '*' || (c) == '+' \ + || (c) == ',' || (c) == '.' || (c) == '/' \ + || (c) == ':' || (c) == ';' || (c) == '<' \ + || (c) == '=' || (c) == '>' || (c) == '?' \ + || (c) == '@' || (c) == '[' || (c) == '\\' \ + || (c) == ']' || (c) == '^' || (c) == '_' \ + || (c) == '`' || (c) == '{' || (c) == '|' \ + || (c) == '}' || (c) == '~') +# define isGRAPH_A(c) (isALPHANUMERIC_A(c) || isPUNCT_A(c)) +# define isPRINT_A(c) (isGRAPH_A(c) || (c) == ' ') + +# ifdef QUESTION_MARK_CTRL +# define _isQMC(c) ((c) == QUESTION_MARK_CTRL) +# else +# define _isQMC(c) 0 +# endif + + /* I (khw) can't think of a way to define all the ASCII controls + * without resorting to a libc (locale-sensitive) call. But we know + * that all controls but the question-mark one are in the range 0-0x3f. + * This makes sure that all the controls that have names are included, + * and all controls that are also considered ASCII in the locale. This + * may include more or fewer than what it actually should, but the + * wrong ones are less-important controls, so likely won't impact + * things (keep in mind that this is compiled only if perl.h isn't + * available). The question mark control is included if available */ +# define isCNTRL_A(c) (((c) < 0x40 && isascii(c)) \ + || (c) == '\0' || (c) == '\a' || (c) == '\b' \ + || (c) == '\f' || (c) == '\n' || (c) == '\r' \ + || (c) == '\t' || (c) == '\v' || _isQMC(c)) + +# define isASCII(c) (isCNTRL_A(c) || isPRINT_A(c)) +# else /* ASCII platform; things are simpler, and isASCII has already + been defined */ +# define isGRAPH_A(c) (((c) > ' ' && (c) < 127)) +# define isPRINT_A(c) (isGRAPH_A(c) || (c) == ' ') +# define isPUNCT_A(c) (isGRAPH_A(c) && (! isALPHANUMERIC_A(c))) +# define isCNTRL_A(c) (isASCII(c) && (! isPRINT_A(c))) +# endif + /* The _L1 macros may be unnecessary for the utilities; I (khw) added them - * during debugging, and it seems best to keep them. */ -# define isPSXSPC_A(c) isSPACE_A(c) /* XXX Assumes SPACE matches '\v' */ + * 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 isALPHA_L1(c) (isUPPER_L1(c) || isLOWER_L1(c)) # define isALPHANUMERIC_L1(c) (isALPHA_L1(c) || isDIGIT_A(c)) # define isBLANK_L1(c) (isBLANK_A(c) \ @@ -1165,7 +1209,6 @@ EXTCONST U32 PL_charclass[]; # define isPRINT_L1(c) (isPRINT_A(c) \ || (FITS_IN_8_BITS(c) \ && NATIVE_TO_LATIN1((U8) c) >= 0xA0)) -# define isPSXSPC_L1(c) isSPACE_L1(c) # define isPUNCT_L1(c) (isPUNCT_A(c) \ || (FITS_IN_8_BITS(c) \ && (NATIVE_TO_LATIN1((U8) c) == 0xA1 \ @@ -1200,6 +1243,7 @@ EXTCONST U32 PL_charclass[]; /* 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 @@ -1213,10 +1257,12 @@ EXTCONST U32 PL_charclass[]; (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) +#define isASCII_A(c) isASCII(c) +#define isASCII_L1(c) isASCII(c) #define isBLANK(c) isBLANK_A(c) #define isCNTRL(c) isCNTRL_A(c) #define isDIGIT(c) isDIGIT_A(c) @@ -1224,7 +1270,9 @@ EXTCONST U32 PL_charclass[]; #define isIDFIRST(c) isIDFIRST_A(c) #define isLOWER(c) isLOWER_A(c) #define isPRINT(c) isPRINT_A(c) +#define isPSXSPC_A(c) isSPACE_A(c) #define isPSXSPC(c) isPSXSPC_A(c) +#define isPSXSPC_L1(c) isSPACE_L1(c) #define isPUNCT(c) isPUNCT_A(c) #define isSPACE(c) isSPACE_A(c) #define isUPPER(c) isUPPER_A(c) @@ -1304,7 +1352,7 @@ EXTCONST U32 PL_charclass[]; ? (c) \ : (IN_UTF8_CTYPE_LOCALE) \ ? PL_latin1_lc[ (U8) (c) ] \ - : function((cast)(c))) + : (cast)function((cast)(c))) /* Note that the result can be larger than a byte in a UTF-8 locale. It * returns a single value, so can't adequately return the upper case of LATIN @@ -1315,7 +1363,7 @@ EXTCONST U32 PL_charclass[]; (! FITS_IN_8_BITS(c) \ ? (c) \ : ((! IN_UTF8_CTYPE_LOCALE) \ - ? function((cast)(c)) \ + ? (cast)function((cast)(c)) \ : ((((U8)(c)) == MICRO_SIGN) \ ? GREEK_CAPITAL_LETTER_MU \ : ((((U8)(c)) == LATIN_SMALL_LETTER_Y_WITH_DIAERESIS) \ @@ -1337,46 +1385,21 @@ EXTCONST U32 PL_charclass[]; _generic_toLOWER_LC(c, function, cast))) /* Use the libc versions for these if available. */ -#if defined(HAS_ISASCII) && ! defined(USE_NEXT_CTYPE) +#if defined(HAS_ISASCII) # define isASCII_LC(c) (FITS_IN_8_BITS(c) && isascii( (U8) (c))) #else # define isASCII_LC(c) isASCII(c) #endif -#if defined(HAS_ISBLANK) && ! defined(USE_NEXT_CTYPE) +#if defined(HAS_ISBLANK) # define isBLANK_LC(c) _generic_LC(c, _CC_BLANK, isblank) #else /* Unlike isASCII, varies if in a UTF-8 locale */ -# define isBLANK_LC(c) (IN_UTF8_CTYPE_LOCALE) ? isBLANK_L1(c) : isBLANK(c) +# define isBLANK_LC(c) ((IN_UTF8_CTYPE_LOCALE) ? isBLANK_L1(c) : isBLANK(c)) #endif -#ifdef USE_NEXT_CTYPE /* NeXT computers */ - -# define _LC_CAST unsigned int /* Needed by _generic_LC. NeXT functions - use this as their input type */ - -# define isALPHA_LC(c) _generic_LC(c, _CC_ALPHA, NXIsAlpha) -# define isALPHANUMERIC_LC(c) _generic_LC(c, _CC_ALPHANUMERIC, NXIsAlNum) -# define isCNTRL_LC(c) _generic_LC(c, _CC_CNTRL, NXIsCntrl) -# define isDIGIT_LC(c) _generic_LC(c, _CC_DIGIT, NXIsDigit) -# define isGRAPH_LC(c) _generic_LC(c, _CC_GRAPH, NXIsGraph) -# define isIDFIRST_LC(c) _generic_LC_underscore(c, _CC_IDFIRST, NXIsAlpha) -# define isLOWER_LC(c) _generic_LC(c, _CC_LOWER, NXIsLower) -# define isPRINT_LC(c) _generic_LC(c, _CC_PRINT, NXIsPrint) -# define isPUNCT_LC(c) _generic_LC(c, _CC_PUNCT, NXIsPunct) -# define isSPACE_LC(c) _generic_LC(c, _CC_SPACE, NXIsSpace) -# define isUPPER_LC(c) _generic_LC(c, _CC_UPPER, NXIsUpper) -# define isWORDCHAR_LC(c) _generic_LC_underscore(c, _CC_WORDCHAR, NXIsAlNum) -# define isXDIGIT_LC(c) _generic_LC(c, _CC_XDIGIT, NXIsXdigit) - -# define toLOWER_LC(c) _generic_toLOWER_LC((c), NXToLower, unsigned int) -# define toUPPER_LC(c) _generic_toUPPER_LC((c), NXToUpper, unsigned int) -# define toFOLD_LC(c) _generic_toFOLD_LC((c), NXToLower, unsigned int) +#define _LC_CAST U8 -#else /* !USE_NEXT_CTYPE */ - -# define _LC_CAST U8 - -# ifdef WIN32 +#ifdef WIN32 /* The Windows functions don't bother to follow the POSIX standard, which * for example says that something can't both be a printable and a control. * But Windows treats the \t control as a printable, and does such things @@ -1387,69 +1410,68 @@ EXTCONST U32 PL_charclass[]; * Not all possible weirdnesses are checked for, just the ones that were * detected on actual Microsoft code pages */ -# define isCNTRL_LC(c) _generic_LC(c, _CC_CNTRL, iscntrl) -# define isSPACE_LC(c) _generic_LC(c, _CC_SPACE, isspace) - -# define isALPHA_LC(c) (_generic_LC(c, _CC_ALPHA, isalpha) && isALPHANUMERIC_LC(c)) -# define isALPHANUMERIC_LC(c) (_generic_LC(c, _CC_ALPHANUMERIC, isalnum) && ! isPUNCT_LC(c)) -# define isDIGIT_LC(c) (_generic_LC(c, _CC_DIGIT, isdigit) && isALPHANUMERIC_LC(c)) -# define isGRAPH_LC(c) (_generic_LC(c, _CC_GRAPH, isgraph) && isPRINT_LC(c)) -# define isIDFIRST_LC(c) (((c) == '_') || (_generic_LC(c, _CC_IDFIRST, isalpha) && ! isPUNCT_LC(c))) -# define isLOWER_LC(c) (_generic_LC(c, _CC_LOWER, islower) && isALPHA_LC(c)) -# define isPRINT_LC(c) (_generic_LC(c, _CC_PRINT, isprint) && ! isCNTRL_LC(c)) -# define isPUNCT_LC(c) (_generic_LC(c, _CC_PUNCT, ispunct) && ! isCNTRL_LC(c)) -# define isUPPER_LC(c) (_generic_LC(c, _CC_UPPER, isupper) && isALPHA_LC(c)) -# define isWORDCHAR_LC(c) (((c) == '_') || isALPHANUMERIC_LC(c)) -# define isXDIGIT_LC(c) (_generic_LC(c, _CC_XDIGIT, isxdigit) && isALPHANUMERIC_LC(c)) - -# define toLOWER_LC(c) _generic_toLOWER_LC((c), tolower, U8) -# define toUPPER_LC(c) _generic_toUPPER_LC((c), toupper, U8) -# define toFOLD_LC(c) _generic_toFOLD_LC((c), tolower, U8) - -# elif defined(CTYPE256) || (!defined(isascii) && !defined(HAS_ISASCII)) +# define isCNTRL_LC(c) _generic_LC(c, _CC_CNTRL, iscntrl) +# define isSPACE_LC(c) _generic_LC(c, _CC_SPACE, isspace) + +# define isALPHA_LC(c) (_generic_LC(c, _CC_ALPHA, isalpha) && isALPHANUMERIC_LC(c)) +# define isALPHANUMERIC_LC(c) (_generic_LC(c, _CC_ALPHANUMERIC, isalnum) && ! isPUNCT_LC(c)) +# define isDIGIT_LC(c) (_generic_LC(c, _CC_DIGIT, isdigit) && isALPHANUMERIC_LC(c)) +# define isGRAPH_LC(c) (_generic_LC(c, _CC_GRAPH, isgraph) && isPRINT_LC(c)) +# define isIDFIRST_LC(c) (((c) == '_') || (_generic_LC(c, _CC_IDFIRST, isalpha) && ! isPUNCT_LC(c))) +# define isLOWER_LC(c) (_generic_LC(c, _CC_LOWER, islower) && isALPHA_LC(c)) +# define isPRINT_LC(c) (_generic_LC(c, _CC_PRINT, isprint) && ! isCNTRL_LC(c)) +# define isPUNCT_LC(c) (_generic_LC(c, _CC_PUNCT, ispunct) && ! isCNTRL_LC(c)) +# define isUPPER_LC(c) (_generic_LC(c, _CC_UPPER, isupper) && isALPHA_LC(c)) +# define isWORDCHAR_LC(c) (((c) == '_') || isALPHANUMERIC_LC(c)) +# define isXDIGIT_LC(c) (_generic_LC(c, _CC_XDIGIT, isxdigit) && isALPHANUMERIC_LC(c)) + +# define toLOWER_LC(c) _generic_toLOWER_LC((c), tolower, U8) +# define toUPPER_LC(c) _generic_toUPPER_LC((c), toupper, U8) +# define toFOLD_LC(c) _generic_toFOLD_LC((c), tolower, U8) + +#elif defined(CTYPE256) || (!defined(isascii) && !defined(HAS_ISASCII)) /* For most other platforms */ -# define isALPHA_LC(c) _generic_LC(c, _CC_ALPHA, isalpha) -# define isALPHANUMERIC_LC(c) _generic_LC(c, _CC_ALPHANUMERIC, isalnum) -# define isCNTRL_LC(c) _generic_LC(c, _CC_CNTRL, iscntrl) -# define isDIGIT_LC(c) _generic_LC(c, _CC_DIGIT, isdigit) -# define isGRAPH_LC(c) _generic_LC(c, _CC_GRAPH, isgraph) -# define isIDFIRST_LC(c) _generic_LC_underscore(c, _CC_IDFIRST, isalpha) -# define isLOWER_LC(c) _generic_LC(c, _CC_LOWER, islower) -# define isPRINT_LC(c) _generic_LC(c, _CC_PRINT, isprint) -# define isPUNCT_LC(c) _generic_LC(c, _CC_PUNCT, ispunct) -# define isSPACE_LC(c) _generic_LC(c, _CC_SPACE, isspace) -# define isUPPER_LC(c) _generic_LC(c, _CC_UPPER, isupper) -# define isWORDCHAR_LC(c) _generic_LC_underscore(c, _CC_WORDCHAR, isalnum) -# define isXDIGIT_LC(c) _generic_LC(c, _CC_XDIGIT, isxdigit) - - -# define toLOWER_LC(c) _generic_toLOWER_LC((c), tolower, U8) -# define toUPPER_LC(c) _generic_toUPPER_LC((c), toupper, U8) -# define toFOLD_LC(c) _generic_toFOLD_LC((c), tolower, U8) - -# else /* The final fallback position */ - -# define isALPHA_LC(c) (isascii(c) && isalpha(c)) -# define isALPHANUMERIC_LC(c) (isascii(c) && isalnum(c)) -# define isCNTRL_LC(c) (isascii(c) && iscntrl(c)) -# define isDIGIT_LC(c) (isascii(c) && isdigit(c)) -# define isGRAPH_LC(c) (isascii(c) && isgraph(c)) -# define isIDFIRST_LC(c) (isascii(c) && (isalpha(c) || (c) == '_')) -# define isLOWER_LC(c) (isascii(c) && islower(c)) -# define isPRINT_LC(c) (isascii(c) && isprint(c)) -# define isPUNCT_LC(c) (isascii(c) && ispunct(c)) -# define isSPACE_LC(c) (isascii(c) && isspace(c)) -# define isUPPER_LC(c) (isascii(c) && isupper(c)) -# define isWORDCHAR_LC(c) (isascii(c) && (isalnum(c) || (c) == '_')) -# define isXDIGIT_LC(c) (isascii(c) && isxdigit(c)) - -# define toLOWER_LC(c) (isascii(c) ? tolower(c) : (c)) -# define toUPPER_LC(c) (isascii(c) ? toupper(c) : (c)) -# define toFOLD_LC(c) (isascii(c) ? tolower(c) : (c)) +# define isALPHA_LC(c) _generic_LC(c, _CC_ALPHA, isalpha) +# define isALPHANUMERIC_LC(c) _generic_LC(c, _CC_ALPHANUMERIC, isalnum) +# define isCNTRL_LC(c) _generic_LC(c, _CC_CNTRL, iscntrl) +# define isDIGIT_LC(c) _generic_LC(c, _CC_DIGIT, isdigit) +# define isGRAPH_LC(c) _generic_LC(c, _CC_GRAPH, isgraph) +# define isIDFIRST_LC(c) _generic_LC_underscore(c, _CC_IDFIRST, isalpha) +# define isLOWER_LC(c) _generic_LC(c, _CC_LOWER, islower) +# define isPRINT_LC(c) _generic_LC(c, _CC_PRINT, isprint) +# define isPUNCT_LC(c) _generic_LC(c, _CC_PUNCT, ispunct) +# define isSPACE_LC(c) _generic_LC(c, _CC_SPACE, isspace) +# define isUPPER_LC(c) _generic_LC(c, _CC_UPPER, isupper) +# define isWORDCHAR_LC(c) _generic_LC_underscore(c, _CC_WORDCHAR, isalnum) +# define isXDIGIT_LC(c) _generic_LC(c, _CC_XDIGIT, isxdigit) + + +# define toLOWER_LC(c) _generic_toLOWER_LC((c), tolower, U8) +# define toUPPER_LC(c) _generic_toUPPER_LC((c), toupper, U8) +# define toFOLD_LC(c) _generic_toFOLD_LC((c), tolower, U8) + +#else /* The final fallback position */ + +# define isALPHA_LC(c) (isascii(c) && isalpha(c)) +# define isALPHANUMERIC_LC(c) (isascii(c) && isalnum(c)) +# define isCNTRL_LC(c) (isascii(c) && iscntrl(c)) +# define isDIGIT_LC(c) (isascii(c) && isdigit(c)) +# define isGRAPH_LC(c) (isascii(c) && isgraph(c)) +# define isIDFIRST_LC(c) (isascii(c) && (isalpha(c) || (c) == '_')) +# define isLOWER_LC(c) (isascii(c) && islower(c)) +# define isPRINT_LC(c) (isascii(c) && isprint(c)) +# define isPUNCT_LC(c) (isascii(c) && ispunct(c)) +# define isSPACE_LC(c) (isascii(c) && isspace(c)) +# define isUPPER_LC(c) (isascii(c) && isupper(c)) +# define isWORDCHAR_LC(c) (isascii(c) && (isalnum(c) || (c) == '_')) +# define isXDIGIT_LC(c) (isascii(c) && isxdigit(c)) + +# define toLOWER_LC(c) (isascii(c) ? tolower(c) : (c)) +# define toUPPER_LC(c) (isascii(c) ? toupper(c) : (c)) +# define toFOLD_LC(c) (isascii(c) ? tolower(c) : (c)) -# endif -#endif /* USE_NEXT_CTYPE */ +#endif #define isIDCONT(c) isWORDCHAR(c) #define isIDCONT_A(c) isWORDCHAR_A(c) @@ -1489,11 +1511,10 @@ EXTCONST U32 PL_charclass[]; #define isLOWER_uni(c) _generic_swash_uni(_CC_LOWER, c) #define isPRINT_uni(c) _generic_swash_uni(_CC_PRINT, c) -/* Posix and regular space are identical above Latin1 */ -#define isPSXSPC_uni(c) _generic_uni(_CC_PSXSPC, is_XPERLSPACE_cp_high, c) - #define isPUNCT_uni(c) _generic_swash_uni(_CC_PUNCT, c) #define isSPACE_uni(c) _generic_uni(_CC_SPACE, is_XPERLSPACE_cp_high, c) +#define isPSXSPC_uni(c) isSPACE_uni(c) + #define isUPPER_uni(c) _generic_swash_uni(_CC_UPPER, c) #define isVERTWS_uni(c) _generic_uni(_CC_VERTSPACE, is_VERTWS_cp_high, c) #define isWORDCHAR_uni(c) _generic_swash_uni(_CC_WORDCHAR, c) @@ -1528,8 +1549,7 @@ EXTCONST U32 PL_charclass[]; _is_uni_perl_idstart, c) #define isLOWER_LC_uvchr(c) _generic_LC_swash_uvchr(isLOWER_LC, _CC_LOWER, c) #define isPRINT_LC_uvchr(c) _generic_LC_swash_uvchr(isPRINT_LC, _CC_PRINT, c) -#define isPSXSPC_LC_uvchr(c) isSPACE_LC_uvchr(c) /* space is identical to posix - space under locale */ +#define isPSXSPC_LC_uvchr(c) isSPACE_LC_uvchr(c) #define isPUNCT_LC_uvchr(c) _generic_LC_swash_uvchr(isPUNCT_LC, _CC_PUNCT, c) #define isSPACE_LC_uvchr(c) _generic_LC_uvchr(isSPACE_LC, \ is_XPERLSPACE_cp_high, c) @@ -1593,7 +1613,7 @@ EXTCONST U32 PL_charclass[]; #ifdef EBCDIC /* Because all controls are UTF-8 invariants in EBCDIC, we can use this * more efficient macro instead of the more general one */ -# define isCNTRL_utf8(p) isCNTRL_L1(p) +# define isCNTRL_utf8(p) isCNTRL_L1(*(p)) #else # define isCNTRL_utf8(p) _generic_utf8(_CC_CNTRL, p, 0) #endif @@ -1615,10 +1635,7 @@ EXTCONST U32 PL_charclass[]; #define isLOWER_utf8(p) _generic_swash_utf8(_CC_LOWER, p) #define isPRINT_utf8(p) _generic_swash_utf8(_CC_PRINT, p) - -/* Posix and regular space are identical above Latin1 */ -#define isPSXSPC_utf8(p) _generic_func_utf8(_CC_PSXSPC, is_XPERLSPACE_high, p) - +#define isPSXSPC_utf8(p) isSPACE_utf8(p) #define isPUNCT_utf8(p) _generic_swash_utf8(_CC_PUNCT, p) #define isSPACE_utf8(p) _generic_func_utf8(_CC_SPACE, is_XPERLSPACE_high, p) #define isUPPER_utf8(p) _generic_swash_utf8(_CC_UPPER, p) @@ -1660,8 +1677,7 @@ EXTCONST U32 PL_charclass[]; #define isIDFIRST_LC_utf8(p) _generic_LC_func_utf8(isIDFIRST_LC, _is_utf8_perl_idstart, p) #define isLOWER_LC_utf8(p) _generic_LC_swash_utf8(isLOWER_LC, _CC_LOWER, p) #define isPRINT_LC_utf8(p) _generic_LC_swash_utf8(isPRINT_LC, _CC_PRINT, p) -#define isPSXSPC_LC_utf8(p) isSPACE_LC_utf8(p) /* space is identical to posix - space under locale */ +#define isPSXSPC_LC_utf8(p) isSPACE_LC_utf8(p) #define isPUNCT_LC_utf8(p) _generic_LC_swash_utf8(isPUNCT_LC, _CC_PUNCT, p) #define isSPACE_LC_utf8(p) _generic_LC_func_utf8(isSPACE_LC, is_XPERLSPACE_high, p) #define isUPPER_LC_utf8(p) _generic_LC_swash_utf8(isUPPER_LC, _CC_UPPER, p) @@ -1699,17 +1715,19 @@ EXTCONST U32 PL_charclass[]; * the outlier from the block that contains the other controls, just like * toCTRL('?') on ASCII yields DEL, the control that is the outlier from the C0 * block. If it weren't special cased, it would yield a non-control. - * The conversion works both ways, so CTRL('D') is 4, and CTRL(4) is D, etc. */ + * The conversion works both ways, so toCTRL('D') is 4, and toCTRL(4) is D, + * etc. */ #ifndef EBCDIC -# define toCTRL(c) (toUPPER(c) ^ 64) +# define toCTRL(c) (__ASSERT_(FITS_IN_8_BITS(c)) toUPPER(((U8)(c))) ^ 64) #else -# define toCTRL(c) ((isPRINT_A(c)) \ - ? UNLIKELY((c) == '?') \ - ? QUESTION_MARK_CTRL \ - : (NATIVE_TO_LATIN1(toUPPER(c)) ^ 64) \ - : UNLIKELY((c) == QUESTION_MARK_CTRL) \ - ? ((c) == '?') \ - : (LATIN1_TO_NATIVE((c) ^ 64))) +# define toCTRL(c) (__ASSERT_(FITS_IN_8_BITS(c)) \ + ((isPRINT_A(c)) \ + ? (UNLIKELY((c) == '?') \ + ? QUESTION_MARK_CTRL \ + : (NATIVE_TO_LATIN1(toUPPER((U8) (c))) ^ 64)) \ + : (UNLIKELY((c) == QUESTION_MARK_CTRL) \ + ? '?' \ + : (LATIN1_TO_NATIVE(((U8) (c)) ^ 64))))) #endif /* Line numbers are unsigned, 32 bits. */ @@ -1746,6 +1764,25 @@ typedef U32 line_t; * 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 @@ -1849,14 +1886,49 @@ PoisonWith(0xEF) for catching access to freed memory. #define MEM_SIZE_MAX ((MEM_SIZE)~0) -/* The +0.0 in MEM_WRAP_CHECK_ is an attempt to foil - * overly eager compilers that will bleat about e.g. - * (U16)n > (size_t)~0/sizeof(U16) always being false. */ + #ifdef PERL_MALLOC_WRAP -#define MEM_WRAP_CHECK(n,t) \ - (void)(UNLIKELY(sizeof(t) > 1 && ((MEM_SIZE)(n)+0.0) > MEM_SIZE_MAX/sizeof(t)) && (croak_memory_wrap(),0)) -#define MEM_WRAP_CHECK_1(n,t,a) \ - (void)(UNLIKELY(sizeof(t) > 1 && ((MEM_SIZE)(n)+0.0) > MEM_SIZE_MAX/sizeof(t)) && (Perl_croak_nocontext("%s",(a)),0)) + +/* This expression will be constant-folded at compile time. It checks + * whether or not the type of the count n is so small (e.g. U8 or U16, or + * U32 on 64-bit systems) that there's no way a wrap-around could occur. + * As well as avoiding the need for a run-time check in some cases, it's + * designed to avoid compiler warnings like: + * comparison is always false due to limited range of data type + */ + +# define _MEM_WRAP_NEEDS_RUNTIME_CHECK(n,t) \ + (sizeof(t) > ((MEM_SIZE)1 << 8*(sizeof(MEM_SIZE) - sizeof(n)))) + +/* This is written in a slightly odd way to avoid various spurious + * compiler warnings. We *want* to write the expression as + * _MEM_WRAP_NEEDS_RUNTIME_CHECK(n,t) && (n > C) + * (for some compile-time constant C), but even when the LHS + * constant-folds to false at compile-time, g++ insists on emitting + * warnings about the RHS (e.g. "comparison is always false"), so instead + * we write it as + * + * (cond ? n : X) > C + * + * where X is a constant with X > C always false. Choosing a value for X + * is tricky. If 0, some compilers will complain about 0 > C always being + * false; if 1, Coverity complains when n happens to be the constant value + * '1', that cond ? 1 : 1 has the same value on both branches; so use C + * for X and hope that nothing else whines. + */ + +# define _MEM_WRAP_WILL_WRAP(n,t) \ + ((_MEM_WRAP_NEEDS_RUNTIME_CHECK(n,t) ? (MEM_SIZE)(n) : \ + MEM_SIZE_MAX/sizeof(t)) > MEM_SIZE_MAX/sizeof(t)) + +# define MEM_WRAP_CHECK(n,t) \ + (void)(UNLIKELY(_MEM_WRAP_WILL_WRAP(n,t)) \ + && (croak_memory_wrap(),0)) + +# define MEM_WRAP_CHECK_1(n,t,a) \ + (void)(UNLIKELY(_MEM_WRAP_WILL_WRAP(n,t)) \ + && (Perl_croak_nocontext("%s",(a)),0)) + #define MEM_WRAP_CHECK_(n,t) MEM_WRAP_CHECK(n,t), #define PERL_STRLEN_ROUNDUP(n) ((void)(((n) > MEM_SIZE_MAX - 2 * PERL_STRLEN_ROUNDUP_QUANTUM) ? (croak_memory_wrap(),0):0),((n-1+PERL_STRLEN_ROUNDUP_QUANTUM)&~((MEM_SIZE)PERL_STRLEN_ROUNDUP_QUANTUM-1))) @@ -2075,11 +2147,5 @@ void Perl_mem_log_del_sv(const SV *sv, const char *filename, const int linenumbe #endif /* HANDY_H */ /* - * Local variables: - * c-indentation-style: bsd - * c-basic-offset: 4 - * indent-tabs-mode: nil - * End: - * * ex: set ts=8 sts=4 sw=4 et: */