X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=utf8.c;h=75226ca0ee31e1c4285685aec6b554b72256ed38;hb=a58d912c5262ef3032191d4aea207683577527bb;hp=785047e673f8f07155c20b8d70212e3bf1682195;hpb=37607a96501d29b9cd8e4976a682c9ac460e70db;p=p5sagit%2Fp5-mst-13.2.git diff --git a/utf8.c b/utf8.c index 785047e..75226ca 100644 --- a/utf8.c +++ b/utf8.c @@ -50,7 +50,7 @@ Perl_uvuni_to_utf8(pTHX_ U8 *d, UV uv) *d++ = UTF_TO_NATIVE(uv); return d; } -#if defined(EBCDIC) || 1 /* always for testing */ +#if defined(EBCDIC) else { STRLEN len = UNISKIP(uv); U8 *p = d+len-1; @@ -163,14 +163,14 @@ Perl_is_utf8_char(pTHX_ U8 *s) slen = len - 1; s++; - /* The initial value is dubious */ + u &= UTF_START_MASK(len); uv = u; ouv = uv; while (slen--) { if (!UTF8_IS_CONTINUATION(*s)) return 0; uv = UTF8_ACCUMULATE(uv, *s); - if (uv < ouv) + if (uv < ouv) return 0; ouv = uv; s++; @@ -243,7 +243,7 @@ Most code should use utf8_to_uvchr() rather than call this directly. UV Perl_utf8n_to_uvuni(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) { - UV uv = *s, ouv; + UV uv = *s, ouv = 0; STRLEN len = 1; bool dowarn = ckWARN_d(WARN_UTF8); STRLEN expectlen = 0; @@ -400,6 +400,7 @@ malformed: case UTF8_WARN_SHORT: Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d)", curlen, curlen == 1 ? "" : "s", expectlen); + expectlen = curlen; /* distance for caller to skip */ break; case UTF8_WARN_OVERFLOW: Perl_sv_catpvf(aTHX_ sv, "(overflow at 0x%"UVxf", byte 0x%02x)", @@ -428,7 +429,7 @@ malformed: if (PL_op) Perl_warner(aTHX_ WARN_UTF8, - "%s in %s", s, PL_op_desc[PL_op->op_type]); + "%s in %s", s, OP_DESC(PL_op)); else Perl_warner(aTHX_ WARN_UTF8, "%s", s); } @@ -507,7 +508,7 @@ Perl_utf8_length(pTHX_ U8 *s, U8 *e) U8 t = UTF8SKIP(s); if (e - s < t) - Perl_croak(aTHX_ "panic: utf8_length: s=%p (%02X) e=%p l=%d - unaligned end",s,*s,e,t); + Perl_croak(aTHX_ "panic: utf8_length: unaligned end"); s += t; len++; } @@ -796,7 +797,7 @@ Perl_utf16_to_utf8_reversed(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen) /* for now these are all defined (inefficiently) in terms of the utf8 versions */ bool -Perl_is_uni_alnum(pTHX_ U32 c) +Perl_is_uni_alnum(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -804,7 +805,7 @@ Perl_is_uni_alnum(pTHX_ U32 c) } bool -Perl_is_uni_alnumc(pTHX_ U32 c) +Perl_is_uni_alnumc(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -812,7 +813,7 @@ Perl_is_uni_alnumc(pTHX_ U32 c) } bool -Perl_is_uni_idfirst(pTHX_ U32 c) +Perl_is_uni_idfirst(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -820,7 +821,7 @@ Perl_is_uni_idfirst(pTHX_ U32 c) } bool -Perl_is_uni_alpha(pTHX_ U32 c) +Perl_is_uni_alpha(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -828,7 +829,7 @@ Perl_is_uni_alpha(pTHX_ U32 c) } bool -Perl_is_uni_ascii(pTHX_ U32 c) +Perl_is_uni_ascii(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -836,7 +837,7 @@ Perl_is_uni_ascii(pTHX_ U32 c) } bool -Perl_is_uni_space(pTHX_ U32 c) +Perl_is_uni_space(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -844,7 +845,7 @@ Perl_is_uni_space(pTHX_ U32 c) } bool -Perl_is_uni_digit(pTHX_ U32 c) +Perl_is_uni_digit(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -852,7 +853,7 @@ Perl_is_uni_digit(pTHX_ U32 c) } bool -Perl_is_uni_upper(pTHX_ U32 c) +Perl_is_uni_upper(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -860,7 +861,7 @@ Perl_is_uni_upper(pTHX_ U32 c) } bool -Perl_is_uni_lower(pTHX_ U32 c) +Perl_is_uni_lower(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -868,7 +869,7 @@ Perl_is_uni_lower(pTHX_ U32 c) } bool -Perl_is_uni_cntrl(pTHX_ U32 c) +Perl_is_uni_cntrl(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -876,7 +877,7 @@ Perl_is_uni_cntrl(pTHX_ U32 c) } bool -Perl_is_uni_graph(pTHX_ U32 c) +Perl_is_uni_graph(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -884,7 +885,7 @@ Perl_is_uni_graph(pTHX_ U32 c) } bool -Perl_is_uni_print(pTHX_ U32 c) +Perl_is_uni_print(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -892,7 +893,7 @@ Perl_is_uni_print(pTHX_ U32 c) } bool -Perl_is_uni_punct(pTHX_ U32 c) +Perl_is_uni_punct(pTHX_ UV c) { U8 tmpbuf[UTF8_MAXLEN+1]; uvchr_to_utf8(tmpbuf, (UV)c); @@ -900,141 +901,131 @@ Perl_is_uni_punct(pTHX_ U32 c) } bool -Perl_is_uni_xdigit(pTHX_ U32 c) +Perl_is_uni_xdigit(pTHX_ UV c) { - U8 tmpbuf[UTF8_MAXLEN+1]; + U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; uvchr_to_utf8(tmpbuf, (UV)c); return is_utf8_xdigit(tmpbuf); } -U32 -Perl_to_uni_upper(pTHX_ U32 c) +UV +Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp) { - U8 tmpbuf[UTF8_MAXLEN+1]; + U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; uvchr_to_utf8(tmpbuf, (UV)c); - return to_utf8_upper(tmpbuf); + return to_utf8_upper(tmpbuf, p, lenp); } -U32 -Perl_to_uni_title(pTHX_ U32 c) +UV +Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp) { - U8 tmpbuf[UTF8_MAXLEN+1]; + U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; uvchr_to_utf8(tmpbuf, (UV)c); - return to_utf8_title(tmpbuf); + return to_utf8_title(tmpbuf, p, lenp); } -U32 -Perl_to_uni_lower(pTHX_ U32 c) +UV +Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp) { - U8 tmpbuf[UTF8_MAXLEN+1]; + U8 tmpbuf[UTF8_MAXLEN_UCLC+1]; + uvchr_to_utf8(tmpbuf, (UV)c); + return to_utf8_lower(tmpbuf, p, lenp); +} + +UV +Perl_to_uni_fold(pTHX_ UV c, U8* p, STRLEN *lenp) +{ + U8 tmpbuf[UTF8_MAXLEN_FOLD+1]; uvchr_to_utf8(tmpbuf, (UV)c); - return to_utf8_lower(tmpbuf); + return to_utf8_fold(tmpbuf, p, lenp); } /* for now these all assume no locale info available for Unicode > 255 */ bool -Perl_is_uni_alnum_lc(pTHX_ U32 c) +Perl_is_uni_alnum_lc(pTHX_ UV c) { return is_uni_alnum(c); /* XXX no locale support yet */ } bool -Perl_is_uni_alnumc_lc(pTHX_ U32 c) +Perl_is_uni_alnumc_lc(pTHX_ UV c) { return is_uni_alnumc(c); /* XXX no locale support yet */ } bool -Perl_is_uni_idfirst_lc(pTHX_ U32 c) +Perl_is_uni_idfirst_lc(pTHX_ UV c) { return is_uni_idfirst(c); /* XXX no locale support yet */ } bool -Perl_is_uni_alpha_lc(pTHX_ U32 c) +Perl_is_uni_alpha_lc(pTHX_ UV c) { return is_uni_alpha(c); /* XXX no locale support yet */ } bool -Perl_is_uni_ascii_lc(pTHX_ U32 c) +Perl_is_uni_ascii_lc(pTHX_ UV c) { return is_uni_ascii(c); /* XXX no locale support yet */ } bool -Perl_is_uni_space_lc(pTHX_ U32 c) +Perl_is_uni_space_lc(pTHX_ UV c) { return is_uni_space(c); /* XXX no locale support yet */ } bool -Perl_is_uni_digit_lc(pTHX_ U32 c) +Perl_is_uni_digit_lc(pTHX_ UV c) { return is_uni_digit(c); /* XXX no locale support yet */ } bool -Perl_is_uni_upper_lc(pTHX_ U32 c) +Perl_is_uni_upper_lc(pTHX_ UV c) { return is_uni_upper(c); /* XXX no locale support yet */ } bool -Perl_is_uni_lower_lc(pTHX_ U32 c) +Perl_is_uni_lower_lc(pTHX_ UV c) { return is_uni_lower(c); /* XXX no locale support yet */ } bool -Perl_is_uni_cntrl_lc(pTHX_ U32 c) +Perl_is_uni_cntrl_lc(pTHX_ UV c) { return is_uni_cntrl(c); /* XXX no locale support yet */ } bool -Perl_is_uni_graph_lc(pTHX_ U32 c) +Perl_is_uni_graph_lc(pTHX_ UV c) { return is_uni_graph(c); /* XXX no locale support yet */ } bool -Perl_is_uni_print_lc(pTHX_ U32 c) +Perl_is_uni_print_lc(pTHX_ UV c) { return is_uni_print(c); /* XXX no locale support yet */ } bool -Perl_is_uni_punct_lc(pTHX_ U32 c) +Perl_is_uni_punct_lc(pTHX_ UV c) { return is_uni_punct(c); /* XXX no locale support yet */ } bool -Perl_is_uni_xdigit_lc(pTHX_ U32 c) +Perl_is_uni_xdigit_lc(pTHX_ UV c) { return is_uni_xdigit(c); /* XXX no locale support yet */ } -U32 -Perl_to_uni_upper_lc(pTHX_ U32 c) -{ - return to_uni_upper(c); /* XXX no locale support yet */ -} - -U32 -Perl_to_uni_title_lc(pTHX_ U32 c) -{ - return to_uni_title(c); /* XXX no locale support yet */ -} - -U32 -Perl_to_uni_lower_lc(pTHX_ U32 c) -{ - return to_uni_lower(c); /* XXX no locale support yet */ -} - bool Perl_is_utf8_alnum(pTHX_ U8 *p) { @@ -1045,13 +1036,13 @@ Perl_is_utf8_alnum(pTHX_ U8 *p) * descendant of isalnum(3), in other words, it doesn't * contain the '_'. --jhi */ PL_utf8_alnum = swash_init("utf8", "IsWord", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_alnum, p); + return swash_fetch(PL_utf8_alnum, p, TRUE); /* return *p == '_' || is_utf8_alpha(p) || is_utf8_digit(p); */ #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */ if (!PL_utf8_alnum) PL_utf8_alnum = swash_init("utf8", "", sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0); - return swash_fetch(PL_utf8_alnum, p); + return swash_fetch(PL_utf8_alnum, p, TRUE); #endif } @@ -1062,13 +1053,13 @@ Perl_is_utf8_alnumc(pTHX_ U8 *p) return FALSE; if (!PL_utf8_alnum) PL_utf8_alnum = swash_init("utf8", "IsAlnumC", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_alnum, p); + return swash_fetch(PL_utf8_alnum, p, TRUE); /* return is_utf8_alpha(p) || is_utf8_digit(p); */ #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */ if (!PL_utf8_alnum) PL_utf8_alnum = swash_init("utf8", "", sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0); - return swash_fetch(PL_utf8_alnum, p); + return swash_fetch(PL_utf8_alnum, p, TRUE); #endif } @@ -1085,7 +1076,7 @@ Perl_is_utf8_alpha(pTHX_ U8 *p) return FALSE; if (!PL_utf8_alpha) PL_utf8_alpha = swash_init("utf8", "IsAlpha", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_alpha, p); + return swash_fetch(PL_utf8_alpha, p, TRUE); } bool @@ -1095,7 +1086,7 @@ Perl_is_utf8_ascii(pTHX_ U8 *p) return FALSE; if (!PL_utf8_ascii) PL_utf8_ascii = swash_init("utf8", "IsAscii", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_ascii, p); + return swash_fetch(PL_utf8_ascii, p, TRUE); } bool @@ -1105,7 +1096,7 @@ Perl_is_utf8_space(pTHX_ U8 *p) return FALSE; if (!PL_utf8_space) PL_utf8_space = swash_init("utf8", "IsSpacePerl", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_space, p); + return swash_fetch(PL_utf8_space, p, TRUE); } bool @@ -1115,7 +1106,7 @@ Perl_is_utf8_digit(pTHX_ U8 *p) return FALSE; if (!PL_utf8_digit) PL_utf8_digit = swash_init("utf8", "IsDigit", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_digit, p); + return swash_fetch(PL_utf8_digit, p, TRUE); } bool @@ -1125,7 +1116,7 @@ Perl_is_utf8_upper(pTHX_ U8 *p) return FALSE; if (!PL_utf8_upper) PL_utf8_upper = swash_init("utf8", "IsUpper", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_upper, p); + return swash_fetch(PL_utf8_upper, p, TRUE); } bool @@ -1135,7 +1126,7 @@ Perl_is_utf8_lower(pTHX_ U8 *p) return FALSE; if (!PL_utf8_lower) PL_utf8_lower = swash_init("utf8", "IsLower", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_lower, p); + return swash_fetch(PL_utf8_lower, p, TRUE); } bool @@ -1145,7 +1136,7 @@ Perl_is_utf8_cntrl(pTHX_ U8 *p) return FALSE; if (!PL_utf8_cntrl) PL_utf8_cntrl = swash_init("utf8", "IsCntrl", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_cntrl, p); + return swash_fetch(PL_utf8_cntrl, p, TRUE); } bool @@ -1155,7 +1146,7 @@ Perl_is_utf8_graph(pTHX_ U8 *p) return FALSE; if (!PL_utf8_graph) PL_utf8_graph = swash_init("utf8", "IsGraph", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_graph, p); + return swash_fetch(PL_utf8_graph, p, TRUE); } bool @@ -1165,7 +1156,7 @@ Perl_is_utf8_print(pTHX_ U8 *p) return FALSE; if (!PL_utf8_print) PL_utf8_print = swash_init("utf8", "IsPrint", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_print, p); + return swash_fetch(PL_utf8_print, p, TRUE); } bool @@ -1175,7 +1166,7 @@ Perl_is_utf8_punct(pTHX_ U8 *p) return FALSE; if (!PL_utf8_punct) PL_utf8_punct = swash_init("utf8", "IsPunct", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_punct, p); + return swash_fetch(PL_utf8_punct, p, TRUE); } bool @@ -1185,7 +1176,7 @@ Perl_is_utf8_xdigit(pTHX_ U8 *p) return FALSE; if (!PL_utf8_xdigit) PL_utf8_xdigit = swash_init("utf8", "IsXDigit", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_xdigit, p); + return swash_fetch(PL_utf8_xdigit, p, TRUE); } bool @@ -1195,40 +1186,99 @@ Perl_is_utf8_mark(pTHX_ U8 *p) return FALSE; if (!PL_utf8_mark) PL_utf8_mark = swash_init("utf8", "IsM", &PL_sv_undef, 0, 0); - return swash_fetch(PL_utf8_mark, p); + return swash_fetch(PL_utf8_mark, p, TRUE); } +/* +=for apidoc A|UV|to_utf8_case|U8 *p|U8* ustrp|STRLEN *lenp|SV **swash|char *normal|char *special + +The "p" contains the pointer to the UTF-8 string encoding +the character that is being converted. + +The "ustrp" is a pointer to the character buffer to put the +conversion result to. The "lenp" is a pointer to the length +of the result. + +The "swash" is a pointer to the swash to use. + +The "normal" is a string like "ToLower" which means the swash +$utf8::ToLower, which is stored in lib/unicore/To/Lower.pl, +and loaded by SWASHGET, using lib/utf8_heavy.pl. + +The "special" is a string like "utf8::ToSpecLower", which means +the hash %utf8::ToSpecLower, which is stored in the same file, +lib/unicore/To/Lower.pl, and also loaded by SWASHGET. The access +to the hash is by Perl_to_utf8_case(). + +=cut + */ + UV -Perl_to_utf8_upper(pTHX_ U8 *p) +Perl_to_utf8_case(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp, SV **swashp,char *normal, char *special) { UV uv; - if (!PL_utf8_toupper) - PL_utf8_toupper = swash_init("utf8", "ToUpper", &PL_sv_undef, 4, 0); - uv = swash_fetch(PL_utf8_toupper, p); - return uv ? UNI_TO_NATIVE(uv) : utf8_to_uvchr(p,0); + if (!*swashp) + *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0); + uv = swash_fetch(*swashp, p, TRUE); + if (uv) + uv = UNI_TO_NATIVE(uv); + else { + HV *hv; + SV *keysv; + HE *he; + + uv = utf8_to_uvchr(p, 0); + + if ((hv = get_hv(special, FALSE)) && + (keysv = sv_2mortal(Perl_newSVpvf(aTHX_ "%04"UVXf, uv))) && + (he = hv_fetch_ent(hv, keysv, FALSE, 0))) { + SV *val = HeVAL(he); + char *s = SvPV(val, *lenp); + U8 c = *(U8*)s; + if (*lenp > 1 || UNI_IS_INVARIANT(c)) + Copy(s, ustrp, *lenp, U8); + else { + /* something in the 0x80..0xFF range */ + ustrp[0] = UTF8_EIGHT_BIT_HI(c); + ustrp[1] = UTF8_EIGHT_BIT_LO(c); + *lenp = 2; + } + return 0; + } + } + if (lenp) + *lenp = UNISKIP(uv); + uvuni_to_utf8(ustrp, uv); + return uv; } UV -Perl_to_utf8_title(pTHX_ U8 *p) +Perl_to_utf8_upper(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) { - UV uv; + return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, + &PL_utf8_toupper, "ToUpper", "utf8::ToSpecUpper"); +} - if (!PL_utf8_totitle) - PL_utf8_totitle = swash_init("utf8", "ToTitle", &PL_sv_undef, 4, 0); - uv = swash_fetch(PL_utf8_totitle, p); - return uv ? UNI_TO_NATIVE(uv) : utf8_to_uvchr(p,0); +UV +Perl_to_utf8_title(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) +{ + return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, + &PL_utf8_totitle, "ToTitle", "utf8::ToSpecTitle"); } UV -Perl_to_utf8_lower(pTHX_ U8 *p) +Perl_to_utf8_lower(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) { - UV uv; + return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, + &PL_utf8_tolower, "ToLower", "utf8::ToSpecLower"); +} - if (!PL_utf8_tolower) - PL_utf8_tolower = swash_init("utf8", "ToLower", &PL_sv_undef, 4, 0); - uv = swash_fetch(PL_utf8_tolower, p); - return uv ? UNI_TO_NATIVE(uv) : utf8_to_uvchr(p,0); +UV +Perl_to_utf8_fold(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp) +{ + return Perl_to_utf8_case(aTHX_ p, ustrp, lenp, + &PL_utf8_tofold, "ToFold", "utf8::ToSpecFold"); } /* a "swash" is a swatch hash */ @@ -1240,10 +1290,15 @@ Perl_swash_init(pTHX_ char* pkg, char* name, SV *listsv, I32 minbits, I32 none) SV* tokenbufsv = sv_2mortal(NEWSV(0,0)); dSP; HV *stash = gv_stashpvn(pkg, strlen(pkg), FALSE); + SV* errsv_save; if (!gv_fetchmeth(stash, "SWASHNEW", 8, -1)) { /* demand load utf8 */ ENTER; + errsv_save = newSVsv(ERRSV); Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpv(pkg,0), Nullsv); + if (!SvTRUE(ERRSV)) + sv_setsv(ERRSV, errsv_save); + SvREFCNT_dec(errsv_save); LEAVE; } SPAGAIN; @@ -1263,10 +1318,14 @@ Perl_swash_init(pTHX_ char* pkg, char* name, SV *listsv, I32 minbits, I32 none) if (PL_curcop == &PL_compiling) /* XXX ought to be handled by lex_start */ sv_setpv(tokenbufsv, PL_tokenbuf); + errsv_save = newSVsv(ERRSV); if (call_method("SWASHNEW", G_SCALAR)) retval = newSVsv(*PL_stack_sp--); else retval = &PL_sv_undef; + if (!SvTRUE(ERRSV)) + sv_setsv(ERRSV, errsv_save); + SvREFCNT_dec(errsv_save); LEAVE; POPSTACK; if (PL_curcop == &PL_compiling) { @@ -1281,22 +1340,38 @@ Perl_swash_init(pTHX_ char* pkg, char* name, SV *listsv, I32 minbits, I32 none) return retval; } + +/* This API is wrong for special case conversions since we may need to + * return several Unicode characters for a single Unicode character + * (see lib/unicore/SpecCase.txt) The SWASHGET in lib/utf8_heavy.pl is + * the lower-level routine, and it is similarly broken for returning + * multiple values. --jhi */ UV -Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr) +Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr, bool do_utf8) { HV* hv = (HV*)SvRV(sv); - /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ - then the "swatch" is a vec() for al the chars which start - with 0xAA..0xYY - So the key in the hash is length of encoded char -1 - */ - U32 klen = UTF8SKIP(ptr) - 1; - U32 off = ptr[klen]; + U32 klen; + U32 off; STRLEN slen; STRLEN needents; - U8 *tmps; + U8 *tmps = NULL; U32 bit; SV *retval; + U8 tmputf8[2]; + UV c = NATIVE_TO_ASCII(*ptr); + + if (!do_utf8 && !UNI_IS_INVARIANT(c)) { + tmputf8[0] = UTF8_EIGHT_BIT_HI(c); + tmputf8[1] = UTF8_EIGHT_BIT_LO(c); + ptr = tmputf8; + } + /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ + * then the "swatch" is a vec() for al the chars which start + * with 0xAA..0xYY + * So the key in the hash (klen) is length of encoded char -1 + */ + klen = UTF8SKIP(ptr) - 1; + off = ptr[klen]; if (klen == 0) { @@ -1322,9 +1397,9 @@ Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr) * NB: this code assumes that swatches are never modified, once generated! */ - if (hv == PL_last_swash_hv && + if (hv == PL_last_swash_hv && klen == PL_last_swash_klen && - (!klen || memEQ((char *)ptr,(char *)PL_last_swash_key,klen)) ) + (!klen || memEQ((char *)ptr, (char *)PL_last_swash_key, klen)) ) { tmps = PL_last_swash_tmps; slen = PL_last_swash_slen; @@ -1340,6 +1415,7 @@ Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr) Unicode tables, not a native character number. */ UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXLEN, NULL, 0); + SV *errsv_save; ENTER; SAVETMPS; save_re_context(); @@ -1348,13 +1424,18 @@ Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr) EXTEND(SP,3); PUSHs((SV*)sv); /* On EBCDIC & ~(0xA0-1) isn't a useful thing to do */ - PUSHs(sv_2mortal(newSViv((klen) ? (code_point & ~(needents - 1)) : 0))); + PUSHs(sv_2mortal(newSViv((klen) ? + (code_point & ~(needents - 1)) : 0))); PUSHs(sv_2mortal(newSViv(needents))); PUTBACK; + errsv_save = newSVsv(ERRSV); if (call_method("SWASHGET", G_SCALAR)) retval = newSVsv(*PL_stack_sp--); else retval = &PL_sv_undef; + if (!SvTRUE(ERRSV)) + sv_setsv(ERRSV, errsv_save); + SvREFCNT_dec(errsv_save); POPSTACK; FREETMPS; LEAVE; @@ -1444,4 +1525,115 @@ Perl_utf8n_to_uvchr(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags) return UNI_TO_NATIVE(uv); } +/* +=for apidoc A|char *|pv_uni_display|SV *dsv|U8 *spv|STRLEN len|STRLEN pvlim|UV flags + +Build to the scalar dsv a displayable version of the string spv, +length len, the displayable version being at most pvlim bytes long +(if longer, the rest is truncated and "..." will be appended). +The flags argument is currently unused but available for future extensions. +The pointer to the PV of the dsv is returned. + +=cut */ +char * +Perl_pv_uni_display(pTHX_ SV *dsv, U8 *spv, STRLEN len, STRLEN pvlim, UV flags) +{ + int truncated = 0; + char *s, *e; + + sv_setpvn(dsv, "", 0); + for (s = (char *)spv, e = s + len; s < e; s += UTF8SKIP(s)) { + UV u; + if (pvlim && SvCUR(dsv) >= pvlim) { + truncated++; + break; + } + u = utf8_to_uvchr((U8*)s, 0); + Perl_sv_catpvf(aTHX_ dsv, "\\x{%"UVxf"}", u); + } + if (truncated) + sv_catpvn(dsv, "...", 3); + + return SvPVX(dsv); +} + +/* +=for apidoc A|char *|sv_uni_display|SV *dsv|SV *ssv|STRLEN pvlim|UV flags + +Build to the scalar dsv a displayable version of the scalar sv, +he displayable version being at most pvlim bytes long +(if longer, the rest is truncated and "..." will be appended). +The flags argument is currently unused but available for future extensions. +The pointer to the PV of the dsv is returned. + +=cut */ +char * +Perl_sv_uni_display(pTHX_ SV *dsv, SV *ssv, STRLEN pvlim, UV flags) +{ + return Perl_pv_uni_display(aTHX_ dsv, (U8*)SvPVX(ssv), SvCUR(ssv), + pvlim, flags); +} + +/* +=for apidoc A|I32|ibcmp_utf8|const char *s1|bool u1|const char *s2|bool u2|register I32 len + +Return true if the strings s1 and s2 differ case-insensitively, false +if not (if they are equal case-insensitively). If u1 is true, the +string s1 is assumed to be in UTF-8-encoded Unicode. If u2 is true, +the string s2 is assumed to be in UTF-8-encoded Unicode. + +For case-insensitiveness, the "casefolding" of Unicode is used +instead of upper/lowercasing both the characters, see +http://www.unicode.org/unicode/reports/tr21/ (Case Mappings). + +=cut */ +I32 +Perl_ibcmp_utf8(pTHX_ const char *s1, bool u1, register I32 len1, const char *s2, bool u2, register I32 len2) +{ + register U8 *a = (U8*)s1; + register U8 *b = (U8*)s2; + register U8 *ae = b + len1; + register U8 *be = b + len2; + STRLEN la, lb; + UV ca, cb; + STRLEN ulen1, ulen2; + U8 tmpbuf1[UTF8_MAXLEN_FOLD+1]; + U8 tmpbuf2[UTF8_MAXLEN_FOLD+1]; + + while (a < ae && b < be) { + if (u1) { + if (a + UTF8SKIP(a) > ae) + break; + ca = utf8_to_uvchr((U8*)a, &la); + } else { + ca = *a; + la = 1; + } + if (u2) { + if (b + UTF8SKIP(b) > be) + break; + cb = utf8_to_uvchr((U8*)b, &lb); + } else { + cb = *b; + lb = 1; + } + if (ca != cb) { + if (u1) + to_uni_fold(NATIVE_TO_UNI(ca), tmpbuf1, &ulen1); + else + ulen1 = 1; + if (u2) + to_uni_fold(NATIVE_TO_UNI(cb), tmpbuf2, &ulen2); + else + ulen2 = 1; + if (ulen1 != ulen2 + || (ulen1 == 1 && PL_fold[ca] != PL_fold[cb]) + || memNE((char *)tmpbuf1, (char *)tmpbuf2, ulen1)) + return 1; /* mismatch */ + } + a += la; + b += lb; + } + return a == ae && b == be ? 0 : 1; /* 0 match, 1 mismatch */ +}