X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/7e75d1a1924b20bbe4b3b5d46333dd5e4b674e25..1b7e229402b233ace142ad59f19a135d3d390e03:/handy.h diff --git a/handy.h b/handy.h index 228662f..e2fb0aa 100644 --- a/handy.h +++ b/handy.h @@ -11,15 +11,8 @@ /* IMPORTANT NOTE: Everything whose name begins with an underscore is for * internal core Perl use only. */ -#ifndef HANDY_H /* Guard against nested #inclusion */ -#define HANDY_H - -#if !defined(__STDC__) -#ifdef NULL -#undef NULL -#endif -# define NULL 0 -#endif +#ifndef PERL_HANDY_H_ /* Guard against nested #inclusion */ +#define PERL_HANDY_H_ #ifndef PERL_CORE # define Null(type) ((type)NULL) @@ -27,11 +20,11 @@ /* =head1 Handy Values -=for apidoc AmU||Nullch +=for apidoc AmnU||Nullch Null character pointer. (No longer available when C is defined.) -=for apidoc AmU||Nullsv +=for apidoc AmnU||Nullsv Null SV pointer. (No longer available when C is defined.) =cut @@ -104,10 +97,17 @@ Null SV pointer. (No longer available when C is defined.) # define HAS_BOOL 1 #endif -/* cast-to-bool. 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 */ +/* +=for apidoc Am|bool|cBOOL|bool expr + +Cast-to-bool. A simple S>> cast may not do the right thing: +if C is defined as C, for example, then the cast from C is +implementation-defined. + +C<(bool)!!(cbool)> in a ternary triggers a bug in xlc on AIX + +=cut +*/ #define cBOOL(cbool) ((cbool) ? (bool)1 : (bool)0) /* Try to figure out __func__ or __FUNCTION__ equivalent, if any. @@ -116,13 +116,11 @@ Null SV pointer. (No longer available when C is defined.) * XXX Similarly, a Configure probe for __FILE__ and __LINE__ is needed. */ #if (defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L) || (defined(__SUNPRO_C)) /* C99 or close enough. */ # define FUNCTION__ __func__ +#elif (defined(USING_MSVC6)) || /* MSVC6 has neither __func__ nor __FUNCTION and no good workarounds, either. */ \ + (defined(__DECC_VER)) /* Tru64 or VMS, and strict C89 being used, but not modern enough cc (in Tur64, -c99 not known, only -std1). */ +# define FUNCTION__ "" #else -# if (defined(USING_MSVC6)) || /* MSVC6 has neither __func__ nor __FUNCTION and no good workarounds, either. */ \ - (defined(__DECC_VER)) /* Tru64 or VMS, and strict C89 being used, but not modern enough cc (in Tur64, -c99 not known, only -std1). */ -# define FUNCTION__ "" -# else -# define FUNCTION__ __FUNCTION__ /* Common extension. */ -# endif +# define FUNCTION__ __FUNCTION__ /* Common extension. */ #endif /* XXX A note on the perl source internal type system. The @@ -174,54 +172,11 @@ typedef U16TYPE U16; typedef I32TYPE I32; typedef U32TYPE U32; -#ifdef HAS_QUAD +#ifdef QUADKIND typedef I64TYPE I64; typedef U64TYPE U64; #endif -/* INT64_C/UINT64_C are C99 from (so they will not be - * available in strict C89 mode), but they are nice, so let's define - * them if necessary. */ -#if defined(HAS_QUAD) -# undef PeRl_INT64_C -# undef PeRl_UINT64_C -/* Prefer the native integer types (int and long) over long long - * (which is not C89) and Win32-specific __int64. */ -# if QUADKIND == QUAD_IS_INT && INTSIZE == 8 -# define PeRl_INT64_C(c) (c) -# define PeRl_UINT64_C(c) CAT2(c,U) -# endif -# if QUADKIND == QUAD_IS_LONG && LONGSIZE == 8 -# define PeRl_INT64_C(c) CAT2(c,L) -# define PeRl_UINT64_C(c) CAT2(c,UL) -# endif -# if QUADKIND == QUAD_IS_LONG_LONG && defined(HAS_LONG_LONG) -# define PeRl_INT64_C(c) CAT2(c,LL) -# define PeRl_UINT64_C(c) CAT2(c,ULL) -# endif -# if QUADKIND == QUAD_IS___INT64 -# define PeRl_INT64_C(c) CAT2(c,I64) -# define PeRl_UINT64_C(c) CAT2(c,UI64) -# endif -# ifndef PeRl_INT64_C -# define PeRl_INT64_C(c) ((I64)(c)) /* last resort */ -# define PeRl_UINT64_C(c) ((U64)(c)) -# endif -/* In OS X the INT64_C/UINT64_C are defined with LL/ULL, which will - * not fly with C89-pedantic gcc, so let's undefine them first so that - * we can redefine them with our native integer preferring versions. */ -# if defined(PERL_DARWIN) && defined(PERL_GCC_PEDANTIC) -# undef INT64_C -# undef UINT64_C -# endif -# ifndef INT64_C -# define INT64_C(c) PeRl_INT64_C(c) -# endif -# ifndef UINT64_C -# define UINT64_C(c) PeRl_UINT64_C(c) -# endif -#endif - #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. @@ -269,22 +224,65 @@ typedef U64TYPE U64; #endif -/* log(2) is pretty close to 0.30103, just in case anyone is grepping for it */ -#define BIT_DIGITS(N) (((N)*146)/485 + 1) /* log2(10) =~ 146/485 */ +/* These C99 typedefs are useful sometimes for, say, loop variables whose + * maximum values are small, but for which speed trumps size. If we have a C99 + * compiler, use that. Otherwise, a plain 'int' should be good enough. + * + * Restrict these to core for now until we are more certain this is a good + * idea. */ +#if defined(PERL_CORE) || defined(PERL_EXT) +# ifdef I_STDINT + typedef int_fast8_t PERL_INT_FAST8_T; + typedef uint_fast8_t PERL_UINT_FAST8_T; + typedef int_fast16_t PERL_INT_FAST16_T; + typedef uint_fast16_t PERL_UINT_FAST16_T; +# else + typedef int PERL_INT_FAST8_T; + typedef unsigned int PERL_UINT_FAST8_T; + typedef int PERL_INT_FAST16_T; + typedef unsigned int PERL_UINT_FAST16_T; +# endif +#endif + +/* log(2) (i.e., log base 10 of 2) is pretty close to 0.30103, just in case + * anyone is grepping for it */ +#define BIT_DIGITS(N) (((N)*146)/485 + 1) /* log10(2) =~ 146/485 */ #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 - * unless under DEBUGGING, where it expands to an assert of its argument, - * followed by a comma (hence the comma operator). If we just used a straight - * assert(), we would get a comma with nothing before it when not DEBUGGING. - * - * We also use empty definition under Coverity since the __ASSERT__ - * checks often check for things that Really Cannot Happen, and Coverity - * detects that and gets all excited. */ +#if defined(PERL_CORE) || defined(PERL_EXT) +# ifndef MIN +# define MIN(a,b) ((a) < (b) ? (a) : (b)) +# endif +# ifndef MAX +# define MAX(a,b) ((a) > (b) ? (a) : (b)) +# endif +#endif + +/* Returns a boolean as to whether the input unsigned number is a power of 2 + * (2**0, 2**1, etc). In other words if it has just a single bit set. + * If not, subtracting 1 would leave the uppermost bit set, so the & would + * yield non-zero */ +#if defined(PERL_CORE) || defined(PERL_EXT) +# define isPOWER_OF_2(n) ((n) && ((n) & ((n)-1)) == 0) +#endif + +/* +=for apidoc Am|void|__ASSERT_|bool expr + +This is a helper macro to avoid preprocessor issues, replaced by nothing +unless under DEBUGGING, where it expands to an assert of its argument, +followed by a comma (hence the comma operator). If we just used a straight +assert(), we would get a comma with nothing before it when not DEBUGGING. + +=cut + +We also use empty definition under Coverity since the __ASSERT__ +checks often check for things that Really Cannot Happen, and Coverity +detects that and gets all excited. */ #if defined(DEBUGGING) && !defined(__COVERITY__) # define __ASSERT_(statement) assert(statement), @@ -293,76 +291,81 @@ typedef U64TYPE U64; #endif /* -=head1 SV-Body Allocation +=head1 SV Manipulation Functions -=for apidoc Ama|SV*|newSVpvs|const char* s -Like C, but takes a literal C-terminated string instead of a +=for apidoc Ama|SV*|newSVpvs|"literal string" s +Like C, but takes a literal string instead of a string/length pair. -=for apidoc Ama|SV*|newSVpvs_flags|const char* s|U32 flags -Like C, but takes a literal C-terminated string instead of +=for apidoc Ama|SV*|newSVpvs_flags|"literal string" s|U32 flags +Like C, but takes a literal string instead of a string/length pair. -=for apidoc Ama|SV*|newSVpvs_share|const char* s -Like C, but takes a literal C-terminated string instead of +=for apidoc Ama|SV*|newSVpvs_share|"literal string" s +Like C, but takes a literal string instead of a string/length pair and omits the hash parameter. -=for apidoc Am|void|sv_catpvs_flags|SV* sv|const char* s|I32 flags -Like C, but takes a literal C-terminated string instead +=for apidoc Am|void|sv_catpvs_flags|SV* sv|"literal string" s|I32 flags +Like C, but takes a literal string instead of a string/length pair. -=for apidoc Am|void|sv_catpvs_nomg|SV* sv|const char* s -Like C, but takes a literal string instead of a -string/length pair. +=for apidoc Am|void|sv_catpvs_nomg|SV* sv|"literal string" s +Like C, but takes a literal string instead of +a string/length pair. -=for apidoc Am|void|sv_catpvs|SV* sv|const char* s -Like C, but takes a literal string instead of a string/length pair. +=for apidoc Am|void|sv_catpvs|SV* sv|"literal string" s +Like C, but takes a literal string instead of a +string/length pair. -=for apidoc Am|void|sv_catpvs_mg|SV* sv|const char* s +=for apidoc Am|void|sv_catpvs_mg|SV* sv|"literal string" s Like C, but takes a literal string instead of a string/length pair. -=for apidoc Am|void|sv_setpvs|SV* sv|const char* s -Like C, but takes a literal string instead of a string/length pair. +=for apidoc Am|void|sv_setpvs|SV* sv|"literal string" s +Like C, but takes a literal string instead of a +string/length pair. -=for apidoc Am|void|sv_setpvs_mg|SV* sv|const char* s +=for apidoc Am|void|sv_setpvs_mg|SV* sv|"literal string" s Like C, but takes a literal string instead of a string/length pair. -=for apidoc Am|SV *|sv_setref_pvs|const char* s -Like C, but takes a literal string instead of a -string/length pair. +=for apidoc Am|SV *|sv_setref_pvs|SV *const rv|const char *const classname|"literal string" s +Like C, but takes a literal string instead of +a string/length pair. =head1 Memory Management -=for apidoc Ama|char*|savepvs|const char* s -Like C, but takes a literal C-terminated string instead of a +=for apidoc Ama|char*|savepvs|"literal string" s +Like C, but takes a literal string instead of a string/length pair. -=for apidoc Ama|char*|savesharedpvs|const char* s +=for apidoc Ama|char*|savesharedpvs|"literal string" s A version of C which allocates the duplicate string in memory which is shared between threads. =head1 GV Functions -=for apidoc Am|HV*|gv_stashpvs|const char* name|I32 create -Like C, but takes a literal string instead of a string/length pair. +=for apidoc Am|HV*|gv_stashpvs|"literal string" name|I32 create +Like C, but takes a literal string instead of a +string/length pair. =head1 Hash Manipulation Functions -=for apidoc Am|SV**|hv_fetchs|HV* tb|const char* key|I32 lval -Like C, but takes a literal string instead of a string/length pair. +=for apidoc Am|SV**|hv_fetchs|HV* tb|"literal string" key|I32 lval +Like C, but takes a literal string instead of a +string/length pair. -=for apidoc Am|SV**|hv_stores|HV* tb|const char* key|NULLOK SV* val -Like C, but takes a literal string instead of a string/length pair +=for apidoc Am|SV**|hv_stores|HV* tb|"literal string" key|SV* val +Like C, but takes a literal string instead of a +string/length pair and omits the hash parameter. =head1 Lexer interface -=for apidoc Amx|void|lex_stuff_pvs|const char *pv|U32 flags +=for apidoc Amx|void|lex_stuff_pvs|"literal string" pv|U32 flags -Like L, but takes a literal string instead of a -string/length pair. +Like L, but takes a literal string instead of +a string/length pair. =cut */ @@ -403,14 +406,7 @@ string/length pair. Perl_gv_fetchpvn_flags(aTHX_ namebeg, len, add, sv_type) #define sv_catxmlpvs(dsv, str, utf8) \ Perl_sv_catxmlpvn(aTHX_ dsv, STR_WITH_LEN(str), utf8) -#define hv_fetchs(hv,key,lval) \ - ((SV **)Perl_hv_common(aTHX_ (hv), NULL, STR_WITH_LEN(key), 0, \ - (lval) ? (HV_FETCH_JUST_SV | HV_FETCH_LVALUE) \ - : HV_FETCH_JUST_SV, NULL, 0)) -#define hv_stores(hv,key,val) \ - ((SV **)Perl_hv_common(aTHX_ (hv), NULL, STR_WITH_LEN(key), 0, \ - (HV_FETCH_ISSTORE|HV_FETCH_JUST_SV), (val), 0)) #define lex_stuff_pvs(pv,flags) Perl_lex_stuff_pvn(aTHX_ STR_WITH_LEN(pv), flags) @@ -421,37 +417,38 @@ string/length pair. =head1 Miscellaneous Functions =for apidoc Am|bool|strNE|char* s1|char* s2 -Test two strings to see if they are different. Returns true or -false. +Test two C-terminated strings to see if they are different. Returns true +or false. =for apidoc Am|bool|strEQ|char* s1|char* s2 -Test two strings to see if they are equal. Returns true or false. +Test two C-terminated strings to see if they are equal. Returns true or +false. =for apidoc Am|bool|strLT|char* s1|char* s2 -Test two strings to see if the first, C, is less than the second, -C. Returns true or false. +Test two C-terminated strings to see if the first, C, is less than the +second, C. Returns true or false. =for apidoc Am|bool|strLE|char* s1|char* s2 -Test two strings to see if the first, C, is less than or equal to the -second, C. Returns true or false. +Test two C-terminated strings to see if the first, C, is less than or +equal to the second, C. Returns true or false. =for apidoc Am|bool|strGT|char* s1|char* s2 -Test two strings to see if the first, C, is greater than the second, -C. Returns true or false. +Test two C-terminated strings to see if the first, C, is greater than +the second, C. Returns true or false. =for apidoc Am|bool|strGE|char* s1|char* s2 -Test two strings to see if the first, C, is greater than or equal to -the second, C. Returns true or false. +Test two C-terminated strings to see if the first, C, is greater than +or equal to the second, C. Returns true or false. =for apidoc Am|bool|strnNE|char* s1|char* s2|STRLEN len -Test two strings to see if they are different. The C parameter -indicates the number of bytes to compare. Returns true or false. (A +Test two C-terminated strings to see if they are different. The C +parameter indicates the number of bytes to compare. Returns true or false. (A wrapper for C). =for apidoc Am|bool|strnEQ|char* s1|char* s2|STRLEN len -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). +Test two C-terminated 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 @@ -464,28 +461,95 @@ are not equal. The C parameter indicates the number of bytes to compare. Returns zero if non-equal, or non-zero if equal. =cut + +New macros should use the following conventions for their names (which are +based on the underlying C library functions): + + (mem | str n? ) (EQ | NE | LT | GT | GE | (( BEGIN | END ) P? )) l? s? + + Each has two main parameters, string-like operands that are compared + against each other, as specified by the macro name. Some macros may + additionally have one or potentially even two length parameters. If a length + parameter applies to both string parameters, it will be positioned third; + otherwise any length parameter immediately follows the string parameter it + applies to. + + If the prefix to the name is 'str', the string parameter is a pointer to a C + language string. Such a string does not contain embedded NUL bytes; its + length may be unknown, but can be calculated by C, since it is + terminated by a NUL, which isn't included in its length. + + The optional 'n' following 'str' means that that there is a third parameter, + giving the maximum number of bytes to look at in each string. Even if both + strings are longer than the length parameter, those extra bytes will be + unexamined. + + The 's' suffix means that the 2nd byte string parameter is a literal C + double-quoted string. Its length will automatically be calculated by the + macro, so no length parameter will ever be needed for it. + + If the prefix is 'mem', the string parameters don't have to be C strings; + they may contain embedded NUL bytes, do not necessarily have a terminating + NUL, and their lengths can be known only through other means, which in + practice are additional parameter(s) passed to the function. All 'mem' + functions have at least one length parameter. Barring any 'l' or 's' suffix, + there is a single length parameter, in position 3, which applies to both + string parameters. The 's' suffix means, as described above, that the 2nd + string is a literal double-quoted C string (hence its length is calculated by + the macro, and the length parameter to the function applies just to the first + string parameter, and hence is positioned just after it). An 'l' suffix + means that the 2nd string parameter has its own length parameter, and the + signature will look like memFOOl(s1, l1, s2, l2). + + BEGIN (and END) are for testing if the 2nd string is an initial (or final) + substring of the 1st string. 'P' if present indicates that the substring + must be a "proper" one in tha mathematical sense that the first one must be + strictly larger than the 2nd. + */ -#define strNE(s1,s2) (strcmp(s1,s2)) -#define strEQ(s1,s2) (!strcmp(s1,s2)) + +#define strNE(s1,s2) (strcmp(s1,s2) != 0) +#define strEQ(s1,s2) (strcmp(s1,s2) == 0) #define strLT(s1,s2) (strcmp(s1,s2) < 0) #define strLE(s1,s2) (strcmp(s1,s2) <= 0) #define strGT(s1,s2) (strcmp(s1,s2) > 0) #define strGE(s1,s2) (strcmp(s1,s2) >= 0) -#define strnNE(s1,s2,l) (strncmp(s1,s2,l)) -#define strnEQ(s1,s2,l) (!strncmp(s1,s2,l)) -#ifdef HAS_MEMCMP -# define memNE(s1,s2,l) (memcmp(s1,s2,l)) -# define memEQ(s1,s2,l) (!memcmp(s1,s2,l)) -#else -# define memNE(s1,s2,l) (bcmp(s1,s2,l)) -# define memEQ(s1,s2,l) (!bcmp(s1,s2,l)) -#endif +#define strnNE(s1,s2,l) (strncmp(s1,s2,l) != 0) +#define strnEQ(s1,s2,l) (strncmp(s1,s2,l) == 0) + +#define memEQ(s1,s2,l) (memcmp(((const void *) (s1)), ((const void *) (s2)), l) == 0) +#define memNE(s1,s2,l) (! memEQ(s1,s2,l)) +/* memEQ and memNE where second comparand is a string constant */ #define memEQs(s1, l, s2) \ - (sizeof(s2)-1 == l && memEQ(s1, ("" s2 ""), (sizeof(s2)-1))) -#define memNEs(s1, l, s2) !memEQs(s1, l, s2) + (((sizeof(s2)-1) == (l)) && memEQ((s1), ("" s2 ""), (sizeof(s2)-1))) +#define memNEs(s1, l, s2) (! memEQs(s1, l, s2)) + +/* Keep these private until we decide it was a good idea */ +#if defined(PERL_CORE) || defined(PERL_EXT) || defined(PERL_EXT_POSIX) + +#define strBEGINs(s1,s2) (strncmp(s1,"" s2 "", sizeof(s2)-1) == 0) + +#define memBEGINs(s1, l, s2) \ + ( (Ptrdiff_t) (l) >= (Ptrdiff_t) sizeof(s2) - 1 \ + && memEQ(s1, "" s2 "", sizeof(s2)-1)) +#define memBEGINPs(s1, l, s2) \ + ( (Ptrdiff_t) (l) > (Ptrdiff_t) sizeof(s2) - 1 \ + && memEQ(s1, "" s2 "", sizeof(s2)-1)) +#define memENDs(s1, l, s2) \ + ( (Ptrdiff_t) (l) >= (Ptrdiff_t) sizeof(s2) - 1 \ + && memEQ(s1 + (l) - (sizeof(s2) - 1), "" s2 "", sizeof(s2)-1)) +#define memENDPs(s1, l, s2) \ + ( (Ptrdiff_t) (l) > (Ptrdiff_t) sizeof(s2) \ + && memEQ(s1 + (l) - (sizeof(s2) - 1), "" s2 "", sizeof(s2)-1)) +#endif /* End of making macros private */ + +#define memLT(s1,s2,l) (memcmp(s1,s2,l) < 0) +#define memLE(s1,s2,l) (memcmp(s1,s2,l) <= 0) +#define memGT(s1,s2,l) (memcmp(s1,s2,l) > 0) +#define memGE(s1,s2,l) (memcmp(s1,s2,l) >= 0) /* * Character classes. @@ -517,70 +581,121 @@ each class. (Not all macros have all variants; each item below lists the ones valid for it.) None are affected by C, and only the ones with C in the name are affected by the current locale. -The base function, e.g., C, takes an octet (either a C or a -C) as input and returns a boolean as to whether or not the character -represented by that octet is (or on non-ASCII platforms, corresponds to) an +The base function, e.g., C, takes any signed or unsigned value, +treating it as a code point, and returns a boolean as to whether or not the +character represented by it is (or on non-ASCII platforms, corresponds to) an ASCII character in the named class based on platform, Unicode, and Perl rules. If the input is a number that doesn't fit in an octet, FALSE is returned. -Variant C (e.g., C) is identical to the base function -with no suffix C<"_A">. +Variant C_A> (e.g., C) is identical to the base function +with no suffix C<"_A">. This variant is used to emphasize by its name that +only ASCII-range characters can return TRUE. -Variant C imposes the Latin-1 (or EBCDIC equivlalent) character set +Variant C_L1> imposes the Latin-1 (or EBCDIC equivalent) character set onto the platform. That is, the code points that are ASCII are unaffected, since ASCII is a subset of Latin-1. But the non-ASCII code points are treated as if they are Latin-1 characters. For example, C will return true when called with the code point 0xDF, which is a word character in both ASCII and EBCDIC (though it represents different characters in each). +If the input is a number that doesn't fit in an octet, FALSE is returned. +(Perl's documentation uses a colloquial definition of Latin-1, to include all +code points below 256.) -Variant C is like the C variant, but accepts any UV code -point as input. If the code point is larger than 255, Unicode rules are used -to determine if it is in the character class. For example, -C returns TRUE, since 0x100 is LATIN CAPITAL LETTER A +Variant C_uvchr> is exactly like the C_L1> variant, for +inputs below 256, but if the code point is larger than 255, Unicode rules are +used to determine if it is in the character class. For example, +C returns TRUE, since 0x100 is LATIN CAPITAL LETTER A WITH MACRON in Unicode, and is a word character. -Variant C is like C, but the input is a pointer to a -(known to be well-formed) UTF-8 encoded string (C or C). The -classification of just the first (possibly multi-byte) character in the string -is tested. - -Variant C is like the C and C variants, but the -result is based on the current locale, which is what C in the name stands -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 +Variant C_utf8_safe> is like C_uvchr>, but is used for UTF-8 +encoded strings. Each call classifies the first character of the string. This +variant takes two parameters. The first, C

, is a +pointer to the first byte of the character to be classified. (Recall that it +may take more than one byte to represent a character in UTF-8 strings.) The +second parameter, C, points to anywhere in the string beyond the first +character, up to one byte past the end of the entire string. The suffix +C<_safe> in the function's name indicates that it will not attempt to read +beyond S>, provided that the constraint S e>> is true (this +is asserted for in C<-DDEBUGGING> builds). If the UTF-8 for the input +character is malformed in some way, the program may croak, or the function may +return FALSE, at the discretion of the implementation, and subject to change in +future releases. + +Variant C_utf8> is like C_utf8_safe>, but takes just a single +parameter, C

, which has the same meaning as the corresponding parameter does +in C_utf8_safe>. The function therefore can't check if it is reading +beyond the end of the string. Starting in Perl v5.32, it will take a second +parameter, becoming a synonym for C_utf8_safe>. At that time every +program that uses it will have to be changed to successfully compile. In the +meantime, the first runtime call to C_utf8> from each call point in the +program will raise a deprecation warning, enabled by default. You can convert +your program now to use C_utf8_safe>, and avoid the warnings, and get an +extra measure of protection, or you can wait until v5.32, when you'll be forced +to add the C parameter. + +Variant C_LC> is like the C_A> and C_L1> variants, +but the result is based on the current locale, which is what C in the name +stands 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. 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 -returns the hard-coded, not-affected-by-locale, Unicode results for larger ones. - -Variant C is like C, but the input is a pointer -to a (known to be well-formed) UTF-8 encoded string (C or C). The -classification of just the first (possibly multi-byte) character in the string -is tested. - -=for apidoc Am|bool|isALPHA|char ch -Returns a boolean indicating whether the specified character is an -alphabetic character, analogous to C. +Variant C_LC_uvchr> acts exactly like C_LC> for inputs less +than 256, but for larger ones it returns the Unicode classification of the code +point. + +Variant C_LC_utf8_safe> is like C_LC_uvchr>, but is used for UTF-8 +encoded strings. Each call classifies the first character of the string. This +variant takes two parameters. The first, C

, is a pointer to the first byte +of the character to be classified. (Recall that it may take more than one byte +to represent a character in UTF-8 strings.) The second parameter, C, +points to anywhere in the string beyond the first character, up to one byte +past the end of the entire string. The suffix C<_safe> in the function's name +indicates that it will not attempt to read beyond S>, provided that +the constraint S e>> is true (this is asserted for in C<-DDEBUGGING> +builds). If the UTF-8 for the input character is malformed in some way, the +program may croak, or the function may return FALSE, at the discretion of the +implementation, and subject to change in future releases. + +Variant C_LC_utf8> is like C_LC_utf8_safe>, but takes just a single +parameter, C

, which has the same meaning as the corresponding parameter does +in C_LC_utf8_safe>. The function therefore can't check if it is reading +beyond the end of the string. Starting in Perl v5.32, it will take a second +parameter, becoming a synonym for C_LC_utf8_safe>. At that time every +program that uses it will have to be changed to successfully compile. In the +meantime, the first runtime call to C_LC_utf8> from each call point in +the program will raise a deprecation warning, enabled by default. You can +convert your program now to use C_LC_utf8_safe>, and avoid the warnings, +and get an extra measure of protection, or you can wait until v5.32, when +you'll be forced to add the C parameter. + +=for apidoc Am|bool|isALPHA|int ch +Returns a boolean indicating whether the specified input is one of C<[A-Za-z]>, +analogous to C. See the L for an explanation of variants -C, C, C, C, C, -C, and C. +C, C, C, C, +C, C, and C. -=for apidoc Am|bool|isALPHANUMERIC|char ch -Returns a boolean indicating whether the specified character is a either an -alphabetic character or decimal digit, analogous to C. +=for apidoc Am|bool|isALPHANUMERIC|int ch +Returns a boolean indicating whether the specified character is one of +C<[A-Za-z0-9]>, analogous to C. See the L for an explanation of variants -C, C, C, -C, C, C, and -C. +C, C, C, +C, C, C, +and C. -=for apidoc Am|bool|isASCII|char ch +A (discouraged from use) synonym is C (where the C suffix means +this corresponds to the C language alphanumeric definition). Also +there are the variants +C, C +C, and C. + +=for apidoc Am|bool|isASCII|int ch Returns a boolean indicating whether the specified character is one of the 128 characters in the ASCII character set, analogous to C. On non-ASCII platforms, it returns TRUE iff this @@ -588,36 +703,36 @@ character corresponds to an ASCII character. Variants C and C are identical to C. See the L for an explanation of variants -C, C, C, C, and -C. Note, however, that some platforms do not have the C +C, C, C, C, and +C. Note, however, that some platforms do not have the C library routine C. In these cases, the variants whose names contain C are the same as the corresponding ones without. Also note, that because all ASCII characters are UTF-8 invariant (meaning they have the exact same representation (always a single byte) whether encoded in UTF-8 or not), C will give the correct results when called with any -byte in any string encoded or not in UTF-8. And similarly C will -work properly on any string encoded or not in UTF-8. +byte in any string encoded or not in UTF-8. And similarly C +will work properly on any string encoded or not in UTF-8. =for apidoc Am|bool|isBLANK|char ch Returns a boolean indicating whether the specified character is a character considered to be a blank, analogous to C. See the L for an explanation of variants -C, C, C, C, C, -C, and C. Note, however, that some -platforms do not have the C library routine C. In these cases, the -variants whose names contain C are the same as the corresponding ones -without. +C, C, C, C, +C, C, and C. Note, +however, that some platforms do not have the C library routine +C. In these cases, the variants whose names contain C are +the same as the corresponding ones without. =for apidoc Am|bool|isCNTRL|char ch Returns a boolean indicating whether the specified character is a control character, analogous to C. See the L for an explanation of variants -C, C, C, C, C, -C, and C -On EBCDIC platforms, you almost always want to use the C variant. +C, C, C, C, +C, C, and C On EBCDIC +platforms, you almost always want to use the C variant. =for apidoc Am|bool|isDIGIT|char ch Returns a boolean indicating whether the specified character is a @@ -625,24 +740,23 @@ digit, analogous to C. Variants C and C are identical to C. See the L for an explanation of variants -C, C, C, C, and -C. +C, C, C, C, and +C. =for apidoc Am|bool|isGRAPH|char ch Returns a boolean indicating whether the specified character is a graphic character, analogous to C. See the L for an explanation of -variants -C, C, C, C, C, -C, and C. +variants C, C, C, C, +C, C, and C. =for apidoc Am|bool|isLOWER|char ch Returns a boolean indicating whether the specified character is a lowercase character, analogous to C. See the L for an explanation of variants -C, C, C, C, C, -C, and C. +C, C, C, C, +C, C, and C. =for apidoc Am|bool|isOCTAL|char ch Returns a boolean indicating whether the specified character is an @@ -657,9 +771,8 @@ Note that the definition of what is punctuation isn't as straightforward as one might desire. See L for details. See the L for an explanation of -variants -C, C, C, C, C, -C, and C. +variants C, C, C, C, +C, C, and C. =for apidoc Am|bool|isSPACE|char ch Returns a boolean indicating whether the specified character is a @@ -672,8 +785,8 @@ in the non-locale variants, was that C did not match a vertical tab. (See L for a macro that matches a vertical tab in all releases.) See the L for an explanation of variants -C, C, C, C, C, -C, and C. +C, C, C, C, +C, C, and C. =for apidoc Am|bool|isPSXSPC|char ch (short for Posix Space) @@ -686,24 +799,23 @@ C forms don't match a Vertical Tab, and the C forms do. Otherwise they are identical. Thus this macro is analogous to what C matches in a regular expression. See the L for an explanation of -variants C, C, C, C, -C, C, and C. +variants C, C, C, C, +C, C, and C. =for apidoc Am|bool|isUPPER|char ch Returns a boolean indicating whether the specified character is an uppercase character, analogous to C. See the L for an explanation of -variants -C, C, C, C, C, -C, and C. +variants C, C, C, C, +C, C, and C. =for apidoc Am|bool|isPRINT|char ch Returns a boolean indicating whether the specified character is a printable character, analogous to C. See the L for an explanation of variants -C, C, C, C, C, -C, and C. +C, C, C, C, +C, C, and C. =for apidoc Am|bool|isWORDCHAR|char ch Returns a boolean indicating whether the specified character is a character @@ -715,10 +827,10 @@ C is a synonym provided for backward compatibility, even though a word character includes more than the standard C language meaning of alphanumeric. See the L for an explanation of -variants -C, C, C, and C. -C, C, and C are also as -described there, but additionally include the platform's native underscore. +variants 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 @@ -726,8 +838,8 @@ digit. In the ASCII range these are C<[0-9A-Fa-f]>. Variants C and C are identical to C. See the L for an explanation of variants -C, C, C, C, and -C. +C, C, C, C, +and C. =for apidoc Am|bool|isIDFIRST|char ch Returns a boolean indicating whether the specified character can be the first @@ -736,8 +848,8 @@ the official Unicode property C. The difference is that this returns true only if the input character also matches L. See the L for an explanation of variants -C, C, C, C, -C, C, and C. +C, C, C, C, +C, C, and C. =for apidoc Am|bool|isIDCONT|char ch Returns a boolean indicating whether the specified character can be the @@ -746,9 +858,9 @@ not quite the same as the official Unicode property C. The difference is that this returns true only if the input character also matches L. See the L for an -explanation of variants C, C, C, -C, C, C, and -C. +explanation of variants C, C, C, +C, C, C, and +C. =head1 Miscellaneous Functions @@ -757,57 +869,109 @@ Returns the value of an ASCII-range hex digit and advances the string pointer. Behaviour is only well defined when isXDIGIT(*str) is true. =head1 Character case changing - -=for apidoc Am|U8|toUPPER|U8 ch +Perl uses "full" Unicode case mappings. This means that converting a single +character to another case may result in a sequence of more than one character. +For example, the uppercase of C> (LATIN SMALL LETTER SHARP S) is the two +character sequence C. This presents some complications The lowercase of +all characters in the range 0..255 is a single character, and thus +C> is furnished. But, C can't exist, as it couldn't +return a valid result for all legal inputs. Instead C> has +an API that does allow every possible legal result to be returned.) Likewise +no other function that is crippled by not being able to give the correct +results for the full range of possible inputs has been implemented here. + +=for apidoc Am|U8|toUPPER|int ch Converts the specified character to uppercase. If the input is anything but an ASCII lowercase character, that input character itself is returned. Variant C is equivalent. -=for apidoc Am|UV|toUPPER_uni|UV cp|U8* s|STRLEN* lenp -Converts the Unicode code point C to its uppercase version, and -stores that in UTF-8 in C, and its length in bytes in C. Note +=for apidoc Am|UV|toUPPER_uvchr|UV cp|U8* s|STRLEN* lenp +Converts the code point C to its uppercase version, and +stores that in UTF-8 in C, and its length in bytes in C. The code +point is interpreted as native if less than 256; otherwise as Unicode. Note that the buffer pointed to by C needs to be at least C bytes since the uppercase version may be longer than the original character. The first code point of the uppercased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more.) -=for apidoc Am|UV|toUPPER_utf8|U8* p|U8* s|STRLEN* lenp -Converts the UTF-8 encoded character at C

to its uppercase version, and +=for apidoc Am|UV|toUPPER_utf8_safe|U8* p|U8* e|U8* s|STRLEN* lenp +Converts the first UTF-8 encoded character in the sequence starting at C

and +extending no further than S> to its uppercase version, and stores that in UTF-8 in C, and its length in bytes in C. Note that the buffer pointed to by C needs to be at least C bytes since the uppercase version may be longer than the original character. The first code point of the uppercased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more). + +The suffix C<_safe> in the function's name indicates that it will not attempt +to read beyond S>, provided that the constraint S e>> is +true (this is asserted for in C<-DDEBUGGING> builds). If the UTF-8 for the +input character is malformed in some way, the program may croak, or the +function may return the REPLACEMENT CHARACTER, at the discretion of the +implementation, and subject to change in future releases. -The input character at C

is assumed to be well-formed. +=for apidoc Am|UV|toUPPER_utf8|U8* p|U8* s|STRLEN* lenp +This is like C>, but doesn't have the C +parameter The function therefore can't check if it is reading +beyond the end of the string. Starting in Perl v5.32, it will take the C +parameter, becoming a synonym for C. At that time every +program that uses it will have to be changed to successfully compile. In the +meantime, the first runtime call to C from each call point in the +program will raise a deprecation warning, enabled by default. You can convert +your program now to use C, and avoid the warnings, and get an +extra measure of protection, or you can wait until v5.32, when you'll be forced +to add the C parameter. =for apidoc Am|U8|toFOLD|U8 ch Converts the specified character to foldcase. If the input is anything but an ASCII uppercase character, that input character itself is returned. Variant C is equivalent. (There is no equivalent C for the full -Latin1 range, as the full generality of L is needed there.) +Latin1 range, as the full generality of L is needed there.) -=for apidoc Am|UV|toFOLD_uni|UV cp|U8* s|STRLEN* lenp -Converts the Unicode code point C to its foldcase version, and -stores that in UTF-8 in C, and its length in bytes in C. Note +=for apidoc Am|UV|toFOLD_uvchr|UV cp|U8* s|STRLEN* lenp +Converts the code point C to its foldcase version, and +stores that in UTF-8 in C, and its length in bytes in C. The code +point is interpreted as native if less than 256; otherwise as Unicode. Note that the buffer pointed to by C needs to be at least C bytes since the foldcase version may be longer than the original character. The first code point of the foldcased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more). -=for apidoc Am|UV|toFOLD_utf8|U8* p|U8* s|STRLEN* lenp -Converts the UTF-8 encoded character at C

to its foldcase version, and +=for apidoc Am|UV|toFOLD_utf8_safe|U8* p|U8* e|U8* s|STRLEN* lenp +Converts the first UTF-8 encoded character in the sequence starting at C

and +extending no further than S> to its foldcase version, and stores that in UTF-8 in C, and its length in bytes in C. Note that the buffer pointed to by C needs to be at least C bytes since the foldcase version may be longer than the original character. The first code point of the foldcased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more). + +The suffix C<_safe> in the function's name indicates that it will not attempt +to read beyond S>, provided that the constraint S e>> is +true (this is asserted for in C<-DDEBUGGING> builds). If the UTF-8 for the +input character is malformed in some way, the program may croak, or the +function may return the REPLACEMENT CHARACTER, at the discretion of the +implementation, and subject to change in future releases. -The input character at C

is assumed to be well-formed. +=for apidoc Am|UV|toFOLD_utf8|U8* p|U8* s|STRLEN* lenp +This is like C>, but doesn't have the C +parameter The function therefore can't check if it is reading +beyond the end of the string. Starting in Perl v5.32, it will take the C +parameter, becoming a synonym for C. At that time every +program that uses it will have to be changed to successfully compile. In the +meantime, the first runtime call to C from each call point in the +program will raise a deprecation warning, enabled by default. You can convert +your program now to use C, and avoid the warnings, and get an +extra measure of protection, or you can wait until v5.32, when you'll be forced +to add the C parameter. =for apidoc Am|U8|toLOWER|U8 ch Converts the specified character to lowercase. If the input is anything but an @@ -822,68 +986,110 @@ undefined if the input doesn't fit in a byte. Converts the specified character to lowercase using the current locale's rules, if possible; otherwise returns the input character itself. -=for apidoc Am|UV|toLOWER_uni|UV cp|U8* s|STRLEN* lenp -Converts the Unicode code point C to its lowercase version, and -stores that in UTF-8 in C, and its length in bytes in C. Note +=for apidoc Am|UV|toLOWER_uvchr|UV cp|U8* s|STRLEN* lenp +Converts the code point C to its lowercase version, and +stores that in UTF-8 in C, and its length in bytes in C. The code +point is interpreted as native if less than 256; otherwise as Unicode. Note that the buffer pointed to by C needs to be at least C bytes since the lowercase version may be longer than the original character. The first code point of the lowercased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more). -=for apidoc Am|UV|toLOWER_utf8|U8* p|U8* s|STRLEN* lenp -Converts the UTF-8 encoded character at C

to its lowercase version, and + +=for apidoc Am|UV|toLOWER_utf8_safe|U8* p|U8* e|U8* s|STRLEN* lenp +Converts the first UTF-8 encoded character in the sequence starting at C

and +extending no further than S> to its lowercase version, and stores that in UTF-8 in C, and its length in bytes in C. Note that the buffer pointed to by C needs to be at least C bytes since the lowercase version may be longer than the original character. The first code point of the lowercased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more). + +The suffix C<_safe> in the function's name indicates that it will not attempt +to read beyond S>, provided that the constraint S e>> is +true (this is asserted for in C<-DDEBUGGING> builds). If the UTF-8 for the +input character is malformed in some way, the program may croak, or the +function may return the REPLACEMENT CHARACTER, at the discretion of the +implementation, and subject to change in future releases. -The input character at C

is assumed to be well-formed. +=for apidoc Am|UV|toLOWER_utf8|U8* p|U8* s|STRLEN* lenp +This is like C>, but doesn't have the C +parameter The function therefore can't check if it is reading +beyond the end of the string. Starting in Perl v5.32, it will take the C +parameter, becoming a synonym for C. At that time every +program that uses it will have to be changed to successfully compile. In the +meantime, the first runtime call to C from each call point in the +program will raise a deprecation warning, enabled by default. You can convert +your program now to use C, and avoid the warnings, and get an +extra measure of protection, or you can wait until v5.32, when you'll be forced +to add the C parameter. =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 C is equivalent. (There is no C for the full Latin1 -range, as the full generality of L is needed there. Titlecase is +range, as the full generality of L is needed there. Titlecase is not a concept used in locale handling, so there is no functionality for that.) -=for apidoc Am|UV|toTITLE_uni|UV cp|U8* s|STRLEN* lenp -Converts the Unicode code point C to its titlecase version, and -stores that in UTF-8 in C, and its length in bytes in C. Note +=for apidoc Am|UV|toTITLE_uvchr|UV cp|U8* s|STRLEN* lenp +Converts the code point C to its titlecase version, and +stores that in UTF-8 in C, and its length in bytes in C. The code +point is interpreted as native if less than 256; otherwise as Unicode. Note that the buffer pointed to by C needs to be at least C bytes since the titlecase version may be longer than the original character. The first code point of the titlecased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more). -=for apidoc Am|UV|toTITLE_utf8|U8* p|U8* s|STRLEN* lenp -Converts the UTF-8 encoded character at C

to its titlecase version, and +=for apidoc Am|UV|toTITLE_utf8_safe|U8* p|U8* e|U8* s|STRLEN* lenp +Converts the first UTF-8 encoded character in the sequence starting at C

and +extending no further than S> to its titlecase version, and stores that in UTF-8 in C, and its length in bytes in C. Note that the buffer pointed to by C needs to be at least C bytes since the titlecase version may be longer than the original character. The first code point of the titlecased version is returned -(but note, as explained just above, that there may be more.) +(but note, as explained at L, that there may be more). + +The suffix C<_safe> in the function's name indicates that it will not attempt +to read beyond S>, provided that the constraint S e>> is +true (this is asserted for in C<-DDEBUGGING> builds). If the UTF-8 for the +input character is malformed in some way, the program may croak, or the +function may return the REPLACEMENT CHARACTER, at the discretion of the +implementation, and subject to change in future releases. -The input character at C

is assumed to be well-formed. +=for apidoc Am|UV|toTITLE_utf8|U8* p|U8* s|STRLEN* lenp +This is like C>, but doesn't have the C +parameter The function therefore can't check if it is reading +beyond the end of the string. Starting in Perl v5.32, it will take the C +parameter, becoming a synonym for C. At that time every +program that uses it will have to be changed to successfully compile. In the +meantime, the first runtime call to C from each call point in the +program will raise a deprecation warning, enabled by default. You can convert +your program now to use C, and avoid the warnings, and get an +extra measure of protection, or you can wait until v5.32, when you'll be forced +to add the C parameter. =cut -XXX Still undocumented isVERTWS_uni and _utf8; it's unclear what their names -really should be. Also toUPPER_LC and toFOLD_LC, which are subject to change. +XXX Still undocumented isVERTWS_uvchr and _utf8; it's unclear what their names +really should be. Also toUPPER_LC and toFOLD_LC, which are subject to change, +and aren't general purpose as they don't work on U+DF, and assert against that. Note that these macros are repeated in Devel::PPPort, so should also be patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc */ -/* 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 WIDEST_UTYPE U64TYPE +/* Specify the widest unsigned type on the platform. */ +#ifdef QUADKIND +# define WIDEST_UTYPE U64 #else # define WIDEST_UTYPE U32 #endif @@ -900,11 +1106,36 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc * of operands. Well, they are, but that is kind of the point. */ #ifndef __COVERITY__ -#define FITS_IN_8_BITS(c) ((sizeof(c) == 1) || !(((WIDEST_UTYPE)(c)) & ~0xFF)) + /* The '| 0' part ensures a compiler error if c is not integer (like e.g., a + * pointer) */ +#define FITS_IN_8_BITS(c) ( (sizeof(c) == 1) \ + || !(((WIDEST_UTYPE)((c) | 0)) & ~0xFF)) #else #define FITS_IN_8_BITS(c) (1) #endif +/* Returns true if c is in the range l..u, where 'l' is non-negative + * Written this way so that after optimization, only one conditional test is + * needed. + * + * This isn't fully general, except for the special cased 'signed char' (which + * should be resolved at compile time): It won't work if 'c' is negative, and + * 'l' is larger than the max for that signed type. Thus if 'c' is a negative + * int, and 'l' is larger than INT_MAX, it will fail. To protect agains this + * happening, there is an assert that will generate a warning if c is larger + * than e.g. INT_MAX if it is an 'unsigned int'. This could be a false + * positive, but khw couldn't figure out a way to make it better. It's good + * enough so far */ +#define inRANGE(c, l, u) (__ASSERT_((l) >= 0) __ASSERT_((u) >= (l)) \ + ((sizeof(c) == 1) \ + ? (((WIDEST_UTYPE) ((((U8) (c))|0) - (l))) <= ((WIDEST_UTYPE) ((u) - (l)))) \ + : (__ASSERT_( (((WIDEST_UTYPE) 1) << (CHARBITS * sizeof(c) - 1) & (c)) \ + /* sign bit of c is 0 */ == 0 \ + || (((~ ((WIDEST_UTYPE) 1) << ((CHARBITS * sizeof(c) - 1) - 1))\ + /* l not larger than largest value in c's signed type */ \ + & ~ ((WIDEST_UTYPE) 0)) & (l)) == 0) \ + ((WIDEST_UTYPE) (((c) - (l)) | 0) <= ((WIDEST_UTYPE) ((u) - (l))))))) + #ifdef EBCDIC # ifndef _ALL_SOURCE /* The native libc isascii() et.al. functions return the wrong results @@ -914,14 +1145,24 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc #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) + * other macros below. + * + * The cast here is used instead of '(c) >= 0', because some compilers emit + * a warning that that test is always true when the parameter is an + * unsigned type. khw supposes that it could be written as + * && ((c) == '\0' || (c) > 0) + * to avoid the message, but the cast will likely avoid extra branches even + * with stupid compilers. + * + * The '| 0' part ensures a compiler error if c is not integer (like e.g., + * a pointer) */ +# define isASCII(c) ((WIDEST_UTYPE)((c) | 0) < 128) #endif -/* 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') +/* Take the eight possible bit patterns of the lower 3 bits and you get the + * lower 3 bits of the 8 octal digits, in both ASCII and EBCDIC, so those bits + * can be ignored. If the rest match '0', we have an octal */ +#define isOCTAL_A(c) (((WIDEST_UTYPE)((c) | 0) & ~7) == '0') #ifdef H_PERL /* If have access to perl.h, lookup in its table */ @@ -934,7 +1175,13 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc * * The first group of these is ordered in what I (khw) estimate to be the * frequency of their use. This gives a slight edge to exiting a loop earlier - * (in reginclass() in regexec.c) */ + * (in reginclass() in regexec.c). Except \v should be last, as it isn't a + * real Posix character class, and some (small) inefficiencies in regular + * expression handling would be introduced by putting it in the middle of those + * that are. Also, cntrl and ascii come after the others as it may be useful + * to group these which have no members that match above Latin1, (or above + * ASCII in the latter case) */ + # define _CC_WORDCHAR 0 /* \w and [:word:] */ # define _CC_DIGIT 1 /* \d and [:digit:] */ # define _CC_ALPHA 2 /* [:alpha:] */ @@ -944,19 +1191,11 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define _CC_PRINT 6 /* [:print:] */ # define _CC_ALPHANUMERIC 7 /* [:alnum:] */ # define _CC_GRAPH 8 /* [:graph:] */ -# define _CC_CASED 9 /* [:lower:] and [:upper:] under /i */ - -#define _FIRST_NON_SWASH_CC 10 -/* The character classes above are implemented with swashes. The second group - * (just below) contains the ones implemented without. These are also sorted - * in rough order of the frequency of their use, except that \v should be last, - * as it isn't a real Posix character class, and some (small) inefficiencies in - * regular expression handling would be introduced by putting it in the middle - * of those that are. Also, cntrl and ascii come after the others as it may be - * useful to group these which have no members that match above Latin1, (or - * above ASCII in the latter case) */ - +# define _CC_CASED 9 /* [:lower:] or [:upper:] under /i */ # define _CC_SPACE 10 /* \s, [:space:] */ +# define _CC_PSXSPC _CC_SPACE /* XXX Temporary, can be removed + when the deprecated isFOO_utf8() + functions are removed */ # define _CC_BLANK 11 /* [:blank:] */ # define _CC_XDIGIT 12 /* [:xdigit:] */ # define _CC_CNTRL 13 /* [:cntrl:] */ @@ -976,6 +1215,9 @@ patched there. The file as of this writing is cpan/Devel-PPPort/parts/inc/misc # define _CC_IS_IN_SOME_FOLD 22 # define _CC_MNEMONIC_CNTRL 23 +# define _CC_IDCONT 24 /* XXX Temporary, can be removed when the deprecated + isFOO_utf8() functions are removed */ + /* This next group is only used on EBCDIC platforms, so theoretically could be * shared with something entirely different that's only on ASCII platforms */ # define _CC_UTF8_START_BYTE_IS_FOR_AT_LEAST_SURROGATE 28 @@ -1016,36 +1258,8 @@ typedef enum { } _char_class_number; #endif -#define POSIX_SWASH_COUNT _FIRST_NON_SWASH_CC #define POSIX_CC_COUNT (_HIGHEST_REGCOMP_DOT_H_SYNC + 1) -#if defined(PERL_IN_UTF8_C) \ - || defined(PERL_IN_REGCOMP_C) \ - || defined(PERL_IN_REGEXEC_C) -# if _CC_WORDCHAR != 0 || _CC_DIGIT != 1 || _CC_ALPHA != 2 || _CC_LOWER != 3 \ - || _CC_UPPER != 4 || _CC_PUNCT != 5 || _CC_PRINT != 6 \ - || _CC_ALPHANUMERIC != 7 || _CC_GRAPH != 8 || _CC_CASED != 9 - #error Need to adjust order of swash_property_names[] -# endif - -/* This is declared static in each of the few files that this is #defined for - * to keep them from being publicly accessible. Hence there is a small amount - * of wasted space */ - -static const char* const swash_property_names[] = { - "XPosixWord", - "XPosixDigit", - "XPosixAlpha", - "XPosixLower", - "XPosixUpper", - "XPosixPunct", - "XPosixPrint", - "XPosixAlnum", - "XPosixGraph", - "Cased" -}; -#endif - START_EXTERN_C # ifdef DOINIT EXTCONST U32 PL_charclass[] = { @@ -1076,17 +1290,28 @@ END_EXTERN_C && ((PL_charclass[(U8) (c)] & _CC_mask_A(classnum)) \ == _CC_mask_A(classnum))) -# define isALPHA_A(c) _generic_isCC_A(c, _CC_ALPHA) +/* On ASCII platforms certain classes form a single range. It's faster to + * special case these. isDIGIT is a single range on all platforms */ +# ifdef EBCDIC +# define isALPHA_A(c) _generic_isCC_A(c, _CC_ALPHA) +# 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 isUPPER_A(c) _generic_isCC_A(c, _CC_UPPER) +# else + /* By folding the upper and lowercase, we can use a single range */ +# define isALPHA_A(c) inRANGE((~('A' ^ 'a') & (c)), 'A', 'Z') +# define isGRAPH_A(c) inRANGE(c, ' ' + 1, 0x7e) +# define isLOWER_A(c) inRANGE(c, 'a', 'z') +# define isPRINT_A(c) inRANGE(c, ' ', 0x7e) +# define isUPPER_A(c) inRANGE(c, 'A', 'Z') +# endif # 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) /* 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 isDIGIT_A(c) inRANGE(c, '0', '9') # 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) /* No non-ASCII xdigits */ @@ -1130,82 +1355,73 @@ END_EXTERN_C /* 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. Most are coded based on first principals. First some ones common - * to both ASCII and EBCDIC */ -# define isDIGIT_A(c) ((c) <= '9' && (c) >= '0') + * to it. Most are coded based on first principles. These are written to + * avoid EBCDIC vs. ASCII #ifdef's as much as possible. */ +# define isDIGIT_A(c) inRANGE(c, '0', '9') # define isBLANK_A(c) ((c) == ' ' || (c) == '\t') -# define isSPACE_A(c) (isBLANK_A(c) \ - || (c) == '\n' \ - || (c) == '\r' \ - || (c) == '\v' \ +# 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 isUPPER_A(c) ((c) <= 'Z' && (c) >= 'A') -# endif - - /* Some more ASCII, non-ASCII common definitions */ + /* On EBCDIC, there are gaps between 'i' and 'j'; 'r' and 's'. Same for + * uppercase. The tests for those aren't necessary on ASCII, but hurt only + * performance (if optimization isn't on), and allow the same code to be + * used for both platform types */ +# define isLOWER_A(c) inRANGE((c), 'a', 'i') \ + || inRANGE((c), 'j', 'r') \ + || inRANGE((c), 's', 'z') +# define isUPPER_A(c) inRANGE((c), 'A', 'I') \ + || inRANGE((c), 'J', 'R') \ + || inRANGE((c), 'S', 'Z') # define isALPHA_A(c) (isUPPER_A(c) || isLOWER_A(c)) # define isALPHANUMERIC_A(c) (isALPHA_A(c) || isDIGIT_A(c)) # define isWORDCHAR_A(c) (isALPHANUMERIC_A(c) || (c) == '_') # define isIDFIRST_A(c) (isALPHA_A(c) || (c) == '_') -# define isXDIGIT_A(c) (isDIGIT_A(c) \ - || ((c) >= 'a' && (c) <= 'f') \ - || ((c) <= 'F' && (c) >= 'A')) +# define isXDIGIT_A(c) ( isDIGIT_A(c) \ + || inRANGE((c), 'a', 'f') \ + || inRANGE((c), 'A', 'F') +# define isPUNCT_A(c) ((c) == '-' || (c) == '!' || (c) == '"' \ + || (c) == '#' || (c) == '$' || (c) == '%' \ + || (c) == '&' || (c) == '\'' || (c) == '(' \ + || (c) == ')' || (c) == '*' || (c) == '+' \ + || (c) == ',' || (c) == '.' || (c) == '/' \ + || (c) == ':' || (c) == ';' || (c) == '<' \ + || (c) == '=' || (c) == '>' || (c) == '?' \ + || (c) == '@' || (c) == '[' || (c) == '\\' \ + || (c) == ']' || (c) == '^' || (c) == '_' \ + || (c) == '`' || (c) == '{' || (c) == '|' \ + || (c) == '}' || (c) == '~') +# define isGRAPH_A(c) (isALPHANUMERIC_A(c) || isPUNCT_A(c)) +# define isPRINT_A(c) (isGRAPH_A(c) || (c) == ' ') # ifdef EBCDIC -# define isPUNCT_A(c) ((c) == '-' || (c) == '!' || (c) == '"' \ - || (c) == '#' || (c) == '$' || (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)) - + /* The below is accurate for the 3 EBCDIC code pages traditionally + * supported by perl. The only difference between them in the controls + * is the position of \n, and that is represented symbolically below */ +# define isCNTRL_A(c) ((c) == '\0' || (c) == '\a' || (c) == '\b' \ + || (c) == '\f' || (c) == '\n' || (c) == '\r' \ + || (c) == '\t' || (c) == '\v' \ + || inRANGE((c), 1, 3) /* SOH, STX, ETX */ \ + || (c) == 7 /* U+7F DEL */ \ + || inRANGE((c), 0x0E, 0x13) /* SO SI DLE \ + DC[1-3] */ \ + || (c) == 0x18 /* U+18 CAN */ \ + || (c) == 0x19 /* U+19 EOM */ \ + || inRANGE((c), 0x1C, 0x1F) /* [FGRU]S */ \ + || (c) == 0x26 /* U+17 ETB */ \ + || (c) == 0x27 /* U+1B ESC */ \ + || (c) == 0x2D /* U+05 ENQ */ \ + || (c) == 0x2E /* U+06 ACK */ \ + || (c) == 0x32 /* U+16 SYN */ \ + || (c) == 0x37 /* U+04 EOT */ \ + || (c) == 0x3C /* U+14 DC4 */ \ + || (c) == 0x3D /* U+15 NAK */ \ + || (c) == 0x3F)/* U+1A SUB */ # 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))) +# else /* isASCII is already defined for ASCII platforms, so can use that to + define isCNTRL */ +# define isCNTRL_A(c) (isASCII(c) && ! isPRINT_A(c)) # endif /* The _L1 macros may be unnecessary for the utilities; I (khw) added them @@ -1224,7 +1440,7 @@ END_EXTERN_C # define isGRAPH_L1(c) (isPRINT_L1(c) && (! isBLANK_L1(c))) # define isLOWER_L1(c) (isLOWER_A(c) \ || (FITS_IN_8_BITS(c) \ - && ((NATIVE_TO_LATIN1((U8) c) >= 0xDF \ + && (( NATIVE_TO_LATIN1((U8) c) >= 0xDF \ && NATIVE_TO_LATIN1((U8) c) != 0xF7) \ || NATIVE_TO_LATIN1((U8) c) == 0xAA \ || NATIVE_TO_LATIN1((U8) c) == 0xBA \ @@ -1234,7 +1450,7 @@ END_EXTERN_C && NATIVE_TO_LATIN1((U8) c) >= 0xA0)) # define isPUNCT_L1(c) (isPUNCT_A(c) \ || (FITS_IN_8_BITS(c) \ - && (NATIVE_TO_LATIN1((U8) c) == 0xA1 \ + && ( NATIVE_TO_LATIN1((U8) c) == 0xA1 \ || NATIVE_TO_LATIN1((U8) c) == 0xA7 \ || NATIVE_TO_LATIN1((U8) c) == 0xAB \ || NATIVE_TO_LATIN1((U8) c) == 0xB6 \ @@ -1243,12 +1459,12 @@ END_EXTERN_C || NATIVE_TO_LATIN1((U8) c) == 0xBF))) # define isSPACE_L1(c) (isSPACE_A(c) \ || (FITS_IN_8_BITS(c) \ - && (NATIVE_TO_LATIN1((U8) c) == 0x85 \ + && ( NATIVE_TO_LATIN1((U8) c) == 0x85 \ || NATIVE_TO_LATIN1((U8) c) == 0xA0))) # define isUPPER_L1(c) (isUPPER_A(c) \ || (FITS_IN_8_BITS(c) \ - && (NATIVE_TO_LATIN1((U8) c) >= 0xC0 \ - && NATIVE_TO_LATIN1((U8) c) <= 0xDE \ + && ( IN_RANGE(NATIVE_TO_LATIN1((U8) c), \ + 0xC0, 0xDE) \ && NATIVE_TO_LATIN1((U8) c) != 0xD7))) # define isWORDCHAR_L1(c) (isIDFIRST_L1(c) || isDIGIT_A(c)) # define isIDFIRST_L1(c) (isALPHA_L1(c) || NATIVE_TO_LATIN1(c) == '_') @@ -1306,13 +1522,18 @@ END_EXTERN_C #define toLOWER(c) (isASCII(c) ? toLOWER_LATIN1(c) : (c)) #define toUPPER(c) (isASCII(c) ? toUPPER_LATIN1_MOD(c) : (c)) which uses table lookup and mask instead of subtraction. (This would - work because the _MOD does not apply in the ASCII range) */ + work because the _MOD does not apply in the ASCII range). + + These actually are UTF-8 invariant casing, not just ASCII, as any non-ASCII + UTF-8 invariants are neither upper nor lower. (Only on EBCDIC platforms are + there non-ASCII invariants, and all of them are controls.) */ #define toLOWER(c) (isUPPER(c) ? (U8)((c) + ('a' - 'A')) : (c)) #define toUPPER(c) (isLOWER(c) ? (U8)((c) - ('a' - 'A')) : (c)) /* In the ASCII range, these are equivalent to what they're here defined to be. * But by creating these definitions, other code doesn't have to be aware of - * this detail */ + * this detail. Actually this works for all UTF-8 invariants, not just the + * ASCII range. (EBCDIC platforms can have non-ASCII invariants.) */ #define toFOLD(c) toLOWER(c) #define toTITLE(c) toUPPER(c) @@ -1370,18 +1591,21 @@ END_EXTERN_C || (char)(c) == '_')) /* These next three are also for internal core Perl use only: case-change - * helper macros */ + * helper macros. The reason for using the PL_latin arrays is in case the + * system function is defective; it ensures uniform results that conform to the + * Unicod standard. It does not handle the anomalies in UTF-8 Turkic locales */ #define _generic_toLOWER_LC(c, function, cast) (! FITS_IN_8_BITS(c) \ ? (c) \ : (IN_UTF8_CTYPE_LOCALE) \ ? PL_latin1_lc[ (U8) (c) ] \ - : (cast)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 * SMALL LETTER SHARP S in a UTF-8 locale (which should be a string of two * values "SS"); instead it asserts against that under DEBUGGING, and - * otherwise returns its input */ + * otherwise returns its input. It does not handle the anomalies in UTF-8 + * Turkic locales. */ #define _generic_toUPPER_LC(c, function, cast) \ (! FITS_IN_8_BITS(c) \ ? (c) \ @@ -1399,7 +1623,8 @@ END_EXTERN_C * returns a single value, so can't adequately return the fold case of LATIN * SMALL LETTER SHARP S in a UTF-8 locale (which should be a string of two * values "ss"); instead it asserts against that under DEBUGGING, and - * otherwise returns its input */ + * otherwise returns its input. It does not handle the anomalies in UTF-8 + * Turkic locales */ #define _generic_toFOLD_LC(c, function, cast) \ ((UNLIKELY((c) == MICRO_SIGN) && IN_UTF8_CTYPE_LOCALE) \ ? GREEK_SMALL_LETTER_MU \ @@ -1508,56 +1733,83 @@ END_EXTERN_C #define isPSXSPC_LC(c) isSPACE_LC(c) /* For internal core Perl use only: the base macros for defining macros like - * isALPHA_uni. 'c' is the code point to check. 'classnum' is the POSIX class - * number defined earlier in this file. _generic_uni() is used for POSIX + * isALPHA_uvchr. 'c' is the code point to check. 'classnum' is the POSIX class + * number defined earlier in this file. _generic_uvchr() is used for POSIX * classes where there is a macro or function 'above_latin1' that takes the * single argument 'c' and returns the desired value. These exist for those * classes which have simple definitions, avoiding the overhead of a hash - * lookup or inversion list binary search. _generic_swash_uni() can be used + * lookup or inversion list binary search. _generic_swash_uvchr() can be used * for classes where that overhead is faster than a direct lookup. - * _generic_uni() won't compile if 'c' isn't unsigned, as it won't match the + * _generic_uvchr() won't compile if 'c' isn't unsigned, as it won't match the * 'above_latin1' prototype. _generic_isCC() macro does 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. */ /* Note that all ignore 'use bytes' */ -#define _generic_uni(classnum, above_latin1, c) ((c) < 256 \ - ? _generic_isCC(c, classnum) \ +#define _generic_uvchr(classnum, above_latin1, c) ((c) < 256 \ + ? _generic_isCC(c, classnum) \ : above_latin1(c)) -#define _generic_swash_uni(classnum, c) ((c) < 256 \ - ? _generic_isCC(c, classnum) \ +#define _generic_swash_uvchr(classnum, c) ((c) < 256 \ + ? _generic_isCC(c, classnum) \ : _is_uni_FOO(classnum, c)) -#define isALPHA_uni(c) _generic_swash_uni(_CC_ALPHA, c) -#define isALPHANUMERIC_uni(c) _generic_swash_uni(_CC_ALPHANUMERIC, c) -#define isASCII_uni(c) isASCII(c) -#define isBLANK_uni(c) _generic_uni(_CC_BLANK, is_HORIZWS_cp_high, c) -#define isCNTRL_uni(c) isCNTRL_L1(c) /* All controls are in Latin1 */ -#define isDIGIT_uni(c) _generic_swash_uni(_CC_DIGIT, c) -#define isGRAPH_uni(c) _generic_swash_uni(_CC_GRAPH, c) -#define isIDCONT_uni(c) _generic_uni(_CC_WORDCHAR, _is_uni_perl_idcont, c) -#define isIDFIRST_uni(c) _generic_uni(_CC_IDFIRST, _is_uni_perl_idstart, c) -#define isLOWER_uni(c) _generic_swash_uni(_CC_LOWER, c) -#define isPRINT_uni(c) _generic_swash_uni(_CC_PRINT, 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) -#define isXDIGIT_uni(c) _generic_uni(_CC_XDIGIT, is_XDIGIT_cp_high, c) - -#define toFOLD_uni(c,s,l) to_uni_fold(c,s,l) -#define toLOWER_uni(c,s,l) to_uni_lower(c,s,l) -#define toTITLE_uni(c,s,l) to_uni_title(c,s,l) -#define toUPPER_uni(c,s,l) to_uni_upper(c,s,l) +#define isALPHA_uvchr(c) _generic_swash_uvchr(_CC_ALPHA, c) +#define isALPHANUMERIC_uvchr(c) _generic_swash_uvchr(_CC_ALPHANUMERIC, c) +#define isASCII_uvchr(c) isASCII(c) +#define isBLANK_uvchr(c) _generic_uvchr(_CC_BLANK, is_HORIZWS_cp_high, c) +#define isCNTRL_uvchr(c) isCNTRL_L1(c) /* All controls are in Latin1 */ +#define isDIGIT_uvchr(c) _generic_swash_uvchr(_CC_DIGIT, c) +#define isGRAPH_uvchr(c) _generic_swash_uvchr(_CC_GRAPH, c) +#define isIDCONT_uvchr(c) \ + _generic_uvchr(_CC_WORDCHAR, _is_uni_perl_idcont, c) +#define isIDFIRST_uvchr(c) \ + _generic_uvchr(_CC_IDFIRST, _is_uni_perl_idstart, c) +#define isLOWER_uvchr(c) _generic_swash_uvchr(_CC_LOWER, c) +#define isPRINT_uvchr(c) _generic_swash_uvchr(_CC_PRINT, c) + +#define isPUNCT_uvchr(c) _generic_swash_uvchr(_CC_PUNCT, c) +#define isSPACE_uvchr(c) _generic_uvchr(_CC_SPACE, is_XPERLSPACE_cp_high, c) +#define isPSXSPC_uvchr(c) isSPACE_uvchr(c) + +#define isUPPER_uvchr(c) _generic_swash_uvchr(_CC_UPPER, c) +#define isVERTWS_uvchr(c) _generic_uvchr(_CC_VERTSPACE, is_VERTWS_cp_high, c) +#define isWORDCHAR_uvchr(c) _generic_swash_uvchr(_CC_WORDCHAR, c) +#define isXDIGIT_uvchr(c) _generic_uvchr(_CC_XDIGIT, is_XDIGIT_cp_high, c) + +#define toFOLD_uvchr(c,s,l) to_uni_fold(c,s,l) +#define toLOWER_uvchr(c,s,l) to_uni_lower(c,s,l) +#define toTITLE_uvchr(c,s,l) to_uni_title(c,s,l) +#define toUPPER_uvchr(c,s,l) to_uni_upper(c,s,l) + +/* For backwards compatibility, even though '_uni' should mean official Unicode + * code points, in Perl it means native for those below 256 */ +#define isALPHA_uni(c) isALPHA_uvchr(c) +#define isALPHANUMERIC_uni(c) isALPHANUMERIC_uvchr(c) +#define isASCII_uni(c) isASCII_uvchr(c) +#define isBLANK_uni(c) isBLANK_uvchr(c) +#define isCNTRL_uni(c) isCNTRL_uvchr(c) +#define isDIGIT_uni(c) isDIGIT_uvchr(c) +#define isGRAPH_uni(c) isGRAPH_uvchr(c) +#define isIDCONT_uni(c) isIDCONT_uvchr(c) +#define isIDFIRST_uni(c) isIDFIRST_uvchr(c) +#define isLOWER_uni(c) isLOWER_uvchr(c) +#define isPRINT_uni(c) isPRINT_uvchr(c) +#define isPUNCT_uni(c) isPUNCT_uvchr(c) +#define isSPACE_uni(c) isSPACE_uvchr(c) +#define isPSXSPC_uni(c) isPSXSPC_uvchr(c) +#define isUPPER_uni(c) isUPPER_uvchr(c) +#define isVERTWS_uni(c) isVERTWS_uvchr(c) +#define isWORDCHAR_uni(c) isWORDCHAR_uvchr(c) +#define isXDIGIT_uni(c) isXDIGIT_uvchr(c) +#define toFOLD_uni(c,s,l) toFOLD_uvchr(c,s,l) +#define toLOWER_uni(c,s,l) toLOWER_uvchr(c,s,l) +#define toTITLE_uni(c,s,l) toTITLE_uvchr(c,s,l) +#define toUPPER_uni(c,s,l) toUPPER_uvchr(c,s,l) /* For internal core Perl use only: the base macros for defining macros like * isALPHA_LC_uvchr. These are like isALPHA_LC, but the input can be any code - * point, not just 0-255. Like _generic_uni, there are two versions, one for + * point, not just 0-255. Like _generic_uvchr, there are two versions, one for * simple class definitions; the other for more complex. These are like - * _generic_uni, so see it for more info. */ + * _generic_uvchr, so see it for more info. */ #define _generic_LC_uvchr(latin1, above_latin1, c) \ (c < 256 ? latin1(c) : above_latin1(c)) #define _generic_LC_swash_uvchr(latin1, classnum, c) \ @@ -1597,62 +1849,121 @@ END_EXTERN_C * 'utf8' parameter. This relies on the fact that ASCII characters have the * same representation whether utf8 or not. Note that it assumes that the utf8 * has been validated, and ignores 'use bytes' */ -#define _generic_utf8(classnum, p, utf8) (UTF8_IS_INVARIANT(*(p)) \ - ? _generic_isCC(*(p), classnum) \ - : (UTF8_IS_DOWNGRADEABLE_START(*(p))) \ - ? _generic_isCC( \ - EIGHT_BIT_UTF8_TO_NATIVE(*(p), \ - *((p)+1 )), \ - classnum) \ - : utf8) +#define _base_generic_utf8(enum_name, name, p, use_locale ) \ + _is_utf8_FOO(CAT2(_CC_, enum_name), \ + (const U8 *) p, \ + "is" STRINGIFY(name) "_utf8", \ + "is" STRINGIFY(name) "_utf8_safe", \ + 1, use_locale, __FILE__,__LINE__) + +#define _generic_utf8(name, p) _base_generic_utf8(name, name, p, 0) + +/* The "_safe" macros make sure that we don't attempt to read beyond 'e', but + * they don't otherwise go out of their way to look for malformed UTF-8. If + * they can return accurate results without knowing if the input is otherwise + * malformed, they do so. For example isASCII is accurate in spite of any + * non-length malformations because it looks only at a single byte. Likewise + * isDIGIT looks just at the first byte for code points 0-255, as all UTF-8 + * variant ones return FALSE. But, if the input has to be well-formed in order + * for the results to be accurate, the macros will test and if malformed will + * call a routine to die + * + * Except for toke.c, the macros do assume that e > p, asserting that on + * DEBUGGING builds. Much code that calls these depends on this being true, + * for other reasons. toke.c is treated specially as using the regular + * assertion breaks it in many ways. All strings that these operate on there + * are supposed to have an extra NUL character at the end, so that *e = \0. A + * bunch of code in toke.c assumes that this is true, so the assertion allows + * for that */ +#ifdef PERL_IN_TOKE_C +# define _utf8_safe_assert(p,e) ((e) > (p) || ((e) == (p) && *(p) == '\0')) +#else +# define _utf8_safe_assert(p,e) ((e) > (p)) +#endif + +#define _generic_utf8_safe(classnum, p, e, above_latin1) \ + (__ASSERT_(_utf8_safe_assert(p, e)) \ + (UTF8_IS_INVARIANT(*(p))) \ + ? _generic_isCC(*(p), classnum) \ + : (UTF8_IS_DOWNGRADEABLE_START(*(p)) \ + ? ((LIKELY((e) - (p) > 1 && UTF8_IS_CONTINUATION(*((p)+1)))) \ + ? _generic_isCC(EIGHT_BIT_UTF8_TO_NATIVE(*(p), *((p)+1 )), \ + classnum) \ + : (_force_out_malformed_utf8_message( \ + (U8 *) (p), (U8 *) (e), 0, 1), 0)) \ + : above_latin1)) /* Like the above, but calls 'above_latin1(p)' to get the utf8 value. * 'above_latin1' can be a macro */ -#define _generic_func_utf8(classnum, above_latin1, p) \ - _generic_utf8(classnum, p, above_latin1(p)) +#define _generic_func_utf8_safe(classnum, above_latin1, p, e) \ + _generic_utf8_safe(classnum, p, e, above_latin1(p, e)) +#define _generic_non_swash_utf8_safe(classnum, above_latin1, p, e) \ + _generic_utf8_safe(classnum, p, e, \ + (UNLIKELY((e) - (p) < UTF8SKIP(p)) \ + ? (_force_out_malformed_utf8_message( \ + (U8 *) (p), (U8 *) (e), 0, 1), 0) \ + : above_latin1(p))) /* Like the above, but passes classnum to _isFOO_utf8(), instead of having an * 'above_latin1' parameter */ -#define _generic_swash_utf8(classnum, p) \ - _generic_utf8(classnum, p, _is_utf8_FOO(classnum, p)) +#define _generic_swash_utf8_safe(classnum, p, e) \ +_generic_utf8_safe(classnum, p, e, _is_utf8_FOO_with_len(classnum, p, e)) /* Like the above, but should be used only when it is known that there are no * characters in the upper-Latin1 range (128-255 on ASCII platforms) which the * class is TRUE for. Hence it can skip the tests for this range. * 'above_latin1' should include its arguments */ -#define _generic_utf8_no_upper_latin1(classnum, p, above_latin1) \ - (UTF8_IS_INVARIANT(*(p)) \ - ? _generic_isCC(*(p), classnum) \ - : (UTF8_IS_ABOVE_LATIN1(*(p))) \ - ? above_latin1 \ - : 0) - -/* NOTE that some of these macros have very similar ones in regcharclass.h. - * For example, there is (at the time of this writing) an 'is_SPACE_utf8()' - * there, differing in name only by an underscore from the one here - * 'isSPACE_utf8(). The difference is that the ones here are probably more - * efficient and smaller, using an O(1) array lookup for Latin1-range code - * points; the regcharclass.h ones are implemented as a series of - * "if-else-if-else ..." */ - -#define isALPHA_utf8(p) _generic_swash_utf8(_CC_ALPHA, p) -#define isALPHANUMERIC_utf8(p) _generic_swash_utf8(_CC_ALPHANUMERIC, p) -#define isASCII_utf8(p) isASCII(*p) /* Because ASCII is invariant under - utf8, the non-utf8 macro works - */ -#define isBLANK_utf8(p) _generic_func_utf8(_CC_BLANK, is_HORIZWS_high, p) +#define _generic_utf8_safe_no_upper_latin1(classnum, p, e, above_latin1) \ + (__ASSERT_(_utf8_safe_assert(p, e)) \ + (UTF8_IS_INVARIANT(*(p))) \ + ? _generic_isCC(*(p), classnum) \ + : (UTF8_IS_DOWNGRADEABLE_START(*(p))) \ + ? 0 /* Note that doesn't check validity for latin1 */ \ + : above_latin1) + + +#define isALPHA_utf8(p) _generic_utf8(ALPHA, p) +#define isALPHANUMERIC_utf8(p) _generic_utf8(ALPHANUMERIC, p) +#define isASCII_utf8(p) _generic_utf8(ASCII, p) +#define isBLANK_utf8(p) _generic_utf8(BLANK, p) +#define isCNTRL_utf8(p) _generic_utf8(CNTRL, p) +#define isDIGIT_utf8(p) _generic_utf8(DIGIT, p) +#define isGRAPH_utf8(p) _generic_utf8(GRAPH, p) +#define isIDCONT_utf8(p) _generic_utf8(IDCONT, p) +#define isIDFIRST_utf8(p) _generic_utf8(IDFIRST, p) +#define isLOWER_utf8(p) _generic_utf8(LOWER, p) +#define isPRINT_utf8(p) _generic_utf8(PRINT, p) +#define isPSXSPC_utf8(p) _generic_utf8(PSXSPC, p) +#define isPUNCT_utf8(p) _generic_utf8(PUNCT, p) +#define isSPACE_utf8(p) _generic_utf8(SPACE, p) +#define isUPPER_utf8(p) _generic_utf8(UPPER, p) +#define isVERTWS_utf8(p) _generic_utf8(VERTSPACE, p) +#define isWORDCHAR_utf8(p) _generic_utf8(WORDCHAR, p) +#define isXDIGIT_utf8(p) _generic_utf8(XDIGIT, p) + +#define isALPHA_utf8_safe(p, e) _generic_swash_utf8_safe(_CC_ALPHA, p, e) +#define isALPHANUMERIC_utf8_safe(p, e) \ + _generic_swash_utf8_safe(_CC_ALPHANUMERIC, p, e) +#define isASCII_utf8_safe(p, e) \ + /* Because ASCII is invariant under utf8, the non-utf8 macro \ + * works */ \ + (__ASSERT_(_utf8_safe_assert(p, e)) isASCII(*(p))) +#define isBLANK_utf8_safe(p, e) \ + _generic_non_swash_utf8_safe(_CC_BLANK, is_HORIZWS_high, p, e) #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_safe(p, e) \ + (__ASSERT_(_utf8_safe_assert(p, e)) isCNTRL_L1(*(p))) #else -# define isCNTRL_utf8(p) _generic_utf8(_CC_CNTRL, p, 0) +# define isCNTRL_utf8_safe(p, e) _generic_utf8_safe(_CC_CNTRL, p, e, 0) #endif -#define isDIGIT_utf8(p) _generic_utf8_no_upper_latin1(_CC_DIGIT, p, \ - _is_utf8_FOO(_CC_DIGIT, p)) -#define isGRAPH_utf8(p) _generic_swash_utf8(_CC_GRAPH, p) -#define isIDCONT_utf8(p) _generic_func_utf8(_CC_WORDCHAR, \ - _is_utf8_perl_idcont, p) +#define isDIGIT_utf8_safe(p, e) \ + _generic_utf8_safe_no_upper_latin1(_CC_DIGIT, p, e, \ + _is_utf8_FOO_with_len(_CC_DIGIT, p, e)) +#define isGRAPH_utf8_safe(p, e) _generic_swash_utf8_safe(_CC_GRAPH, p, e) +#define isIDCONT_utf8_safe(p, e) _generic_func_utf8_safe(_CC_WORDCHAR, \ + _is_utf8_perl_idcont_with_len, p, e) /* To prevent S_scan_word in toke.c from hanging, we have to make sure that * IDFIRST is an alnum. See @@ -1660,65 +1971,133 @@ END_EXTERN_C * ever wanted to know about. (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) _generic_func_utf8(_CC_IDFIRST, \ - _is_utf8_perl_idstart, p) - -#define isLOWER_utf8(p) _generic_swash_utf8(_CC_LOWER, p) -#define isPRINT_utf8(p) _generic_swash_utf8(_CC_PRINT, 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) -#define isVERTWS_utf8(p) _generic_func_utf8(_CC_VERTSPACE, is_VERTWS_high, p) -#define isWORDCHAR_utf8(p) _generic_swash_utf8(_CC_WORDCHAR, p) -#define isXDIGIT_utf8(p) _generic_utf8_no_upper_latin1(_CC_XDIGIT, p, \ - is_XDIGIT_high(p)) +#define isIDFIRST_utf8_safe(p, e) \ + _generic_func_utf8_safe(_CC_IDFIRST, \ + _is_utf8_perl_idstart_with_len, (U8 *) (p), (U8 *) (e)) + +#define isLOWER_utf8_safe(p, e) _generic_swash_utf8_safe(_CC_LOWER, p, e) +#define isPRINT_utf8_safe(p, e) _generic_swash_utf8_safe(_CC_PRINT, p, e) +#define isPSXSPC_utf8_safe(p, e) isSPACE_utf8_safe(p, e) +#define isPUNCT_utf8_safe(p, e) _generic_swash_utf8_safe(_CC_PUNCT, p, e) +#define isSPACE_utf8_safe(p, e) \ + _generic_non_swash_utf8_safe(_CC_SPACE, is_XPERLSPACE_high, p, e) +#define isUPPER_utf8_safe(p, e) _generic_swash_utf8_safe(_CC_UPPER, p, e) +#define isVERTWS_utf8_safe(p, e) \ + _generic_non_swash_utf8_safe(_CC_VERTSPACE, is_VERTWS_high, p, e) +#define isWORDCHAR_utf8_safe(p, e) \ + _generic_swash_utf8_safe(_CC_WORDCHAR, p, e) +#define isXDIGIT_utf8_safe(p, e) \ + _generic_utf8_safe_no_upper_latin1(_CC_XDIGIT, p, e, \ + (UNLIKELY((e) - (p) < UTF8SKIP(p)) \ + ? (_force_out_malformed_utf8_message( \ + (U8 *) (p), (U8 *) (e), 0, 1), 0) \ + : is_XDIGIT_high(p))) #define toFOLD_utf8(p,s,l) to_utf8_fold(p,s,l) #define toLOWER_utf8(p,s,l) to_utf8_lower(p,s,l) #define toTITLE_utf8(p,s,l) to_utf8_title(p,s,l) #define toUPPER_utf8(p,s,l) to_utf8_upper(p,s,l) +/* For internal core use only, subject to change */ +#define _toFOLD_utf8_flags(p,e,s,l,f) _to_utf8_fold_flags (p,e,s,l,f, "", 0) +#define _toLOWER_utf8_flags(p,e,s,l,f) _to_utf8_lower_flags(p,e,s,l,f, "", 0) +#define _toTITLE_utf8_flags(p,e,s,l,f) _to_utf8_title_flags(p,e,s,l,f, "", 0) +#define _toUPPER_utf8_flags(p,e,s,l,f) _to_utf8_upper_flags(p,e,s,l,f, "", 0) + +#define toFOLD_utf8_safe(p,e,s,l) _toFOLD_utf8_flags(p,e,s,l, FOLD_FLAGS_FULL) +#define toLOWER_utf8_safe(p,e,s,l) _toLOWER_utf8_flags(p,e,s,l, 0) +#define toTITLE_utf8_safe(p,e,s,l) _toTITLE_utf8_flags(p,e,s,l, 0) +#define toUPPER_utf8_safe(p,e,s,l) _toUPPER_utf8_flags(p,e,s,l, 0) + /* For internal core Perl use only: the base macros for defining macros like * isALPHA_LC_utf8. These are like _generic_utf8, but if the first code point * in 'p' is within the 0-255 range, it uses locale rules from the passed-in * 'macro' parameter */ -#define _generic_LC_utf8(macro, p, utf8) \ - (UTF8_IS_INVARIANT(*(p)) \ - ? macro(*(p)) \ - : (UTF8_IS_DOWNGRADEABLE_START(*(p))) \ - ? macro(EIGHT_BIT_UTF8_TO_NATIVE(*(p), *((p)+1)))\ - : utf8) - -#define _generic_LC_swash_utf8(macro, classnum, p) \ - _generic_LC_utf8(macro, p, _is_utf8_FOO(classnum, p)) -#define _generic_LC_func_utf8(macro, above_latin1, p) \ - _generic_LC_utf8(macro, p, above_latin1(p)) - -#define isALPHANUMERIC_LC_utf8(p) _generic_LC_swash_utf8(isALPHANUMERIC_LC, \ - _CC_ALPHANUMERIC, p) -#define isALPHA_LC_utf8(p) _generic_LC_swash_utf8(isALPHA_LC, _CC_ALPHA, p) -#define isASCII_LC_utf8(p) isASCII_LC(*p) -#define isBLANK_LC_utf8(p) _generic_LC_func_utf8(isBLANK_LC, \ - is_HORIZWS_high, p) -#define isCNTRL_LC_utf8(p) _generic_LC_utf8(isCNTRL_LC, p, 0) -#define isDIGIT_LC_utf8(p) _generic_LC_swash_utf8(isDIGIT_LC, _CC_DIGIT, p) -#define isGRAPH_LC_utf8(p) _generic_LC_swash_utf8(isGRAPH_LC, _CC_GRAPH, p) -#define isIDCONT_LC_utf8(p) _generic_LC_func_utf8(isIDCONT_LC, \ - _is_utf8_perl_idcont, p) -#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) -#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) -#define isWORDCHAR_LC_utf8(p) _generic_LC_swash_utf8(isWORDCHAR_LC, \ - _CC_WORDCHAR, p) -#define isXDIGIT_LC_utf8(p) _generic_LC_func_utf8(isXDIGIT_LC, \ - is_XDIGIT_high, p) +#define _generic_LC_utf8(name, p) _base_generic_utf8(name, name, p, 1) + +#define isALPHA_LC_utf8(p) _generic_LC_utf8(ALPHA, p) +#define isALPHANUMERIC_LC_utf8(p) _generic_LC_utf8(ALPHANUMERIC, p) +#define isASCII_LC_utf8(p) _generic_LC_utf8(ASCII, p) +#define isBLANK_LC_utf8(p) _generic_LC_utf8(BLANK, p) +#define isCNTRL_LC_utf8(p) _generic_LC_utf8(CNTRL, p) +#define isDIGIT_LC_utf8(p) _generic_LC_utf8(DIGIT, p) +#define isGRAPH_LC_utf8(p) _generic_LC_utf8(GRAPH, p) +#define isIDCONT_LC_utf8(p) _generic_LC_utf8(IDCONT, p) +#define isIDFIRST_LC_utf8(p) _generic_LC_utf8(IDFIRST, p) +#define isLOWER_LC_utf8(p) _generic_LC_utf8(LOWER, p) +#define isPRINT_LC_utf8(p) _generic_LC_utf8(PRINT, p) +#define isPSXSPC_LC_utf8(p) _generic_LC_utf8(PSXSPC, p) +#define isPUNCT_LC_utf8(p) _generic_LC_utf8(PUNCT, p) +#define isSPACE_LC_utf8(p) _generic_LC_utf8(SPACE, p) +#define isUPPER_LC_utf8(p) _generic_LC_utf8(UPPER, p) +#define isWORDCHAR_LC_utf8(p) _generic_LC_utf8(WORDCHAR, p) +#define isXDIGIT_LC_utf8(p) _generic_LC_utf8(XDIGIT, p) + +/* For internal core Perl use only: the base macros for defining macros like + * isALPHA_LC_utf8_safe. These are like _generic_utf8, but if the first code + * point in 'p' is within the 0-255 range, it uses locale rules from the + * passed-in 'macro' parameter */ +#define _generic_LC_utf8_safe(macro, p, e, above_latin1) \ + (__ASSERT_(_utf8_safe_assert(p, e)) \ + (UTF8_IS_INVARIANT(*(p))) \ + ? macro(*(p)) \ + : (UTF8_IS_DOWNGRADEABLE_START(*(p)) \ + ? ((LIKELY((e) - (p) > 1 && UTF8_IS_CONTINUATION(*((p)+1)))) \ + ? macro(EIGHT_BIT_UTF8_TO_NATIVE(*(p), *((p)+1))) \ + : (_force_out_malformed_utf8_message( \ + (U8 *) (p), (U8 *) (e), 0, 1), 0)) \ + : above_latin1)) + +#define _generic_LC_swash_utf8_safe(macro, classnum, p, e) \ + _generic_LC_utf8_safe(macro, p, e, \ + _is_utf8_FOO_with_len(classnum, p, e)) + +#define _generic_LC_func_utf8_safe(macro, above_latin1, p, e) \ + _generic_LC_utf8_safe(macro, p, e, above_latin1(p, e)) + +#define _generic_LC_non_swash_utf8_safe(classnum, above_latin1, p, e) \ + _generic_LC_utf8_safe(classnum, p, e, \ + (UNLIKELY((e) - (p) < UTF8SKIP(p)) \ + ? (_force_out_malformed_utf8_message( \ + (U8 *) (p), (U8 *) (e), 0, 1), 0) \ + : above_latin1(p))) + +#define isALPHANUMERIC_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isALPHANUMERIC_LC, \ + _CC_ALPHANUMERIC, p, e) +#define isALPHA_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isALPHA_LC, _CC_ALPHA, p, e) +#define isASCII_LC_utf8_safe(p, e) \ + (__ASSERT_(_utf8_safe_assert(p, e)) isASCII_LC(*(p))) +#define isBLANK_LC_utf8_safe(p, e) \ + _generic_LC_non_swash_utf8_safe(isBLANK_LC, is_HORIZWS_high, p, e) +#define isCNTRL_LC_utf8_safe(p, e) \ + _generic_LC_utf8_safe(isCNTRL_LC, p, e, 0) +#define isDIGIT_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isDIGIT_LC, _CC_DIGIT, p, e) +#define isGRAPH_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isGRAPH_LC, _CC_GRAPH, p, e) +#define isIDCONT_LC_utf8_safe(p, e) \ + _generic_LC_func_utf8_safe(isIDCONT_LC, \ + _is_utf8_perl_idcont_with_len, p, e) +#define isIDFIRST_LC_utf8_safe(p, e) \ + _generic_LC_func_utf8_safe(isIDFIRST_LC, \ + _is_utf8_perl_idstart_with_len, p, e) +#define isLOWER_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isLOWER_LC, _CC_LOWER, p, e) +#define isPRINT_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isPRINT_LC, _CC_PRINT, p, e) +#define isPSXSPC_LC_utf8_safe(p, e) isSPACE_LC_utf8_safe(p, e) +#define isPUNCT_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isPUNCT_LC, _CC_PUNCT, p, e) +#define isSPACE_LC_utf8_safe(p, e) \ + _generic_LC_non_swash_utf8_safe(isSPACE_LC, is_XPERLSPACE_high, p, e) +#define isUPPER_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isUPPER_LC, _CC_UPPER, p, e) +#define isWORDCHAR_LC_utf8_safe(p, e) \ + _generic_LC_swash_utf8_safe(isWORDCHAR_LC, _CC_WORDCHAR, p, e) +#define isXDIGIT_LC_utf8_safe(p, e) \ + _generic_LC_non_swash_utf8_safe(isXDIGIT_LC, is_XDIGIT_high, p, e) /* Macros for backwards compatibility and for completeness when the ASCII and * Latin1 values are identical */ @@ -1728,6 +2107,7 @@ END_EXTERN_C #define isOCTAL_L1(c) isOCTAL_A(c) #define isXDIGIT_L1(c) isXDIGIT_A(c) #define isALNUM(c) isWORDCHAR(c) +#define isALNUM_A(c) isALNUM(c) #define isALNUMU(c) isWORDCHAR_L1(c) #define isALNUM_LC(c) isWORDCHAR_LC(c) #define isALNUM_uni(c) isWORDCHAR_uni(c) @@ -1824,7 +2204,7 @@ typedef U32 line_t; =for apidoc Am|void|Newx|void* ptr|int nitems|type The XSUB-writer's interface to the C C function. -Memory obtained by this should B be freed with L<"Safefree">. +Memory obtained by this should B be freed with L. 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 @@ -1836,29 +2216,29 @@ there for use in XS modules supporting older perls. The XSUB-writer's interface to the C C function, with cast. See also C>. -Memory obtained by this should B be freed with L<"Safefree">. +Memory obtained by this should B be freed with L. =for apidoc Am|void|Newxz|void* ptr|int nitems|type The XSUB-writer's interface to the C C function. The allocated memory is zeroed with C. See also C>. -Memory obtained by this should B be freed with L<"Safefree">. +Memory obtained by this should B be freed with L. =for apidoc Am|void|Renew|void* ptr|int nitems|type The XSUB-writer's interface to the C C function. -Memory obtained by this should B be freed with L<"Safefree">. +Memory obtained by this should B be freed with L. =for apidoc Am|void|Renewc|void* ptr|int nitems|type|cast The XSUB-writer's interface to the C C function, with cast. -Memory obtained by this should B be freed with L<"Safefree">. +Memory obtained by this should B be freed with L. =for apidoc Am|void|Safefree|void* ptr The XSUB-writer's interface to the C C function. -This should B be used on memory obtained using L<"Newx"> and friends. +This should B be used on memory obtained using L and friends. =for apidoc Am|void|Move|void* src|void* dest|int nitems|type The XSUB-writer's interface to the C C function. The C is the @@ -1919,8 +2299,9 @@ PoisonWith(0xEF) for catching access to freed memory. #define NEWSV(x,len) newSV(len) #endif -#define MEM_SIZE_MAX ((MEM_SIZE)~0) +#define MEM_SIZE_MAX ((MEM_SIZE)-1) +#define _PERL_STRLEN_ROUNDUP_UNCHECKED(n) (((n) - 1 + PERL_STRLEN_ROUNDUP_QUANTUM) & ~((MEM_SIZE)PERL_STRLEN_ROUNDUP_QUANTUM - 1)) #ifdef PERL_MALLOC_WRAP @@ -1967,17 +2348,22 @@ PoisonWith(0xEF) for catching access to freed memory. (void)(UNLIKELY(_MEM_WRAP_WILL_WRAP(n,t)) \ && (Perl_croak_nocontext("%s",(a)),0)) +/* "a" arg must be a string literal */ +# define MEM_WRAP_CHECK_s(n,t,a) \ + (void)(UNLIKELY(_MEM_WRAP_WILL_WRAP(n,t)) \ + && (Perl_croak_nocontext("" 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))) +#define PERL_STRLEN_ROUNDUP(n) ((void)(((n) > MEM_SIZE_MAX - 2 * PERL_STRLEN_ROUNDUP_QUANTUM) ? (croak_memory_wrap(),0) : 0), _PERL_STRLEN_ROUNDUP_UNCHECKED(n)) #else #define MEM_WRAP_CHECK(n,t) #define MEM_WRAP_CHECK_1(n,t,a) -#define MEM_WRAP_CHECK_2(n,t,a,b) +#define MEM_WRAP_CHECK_s(n,t,a) #define MEM_WRAP_CHECK_(n,t) -#define PERL_STRLEN_ROUNDUP(n) (((n-1+PERL_STRLEN_ROUNDUP_QUANTUM)&~((MEM_SIZE)PERL_STRLEN_ROUNDUP_QUANTUM-1))) +#define PERL_STRLEN_ROUNDUP(n) _PERL_STRLEN_ROUNDUP_UNCHECKED(n) #endif @@ -2014,12 +2400,6 @@ PoisonWith(0xEF) for catching access to freed memory. * - lots of ENV reads */ -PERL_CALLCONV Malloc_t Perl_mem_log_alloc(const UV n, const UV typesize, const char *type_name, Malloc_t newalloc, const char *filename, const int linenumber, const char *funcname); - -PERL_CALLCONV Malloc_t Perl_mem_log_realloc(const UV n, const UV typesize, const char *type_name, Malloc_t oldalloc, Malloc_t newalloc, const char *filename, const int linenumber, const char *funcname); - -PERL_CALLCONV Malloc_t Perl_mem_log_free(Malloc_t oldalloc, const char *filename, const int linenumber, const char *funcname); - # ifdef PERL_CORE # ifndef PERL_MEM_LOG_NOIMPL enum mem_log_type { @@ -2077,18 +2457,20 @@ void Perl_mem_log_del_sv(const SV *sv, const char *filename, const int linenumbe #define Safefree(d) safefree(MEM_LOG_FREE((Malloc_t)(d))) #endif -#define Move(s,d,n,t) (MEM_WRAP_CHECK_(n,t) (void)memmove((char*)(d),(const char*)(s), (n) * sizeof(t))) -#define Copy(s,d,n,t) (MEM_WRAP_CHECK_(n,t) (void)memcpy((char*)(d),(const char*)(s), (n) * sizeof(t))) -#define Zero(d,n,t) (MEM_WRAP_CHECK_(n,t) (void)memzero((char*)(d), (n) * sizeof(t))) +/* assert that a valid ptr has been supplied - use this instead of assert(ptr) * + * as it handles cases like constant string arguments without throwing warnings * + * the cast is required, as is the inequality check, to avoid warnings */ +#define perl_assert_ptr(p) assert( ((void*)(p)) != 0 ) -#define MoveD(s,d,n,t) (MEM_WRAP_CHECK_(n,t) memmove((char*)(d),(const char*)(s), (n) * sizeof(t))) -#define CopyD(s,d,n,t) (MEM_WRAP_CHECK_(n,t) memcpy((char*)(d),(const char*)(s), (n) * sizeof(t))) -#ifdef HAS_MEMSET -#define ZeroD(d,n,t) (MEM_WRAP_CHECK_(n,t) memzero((char*)(d), (n) * sizeof(t))) -#else -/* Using bzero(), which returns void. */ -#define ZeroD(d,n,t) (MEM_WRAP_CHECK_(n,t) memzero((char*)(d), (n) * sizeof(t)),d) -#endif + +#define Move(s,d,n,t) (MEM_WRAP_CHECK_(n,t) perl_assert_ptr(d), perl_assert_ptr(s), (void)memmove((char*)(d),(const char*)(s), (n) * sizeof(t))) +#define Copy(s,d,n,t) (MEM_WRAP_CHECK_(n,t) perl_assert_ptr(d), perl_assert_ptr(s), (void)memcpy((char*)(d),(const char*)(s), (n) * sizeof(t))) +#define Zero(d,n,t) (MEM_WRAP_CHECK_(n,t) perl_assert_ptr(d), (void)memzero((char*)(d), (n) * sizeof(t))) + +/* Like above, but returns a pointer to 'd' */ +#define MoveD(s,d,n,t) (MEM_WRAP_CHECK_(n,t) perl_assert_ptr(d), perl_assert_ptr(s), memmove((char*)(d),(const char*)(s), (n) * sizeof(t))) +#define CopyD(s,d,n,t) (MEM_WRAP_CHECK_(n,t) perl_assert_ptr(d), perl_assert_ptr(s), memcpy((char*)(d),(const char*)(s), (n) * sizeof(t))) +#define ZeroD(d,n,t) (MEM_WRAP_CHECK_(n,t) perl_assert_ptr(d), memzero((char*)(d), (n) * sizeof(t))) #define PoisonWith(d,n,t,b) (MEM_WRAP_CHECK_(n,t) (void)memset((char*)(d), (U8)(b), (n) * sizeof(t))) #define PoisonNew(d,n,t) PoisonWith(d,n,t,0xAB) @@ -2101,29 +2483,35 @@ void Perl_mem_log_del_sv(const SV *sv, const char *filename, const int linenumbe # define PERL_POISON_EXPR(x) #endif -#ifdef USE_STRUCT_COPY #define StructCopy(s,d,t) (*((t*)(d)) = *((t*)(s))) -#else -#define StructCopy(s,d,t) Copy(s,d,1,t) -#endif -/* C_ARRAY_LENGTH is the number of elements in the C array (so you - * want your zero-based indices to be less than but not equal to). - * - * C_ARRAY_END is one past the last: half-open/half-closed range, - * not last-inclusive range. */ +/* +=head1 Handy Values + +=for apidoc Am|STRLEN|C_ARRAY_LENGTH|void *a + +Returns the number of elements in the input C array (so you want your +zero-based indices to be less than but not equal to). + +=for apidoc Am|void *|C_ARRAY_END|void *a + +Returns a pointer to one element past the final element of the input C array. + +=cut + +C_ARRAY_END is one past the last: half-open/half-closed range, not +last-inclusive range. +*/ #define C_ARRAY_LENGTH(a) (sizeof(a)/sizeof((a)[0])) #define C_ARRAY_END(a) ((a) + C_ARRAY_LENGTH(a)) #ifdef NEED_VA_COPY # ifdef va_copy # define Perl_va_copy(s, d) va_copy(d, s) +# elif defined(__va_copy) +# define Perl_va_copy(s, d) __va_copy(d, s) # else -# if defined(__va_copy) -# define Perl_va_copy(s, d) __va_copy(d, s) -# else -# define Perl_va_copy(s, d) Copy(s, d, 1, va_list) -# endif +# define Perl_va_copy(s, d) Copy(s, d, 1, va_list) # endif #endif @@ -2150,6 +2538,12 @@ void Perl_mem_log_del_sv(const SV *sv, const char *filename, const int linenumbe #ifdef PERL_CORE # define deprecate(s) Perl_ck_warner_d(aTHX_ packWARN(WARN_DEPRECATED), \ "Use of " s " is deprecated") +# define deprecate_disappears_in(when,message) \ + Perl_ck_warner_d(aTHX_ packWARN(WARN_DEPRECATED), \ + message ", and will disappear in Perl " when) +# define deprecate_fatal_in(when,message) \ + Perl_ck_warner_d(aTHX_ packWARN(WARN_DEPRECATED), \ + message ". Its use will be fatal in Perl " when) #endif /* Internal macros to deal with gids and uids */ @@ -2158,32 +2552,28 @@ void Perl_mem_log_del_sv(const SV *sv, const char *filename, const int linenumbe # if Uid_t_size > IVSIZE # define sv_setuid(sv, uid) sv_setnv((sv), (NV)(uid)) # define SvUID(sv) SvNV(sv) +# elif Uid_t_sign <= 0 +# define sv_setuid(sv, uid) sv_setiv((sv), (IV)(uid)) +# define SvUID(sv) SvIV(sv) # else -# if Uid_t_sign <= 0 -# define sv_setuid(sv, uid) sv_setiv((sv), (IV)(uid)) -# define SvUID(sv) SvIV(sv) -# else -# define sv_setuid(sv, uid) sv_setuv((sv), (UV)(uid)) -# define SvUID(sv) SvUV(sv) -# endif +# define sv_setuid(sv, uid) sv_setuv((sv), (UV)(uid)) +# define SvUID(sv) SvUV(sv) # endif /* Uid_t_size */ # if Gid_t_size > IVSIZE # define sv_setgid(sv, gid) sv_setnv((sv), (NV)(gid)) # define SvGID(sv) SvNV(sv) +# elif Gid_t_sign <= 0 +# define sv_setgid(sv, gid) sv_setiv((sv), (IV)(gid)) +# define SvGID(sv) SvIV(sv) # else -# if Gid_t_sign <= 0 -# define sv_setgid(sv, gid) sv_setiv((sv), (IV)(gid)) -# define SvGID(sv) SvIV(sv) -# else -# define sv_setgid(sv, gid) sv_setuv((sv), (UV)(gid)) -# define SvGID(sv) SvUV(sv) -# endif +# define sv_setgid(sv, gid) sv_setuv((sv), (UV)(gid)) +# define SvGID(sv) SvUV(sv) # endif /* Gid_t_size */ #endif -#endif /* HANDY_H */ +#endif /* PERL_HANDY_H_ */ /* * ex: set ts=8 sts=4 sw=4 et: