X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/c82ecf346a8512f22f25188e450d92938c245421..1c43698bb7a30756ad27d5ff52376ba1d596e8bb:/inline.h diff --git a/inline.h b/inline.h index 6b24ae5..0792694 100644 --- a/inline.h +++ b/inline.h @@ -14,7 +14,7 @@ /* ------------------------------- av.h ------------------------------- */ -PERL_STATIC_INLINE I32 +PERL_STATIC_INLINE SSize_t S_av_top_index(pTHX_ AV *av) { PERL_ARGS_ASSERT_AV_TOP_INDEX; @@ -23,20 +23,6 @@ S_av_top_index(pTHX_ AV *av) return AvFILL(av); } -/* ------------------------------- cop.h ------------------------------ */ - -#ifdef USE_ITHREADS -PERL_STATIC_INLINE void -S_CopFILE_free(pTHX_ COP * const c) -{ - GV * const gv = CopFILEGV(c); - if (!gv) return; - if (SvREFCNT(gv) == 1) PL_filegvpad[c->cop_filegvoff] = NULL; - SvREFCNT_dec_NN(gv); - c->cop_filegvoff = 0; -} -#endif - /* ------------------------------- cv.h ------------------------------- */ PERL_STATIC_INLINE I32 * @@ -76,6 +62,26 @@ S_strip_spaces(pTHX_ const char * orig, STRLEN * const len) } #endif +/* ------------------------------- mg.h ------------------------------- */ + +#if defined(PERL_CORE) || defined(PERL_EXT) +/* assumes get-magic and stringification have already occurred */ +PERL_STATIC_INLINE STRLEN +S_MgBYTEPOS(pTHX_ MAGIC *mg, SV *sv, const char *s, STRLEN len) +{ + assert(mg->mg_type == PERL_MAGIC_regex_global); + assert(mg->mg_len != -1); + if (mg->mg_flags & MGf_BYTES || !DO_UTF8(sv)) + return (STRLEN)mg->mg_len; + else { + const STRLEN pos = (STRLEN)mg->mg_len; + /* Without this check, we may read past the end of the buffer: */ + if (pos > sv_or_pv_len_utf8(sv, s, len)) return len+1; + return sv_or_pv_pos_u2b(sv, s, pos, NULL); + } +} +#endif + /* ----------------------------- regexp.h ----------------------------- */ PERL_STATIC_INLINE struct regexp * @@ -122,7 +128,6 @@ PERL_STATIC_INLINE void S_SvREFCNT_dec_NN(pTHX_ SV *sv) { U32 rc = SvREFCNT(sv); - PERL_ARGS_ASSERT_SVREFCNT_DEC_NN; if (LIKELY(rc > 1)) SvREFCNT(sv) = rc - 1; else @@ -166,10 +171,11 @@ S_SvPADSTALE_off(SV *sv) assert(SvFLAGS(sv) & SVs_PADMY); return SvFLAGS(sv) &= ~SVs_PADSTALE; } -#ifdef PERL_CORE +#if defined(PERL_CORE) || defined (PERL_EXT) PERL_STATIC_INLINE STRLEN S_sv_or_pv_pos_u2b(pTHX_ SV *sv, const char *pv, STRLEN pos, STRLEN *lenp) { + PERL_ARGS_ASSERT_SV_OR_PV_POS_U2B; if (SvGAMAGIC(sv)) { U8 *hopped = utf8_hop((U8 *)pv, pos); if (lenp) *lenp = (STRLEN)(utf8_hop(hopped, *lenp) - hopped); @@ -182,36 +188,170 @@ S_sv_or_pv_pos_u2b(pTHX_ SV *sv, const char *pv, STRLEN pos, STRLEN *lenp) /* ------------------------------- handy.h ------------------------------- */ /* saves machine code for a common noreturn idiom typically used in Newx*() */ -#ifdef __clang__ -#pragma clang diagnostic push -#pragma clang diagnostic ignored "-Wunused-function" +#ifdef GCC_DIAG_PRAGMA +GCC_DIAG_IGNORE(-Wunused-function) /* Intentionally left semicolonless. */ #endif static void S_croak_memory_wrap(void) { Perl_croak_nocontext("%s",PL_memory_wrap); } -#ifdef __clang__ -#pragma clang diagnostic pop +#ifdef GCC_DIAG_PRAGMA +GCC_DIAG_RESTORE /* Intentionally left semicolonless. */ #endif /* ------------------------------- utf8.h ------------------------------- */ -/* These exist only to replace the macros they formerly were so that their use - * can be deprecated */ +PERL_STATIC_INLINE void +S_append_utf8_from_native_byte(const U8 byte, U8** dest) +{ + /* Takes an input 'byte' (Latin1 or EBCDIC) and appends it to the UTF-8 + * encoded string at '*dest', updating '*dest' to include it */ -PERL_STATIC_INLINE bool -S_isIDFIRST_lazy(pTHX_ const char* p) + PERL_ARGS_ASSERT_APPEND_UTF8_FROM_NATIVE_BYTE; + + if (NATIVE_BYTE_IS_INVARIANT(byte)) + *(*dest)++ = byte; + else { + *(*dest)++ = UTF8_EIGHT_BIT_HI(byte); + *(*dest)++ = UTF8_EIGHT_BIT_LO(byte); + } +} + +/* + +A helper function for the macro isUTF8_CHAR(), which should be used instead of +this function. The macro will handle smaller code points directly saving time, +using this function as a fall-back for higher code points. + +Tests if the first bytes of string C form a valid UTF-8 character. 0 is +returned if the bytes starting at C up to but not including C do not form a +complete well-formed UTF-8 character; otherwise the number of bytes in the +character is returned. + +Note that an INVARIANT (i.e. ASCII on non-EBCDIC) character is a valid UTF-8 +character. + +=cut */ +PERL_STATIC_INLINE STRLEN +S__is_utf8_char_slow(const U8 *s, const U8 *e) { - PERL_ARGS_ASSERT_ISIDFIRST_LAZY; + dTHX; /* The function called below requires thread context */ - return isIDFIRST_lazy_if(p,1); + STRLEN actual_len; + + PERL_ARGS_ASSERT__IS_UTF8_CHAR_SLOW; + + assert(e >= s); + utf8n_to_uvchr(s, e - s, &actual_len, UTF8_CHECK_ONLY); + + return (actual_len == (STRLEN) -1) ? 0 : actual_len; } +/* ------------------------------- perl.h ----------------------------- */ + +/* +=head1 Miscellaneous Functions + +=for apidoc AiR|bool|is_safe_syscall|const char *pv|STRLEN len|const char *what|const char *op_name + +Test that the given C doesn't contain any internal C characters. +If it does, set C to ENOENT, optionally warn, and return FALSE. + +Return TRUE if the name is safe. + +Used by the IS_SAFE_SYSCALL() macro. + +=cut +*/ + +PERL_STATIC_INLINE bool +S_is_safe_syscall(pTHX_ const char *pv, STRLEN len, const char *what, const char *op_name) { + /* While the Windows CE API provides only UCS-16 (or UTF-16) APIs + * perl itself uses xce*() functions which accept 8-bit strings. + */ + + PERL_ARGS_ASSERT_IS_SAFE_SYSCALL; + + if (pv && len > 1) { + char *null_at; + if (UNLIKELY((null_at = (char *)memchr(pv, 0, len-1)) != NULL)) { + SETERRNO(ENOENT, LIB_INVARG); + Perl_ck_warner(aTHX_ packWARN(WARN_SYSCALLS), + "Invalid \\0 character in %s for %s: %s\\0%s", + what, op_name, pv, null_at+1); + return FALSE; + } + } + + return TRUE; +} + +/* + +Return true if the supplied filename has a newline character +immediately before the final NUL. + +My original look at this incorrectly used the len from SvPV(), but +that's incorrect, since we allow for a NUL in pv[len-1]. + +So instead, strlen() and work from there. + +This allow for the user reading a filename, forgetting to chomp it, +then calling: + + open my $foo, "$file\0"; + +*/ + +#ifdef PERL_CORE + PERL_STATIC_INLINE bool -S_isALNUM_lazy(pTHX_ const char* p) +S_should_warn_nl(const char *pv) { + STRLEN len; + + PERL_ARGS_ASSERT_SHOULD_WARN_NL; + + len = strlen(pv); + + return len > 0 && pv[len-1] == '\n'; +} + +#endif + +/* ------------------ pp.c, regcomp.c, toke.c, universal.c ------------ */ + +#define MAX_CHARSET_NAME_LENGTH 2 + +PERL_STATIC_INLINE const char * +get_regex_charset_name(const U32 flags, STRLEN* const lenp) { - PERL_ARGS_ASSERT_ISALNUM_LAZY; + /* Returns a string that corresponds to the name of the regex character set + * given by 'flags', and *lenp is set the length of that string, which + * cannot exceed MAX_CHARSET_NAME_LENGTH characters */ - return isALNUM_lazy_if(p,1); + *lenp = 1; + switch (get_regex_charset(flags)) { + case REGEX_DEPENDS_CHARSET: return DEPENDS_PAT_MODS; + case REGEX_LOCALE_CHARSET: return LOCALE_PAT_MODS; + case REGEX_UNICODE_CHARSET: return UNICODE_PAT_MODS; + case REGEX_ASCII_RESTRICTED_CHARSET: return ASCII_RESTRICT_PAT_MODS; + case REGEX_ASCII_MORE_RESTRICTED_CHARSET: + *lenp = 2; + return ASCII_MORE_RESTRICT_PAT_MODS; + } + /* The NOT_REACHED; hides an assert() which has a rather complex + * definition in perl.h. */ + NOT_REACHED; /* NOTREACHED */ + return "?"; /* Unknown */ } + +/* + * Local variables: + * c-indentation-style: bsd + * c-basic-offset: 4 + * indent-tabs-mode: nil + * End: + * + * ex: set ts=8 sts=4 sw=4 et: + */