3 * Copyright (c) 1998-2002, Larry Wall
5 * You may distribute under the terms of either the GNU General Public
6 * License or the Artistic License, as specified in the README file.
11 * 'What a fix!' said Sam. 'That's the one place in all the lands we've ever
12 * heard of that we don't want to see any closer; and that's the one place
13 * we're trying to get to! And that's just where we can't get, nohow.'
15 * 'Well do I understand your speech,' he answered in the same language;
16 * 'yet few strangers do so. Why then do you not speak in the Common Tongue,
17 * as is the custom in the West, if you wish to be answered?'
19 * ...the travellers perceived that the floor was paved with stones of many
20 * hues; branching runes and strange devices intertwined beneath their feet.
24 #define PERL_IN_UTF8_C
28 =head1 Unicode Support
30 =for apidoc A|U8 *|uvuni_to_utf8_flags|U8 *d|UV uv|UV flags
32 Adds the UTF8 representation of the Unicode codepoint C<uv> to the end
33 of the string C<d>; C<d> should be have at least C<UTF8_MAXLEN+1> free
34 bytes available. The return value is the pointer to the byte after the
35 end of the new character. In other words,
37 d = uvuni_to_utf8_flags(d, uv, flags);
41 d = uvuni_to_utf8(d, uv);
43 (which is equivalent to)
45 d = uvuni_to_utf8_flags(d, uv, 0);
47 is the recommended Unicode-aware way of saying
55 Perl_uvuni_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags)
57 if (ckWARN(WARN_UTF8)) {
58 if (UNICODE_IS_SURROGATE(uv) &&
59 !(flags & UNICODE_ALLOW_SURROGATE))
60 Perl_warner(aTHX_ WARN_UTF8, "UTF-16 surrogate 0x%04"UVxf, uv);
62 ((uv >= 0xFDD0 && uv <= 0xFDEF &&
63 !(flags & UNICODE_ALLOW_FDD0))
65 ((uv & 0xFFFF) == 0xFFFE &&
66 !(flags & UNICODE_ALLOW_FFFE))
68 ((uv & 0xFFFF) == 0xFFFF &&
69 !(flags & UNICODE_ALLOW_FFFF))) &&
70 /* UNICODE_ALLOW_SUPER includes
71 * FFFEs and FFFFs beyond 0x10FFFF. */
72 ((uv <= PERL_UNICODE_MAX) ||
73 !(flags & UNICODE_ALLOW_SUPER))
75 Perl_warner(aTHX_ WARN_UTF8,
76 "Unicode character 0x%04"UVxf" is illegal", uv);
78 if (UNI_IS_INVARIANT(uv)) {
79 *d++ = UTF_TO_NATIVE(uv);
84 STRLEN len = UNISKIP(uv);
87 *p-- = UTF_TO_NATIVE((uv & UTF_CONTINUATION_MASK) | UTF_CONTINUATION_MARK);
88 uv >>= UTF_ACCUMULATION_SHIFT;
90 *p = UTF_TO_NATIVE((uv & UTF_START_MASK(len)) | UTF_START_MARK(len));
93 #else /* Non loop style */
95 *d++ = (( uv >> 6) | 0xc0);
96 *d++ = (( uv & 0x3f) | 0x80);
100 *d++ = (( uv >> 12) | 0xe0);
101 *d++ = (((uv >> 6) & 0x3f) | 0x80);
102 *d++ = (( uv & 0x3f) | 0x80);
106 *d++ = (( uv >> 18) | 0xf0);
107 *d++ = (((uv >> 12) & 0x3f) | 0x80);
108 *d++ = (((uv >> 6) & 0x3f) | 0x80);
109 *d++ = (( uv & 0x3f) | 0x80);
112 if (uv < 0x4000000) {
113 *d++ = (( uv >> 24) | 0xf8);
114 *d++ = (((uv >> 18) & 0x3f) | 0x80);
115 *d++ = (((uv >> 12) & 0x3f) | 0x80);
116 *d++ = (((uv >> 6) & 0x3f) | 0x80);
117 *d++ = (( uv & 0x3f) | 0x80);
120 if (uv < 0x80000000) {
121 *d++ = (( uv >> 30) | 0xfc);
122 *d++ = (((uv >> 24) & 0x3f) | 0x80);
123 *d++ = (((uv >> 18) & 0x3f) | 0x80);
124 *d++ = (((uv >> 12) & 0x3f) | 0x80);
125 *d++ = (((uv >> 6) & 0x3f) | 0x80);
126 *d++ = (( uv & 0x3f) | 0x80);
130 if (uv < UTF8_QUAD_MAX)
133 *d++ = 0xfe; /* Can't match U+FEFF! */
134 *d++ = (((uv >> 30) & 0x3f) | 0x80);
135 *d++ = (((uv >> 24) & 0x3f) | 0x80);
136 *d++ = (((uv >> 18) & 0x3f) | 0x80);
137 *d++ = (((uv >> 12) & 0x3f) | 0x80);
138 *d++ = (((uv >> 6) & 0x3f) | 0x80);
139 *d++ = (( uv & 0x3f) | 0x80);
144 *d++ = 0xff; /* Can't match U+FFFE! */
145 *d++ = 0x80; /* 6 Reserved bits */
146 *d++ = (((uv >> 60) & 0x0f) | 0x80); /* 2 Reserved bits */
147 *d++ = (((uv >> 54) & 0x3f) | 0x80);
148 *d++ = (((uv >> 48) & 0x3f) | 0x80);
149 *d++ = (((uv >> 42) & 0x3f) | 0x80);
150 *d++ = (((uv >> 36) & 0x3f) | 0x80);
151 *d++ = (((uv >> 30) & 0x3f) | 0x80);
152 *d++ = (((uv >> 24) & 0x3f) | 0x80);
153 *d++ = (((uv >> 18) & 0x3f) | 0x80);
154 *d++ = (((uv >> 12) & 0x3f) | 0x80);
155 *d++ = (((uv >> 6) & 0x3f) | 0x80);
156 *d++ = (( uv & 0x3f) | 0x80);
160 #endif /* Loop style */
164 Perl_uvuni_to_utf8(pTHX_ U8 *d, UV uv)
166 return Perl_uvuni_to_utf8_flags(aTHX_ d, uv, 0);
171 =for apidoc A|STRLEN|is_utf8_char|U8 *s
173 Tests if some arbitrary number of bytes begins in a valid UTF-8
174 character. Note that an INVARIANT (i.e. ASCII) character is a valid UTF-8 character.
175 The actual number of bytes in the UTF-8 character will be returned if
176 it is valid, otherwise 0.
181 Perl_is_utf8_char(pTHX_ U8 *s)
187 if (UTF8_IS_INVARIANT(u))
190 if (!UTF8_IS_START(u))
195 if (len < 2 || !UTF8_IS_CONTINUATION(s[1]))
200 u &= UTF_START_MASK(len);
204 if (!UTF8_IS_CONTINUATION(*s))
206 uv = UTF8_ACCUMULATE(uv, *s);
213 if (UNISKIP(uv) < len)
220 =for apidoc A|bool|is_utf8_string|U8 *s|STRLEN len
222 Returns true if first C<len> bytes of the given string form a valid UTF8
223 string, false otherwise. Note that 'a valid UTF8 string' does not mean
224 'a string that contains UTF8' because a valid ASCII string is a valid
231 Perl_is_utf8_string(pTHX_ U8 *s, STRLEN len)
238 len = strlen((char *)s);
254 =for apidoc A|UV|utf8n_to_uvuni|U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags
256 Bottom level UTF-8 decode routine.
257 Returns the unicode code point value of the first character in the string C<s>
258 which is assumed to be in UTF8 encoding and no longer than C<curlen>;
259 C<retlen> will be set to the length, in bytes, of that character.
261 If C<s> does not point to a well-formed UTF8 character, the behaviour
262 is dependent on the value of C<flags>: if it contains UTF8_CHECK_ONLY,
263 it is assumed that the caller will raise a warning, and this function
264 will silently just set C<retlen> to C<-1> and return zero. If the
265 C<flags> does not contain UTF8_CHECK_ONLY, warnings about
266 malformations will be given, C<retlen> will be set to the expected
267 length of the UTF-8 character in bytes, and zero will be returned.
269 The C<flags> can also contain various flags to allow deviations from
270 the strict UTF-8 encoding (see F<utf8.h>).
272 Most code should use utf8_to_uvchr() rather than call this directly.
278 Perl_utf8n_to_uvuni(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags)
283 bool dowarn = ckWARN_d(WARN_UTF8);
285 STRLEN expectlen = 0;
288 /* This list is a superset of the UTF8_ALLOW_XXX. */
290 #define UTF8_WARN_EMPTY 1
291 #define UTF8_WARN_CONTINUATION 2
292 #define UTF8_WARN_NON_CONTINUATION 3
293 #define UTF8_WARN_FE_FF 4
294 #define UTF8_WARN_SHORT 5
295 #define UTF8_WARN_OVERFLOW 6
296 #define UTF8_WARN_SURROGATE 7
297 #define UTF8_WARN_BOM 8
298 #define UTF8_WARN_LONG 9
299 #define UTF8_WARN_FFFF 10
302 !(flags & UTF8_ALLOW_EMPTY)) {
303 warning = UTF8_WARN_EMPTY;
307 if (UTF8_IS_INVARIANT(uv)) {
310 return (UV) (NATIVE_TO_UTF(*s));
313 if (UTF8_IS_CONTINUATION(uv) &&
314 !(flags & UTF8_ALLOW_CONTINUATION)) {
315 warning = UTF8_WARN_CONTINUATION;
319 if (UTF8_IS_START(uv) && curlen > 1 && !UTF8_IS_CONTINUATION(s[1]) &&
320 !(flags & UTF8_ALLOW_NON_CONTINUATION)) {
321 warning = UTF8_WARN_NON_CONTINUATION;
326 uv = NATIVE_TO_UTF(uv);
328 if ((uv == 0xfe || uv == 0xff) &&
329 !(flags & UTF8_ALLOW_FE_FF)) {
330 warning = UTF8_WARN_FE_FF;
335 if (!(uv & 0x20)) { len = 2; uv &= 0x1f; }
336 else if (!(uv & 0x10)) { len = 3; uv &= 0x0f; }
337 else if (!(uv & 0x08)) { len = 4; uv &= 0x07; }
338 else if (!(uv & 0x04)) { len = 5; uv &= 0x03; }
340 else if (!(uv & 0x02)) { len = 6; uv &= 0x01; }
341 else { len = 7; uv &= 0x01; }
343 else if (!(uv & 0x02)) { len = 6; uv &= 0x01; }
344 else if (!(uv & 0x01)) { len = 7; uv = 0; }
345 else { len = 13; uv = 0; } /* whoa! */
353 if ((curlen < expectlen) &&
354 !(flags & UTF8_ALLOW_SHORT)) {
355 warning = UTF8_WARN_SHORT;
364 if (!UTF8_IS_CONTINUATION(*s) &&
365 !(flags & UTF8_ALLOW_NON_CONTINUATION)) {
367 warning = UTF8_WARN_NON_CONTINUATION;
371 uv = UTF8_ACCUMULATE(uv, *s);
373 /* These cannot be allowed. */
375 if (!(flags & UTF8_ALLOW_LONG)) {
376 warning = UTF8_WARN_LONG;
380 else { /* uv < ouv */
381 /* This cannot be allowed. */
382 warning = UTF8_WARN_OVERFLOW;
390 if (UNICODE_IS_SURROGATE(uv) &&
391 !(flags & UTF8_ALLOW_SURROGATE)) {
392 warning = UTF8_WARN_SURROGATE;
394 } else if (UNICODE_IS_BYTE_ORDER_MARK(uv) &&
395 !(flags & UTF8_ALLOW_BOM)) {
396 warning = UTF8_WARN_BOM;
398 } else if ((expectlen > UNISKIP(uv)) &&
399 !(flags & UTF8_ALLOW_LONG)) {
400 warning = UTF8_WARN_LONG;
402 } else if (UNICODE_IS_ILLEGAL(uv) &&
403 !(flags & UTF8_ALLOW_FFFF)) {
404 warning = UTF8_WARN_FFFF;
412 if (flags & UTF8_CHECK_ONLY) {
419 SV* sv = sv_2mortal(newSVpv("Malformed UTF-8 character ", 0));
422 case 0: /* Intentionally empty. */ break;
423 case UTF8_WARN_EMPTY:
424 Perl_sv_catpvf(aTHX_ sv, "(empty string)");
426 case UTF8_WARN_CONTINUATION:
427 Perl_sv_catpvf(aTHX_ sv, "(unexpected continuation byte 0x%02"UVxf", with no preceding start byte)", uv);
429 case UTF8_WARN_NON_CONTINUATION:
431 Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", immediately after start byte 0x%02"UVxf")",
432 (UV)s[1], startbyte);
434 Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", %d byte%s after start byte 0x%02"UVxf", expected %d bytes)",
435 (UV)s[1], s - s0, s - s0 > 1 ? "s" : "", startbyte, expectlen);
438 case UTF8_WARN_FE_FF:
439 Perl_sv_catpvf(aTHX_ sv, "(byte 0x%02"UVxf")", uv);
441 case UTF8_WARN_SHORT:
442 Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")",
443 curlen, curlen == 1 ? "" : "s", expectlen, startbyte);
444 expectlen = curlen; /* distance for caller to skip */
446 case UTF8_WARN_OVERFLOW:
447 Perl_sv_catpvf(aTHX_ sv, "(overflow at 0x%"UVxf", byte 0x%02x, after start byte 0x%02"UVxf")",
450 case UTF8_WARN_SURROGATE:
451 Perl_sv_catpvf(aTHX_ sv, "(UTF-16 surrogate 0x%04"UVxf")", uv);
454 Perl_sv_catpvf(aTHX_ sv, "(byte order mark 0x%04"UVxf")", uv);
457 Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")",
458 expectlen, expectlen == 1 ? "": "s", UNISKIP(uv), startbyte);
461 Perl_sv_catpvf(aTHX_ sv, "(character 0x%04"UVxf")", uv);
464 Perl_sv_catpvf(aTHX_ sv, "(unknown reason)");
472 Perl_warner(aTHX_ WARN_UTF8,
473 "%s in %s", s, OP_DESC(PL_op));
475 Perl_warner(aTHX_ WARN_UTF8, "%s", s);
480 *retlen = expectlen ? expectlen : len;
486 =for apidoc A|UV|utf8_to_uvchr|U8 *s|STRLEN *retlen
488 Returns the native character value of the first character in the string C<s>
489 which is assumed to be in UTF8 encoding; C<retlen> will be set to the
490 length, in bytes, of that character.
492 If C<s> does not point to a well-formed UTF8 character, zero is
493 returned and retlen is set, if possible, to -1.
499 Perl_utf8_to_uvchr(pTHX_ U8 *s, STRLEN *retlen)
501 return Perl_utf8n_to_uvchr(aTHX_ s, UTF8_MAXLEN, retlen, 0);
505 =for apidoc A|UV|utf8_to_uvuni|U8 *s|STRLEN *retlen
507 Returns the Unicode code point of the first character in the string C<s>
508 which is assumed to be in UTF8 encoding; C<retlen> will be set to the
509 length, in bytes, of that character.
511 This function should only be used when returned UV is considered
512 an index into the Unicode semantic tables (e.g. swashes).
514 If C<s> does not point to a well-formed UTF8 character, zero is
515 returned and retlen is set, if possible, to -1.
521 Perl_utf8_to_uvuni(pTHX_ U8 *s, STRLEN *retlen)
523 /* Call the low level routine asking for checks */
524 return Perl_utf8n_to_uvuni(aTHX_ s, UTF8_MAXLEN, retlen, 0);
528 =for apidoc A|STRLEN|utf8_length|U8 *s|U8 *e
530 Return the length of the UTF-8 char encoded string C<s> in characters.
531 Stops at C<e> (inclusive). If C<e E<lt> s> or if the scan would end
532 up past C<e>, croaks.
538 Perl_utf8_length(pTHX_ U8 *s, U8 *e)
542 /* Note: cannot use UTF8_IS_...() too eagerly here since e.g.
543 * the bitops (especially ~) can create illegal UTF-8.
544 * In other words: in Perl UTF-8 is not just for Unicode. */
547 Perl_croak(aTHX_ "panic: utf8_length: unexpected end");
552 Perl_croak(aTHX_ "panic: utf8_length: unaligned end");
561 =for apidoc A|IV|utf8_distance|U8 *a|U8 *b
563 Returns the number of UTF8 characters between the UTF-8 pointers C<a>
566 WARNING: use only if you *know* that the pointers point inside the
573 Perl_utf8_distance(pTHX_ U8 *a, U8 *b)
577 /* Note: cannot use UTF8_IS_...() too eagerly here since e.g.
578 * the bitops (especially ~) can create illegal UTF-8.
579 * In other words: in Perl UTF-8 is not just for Unicode. */
586 Perl_croak(aTHX_ "panic: utf8_distance: unaligned end");
596 Perl_croak(aTHX_ "panic: utf8_distance: unaligned end");
606 =for apidoc A|U8 *|utf8_hop|U8 *s|I32 off
608 Return the UTF-8 pointer C<s> displaced by C<off> characters, either
611 WARNING: do not use the following unless you *know* C<off> is within
612 the UTF-8 data pointed to by C<s> *and* that on entry C<s> is aligned
613 on the first byte of character or just after the last byte of a character.
619 Perl_utf8_hop(pTHX_ U8 *s, I32 off)
621 /* Note: cannot use UTF8_IS_...() too eagerly here since e.g
622 * the bitops (especially ~) can create illegal UTF-8.
623 * In other words: in Perl UTF-8 is not just for Unicode. */
632 while (UTF8_IS_CONTINUATION(*s))
640 =for apidoc A|U8 *|utf8_to_bytes|U8 *s|STRLEN *len
642 Converts a string C<s> of length C<len> from UTF8 into byte encoding.
643 Unlike C<bytes_to_utf8>, this over-writes the original string, and
644 updates len to contain the new length.
645 Returns zero on failure, setting C<len> to -1.
651 Perl_utf8_to_bytes(pTHX_ U8 *s, STRLEN *len)
657 /* ensure valid UTF8 and chars < 256 before updating string */
658 for (send = s + *len; s < send; ) {
661 if (!UTF8_IS_INVARIANT(c) &&
662 (!UTF8_IS_DOWNGRADEABLE_START(c) || (s >= send)
663 || !(c = *s++) || !UTF8_IS_CONTINUATION(c))) {
672 *d++ = (U8)utf8_to_uvchr(s, &ulen);
681 =for apidoc A|U8 *|bytes_from_utf8|U8 *s|STRLEN *len|bool *is_utf8
683 Converts a string C<s> of length C<len> from UTF8 into byte encoding.
684 Unlike <utf8_to_bytes> but like C<bytes_to_utf8>, returns a pointer to
685 the newly-created string, and updates C<len> to contain the new
686 length. Returns the original string if no conversion occurs, C<len>
687 is unchanged. Do nothing if C<is_utf8> points to 0. Sets C<is_utf8> to
688 0 if C<s> is converted or contains all 7bit characters.
694 Perl_bytes_from_utf8(pTHX_ U8 *s, STRLEN *len, bool *is_utf8)
704 /* ensure valid UTF8 and chars < 256 before converting string */
705 for (send = s + *len; s < send;) {
707 if (!UTF8_IS_INVARIANT(c)) {
708 if (UTF8_IS_DOWNGRADEABLE_START(c) && s < send &&
709 (c = *s++) && UTF8_IS_CONTINUATION(c))
718 Newz(801, d, (*len) - count + 1, U8);
719 s = start; start = d;
722 if (!UTF8_IS_INVARIANT(c)) {
723 /* Then it is two-byte encoded */
724 c = UTF8_ACCUMULATE(NATIVE_TO_UTF(c), *s++);
725 c = ASCII_TO_NATIVE(c);
735 =for apidoc A|U8 *|bytes_to_utf8|U8 *s|STRLEN *len
737 Converts a string C<s> of length C<len> from ASCII into UTF8 encoding.
738 Returns a pointer to the newly-created string, and sets C<len> to
739 reflect the new length.
745 Perl_bytes_to_utf8(pTHX_ U8 *s, STRLEN *len)
752 Newz(801, d, (*len) * 2 + 1, U8);
756 UV uv = NATIVE_TO_ASCII(*s++);
757 if (UNI_IS_INVARIANT(uv))
758 *d++ = UTF_TO_NATIVE(uv);
760 *d++ = UTF8_EIGHT_BIT_HI(uv);
761 *d++ = UTF8_EIGHT_BIT_LO(uv);
770 * Convert native (big-endian) or reversed (little-endian) UTF-16 to UTF-8.
772 * Destination must be pre-extended to 3/2 source. Do not use in-place.
773 * We optimize for native, for obvious reasons. */
776 Perl_utf16_to_utf8(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
782 Perl_croak(aTHX_ "panic: utf16_to_utf8: odd bytelen");
787 UV uv = (p[0] << 8) + p[1]; /* UTF-16BE */
794 *d++ = (( uv >> 6) | 0xc0);
795 *d++ = (( uv & 0x3f) | 0x80);
798 if (uv >= 0xd800 && uv < 0xdbff) { /* surrogates */
800 if (low < 0xdc00 || low >= 0xdfff)
801 Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
802 uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000;
805 *d++ = (( uv >> 12) | 0xe0);
806 *d++ = (((uv >> 6) & 0x3f) | 0x80);
807 *d++ = (( uv & 0x3f) | 0x80);
811 *d++ = (( uv >> 18) | 0xf0);
812 *d++ = (((uv >> 12) & 0x3f) | 0x80);
813 *d++ = (((uv >> 6) & 0x3f) | 0x80);
814 *d++ = (( uv & 0x3f) | 0x80);
818 *newlen = d - dstart;
822 /* Note: this one is slightly destructive of the source. */
825 Perl_utf16_to_utf8_reversed(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
828 U8* send = s + bytelen;
835 return utf16_to_utf8(p, d, bytelen, newlen);
838 /* for now these are all defined (inefficiently) in terms of the utf8 versions */
841 Perl_is_uni_alnum(pTHX_ UV c)
843 U8 tmpbuf[UTF8_MAXLEN+1];
844 uvchr_to_utf8(tmpbuf, c);
845 return is_utf8_alnum(tmpbuf);
849 Perl_is_uni_alnumc(pTHX_ UV c)
851 U8 tmpbuf[UTF8_MAXLEN+1];
852 uvchr_to_utf8(tmpbuf, c);
853 return is_utf8_alnumc(tmpbuf);
857 Perl_is_uni_idfirst(pTHX_ UV c)
859 U8 tmpbuf[UTF8_MAXLEN+1];
860 uvchr_to_utf8(tmpbuf, c);
861 return is_utf8_idfirst(tmpbuf);
865 Perl_is_uni_alpha(pTHX_ UV c)
867 U8 tmpbuf[UTF8_MAXLEN+1];
868 uvchr_to_utf8(tmpbuf, c);
869 return is_utf8_alpha(tmpbuf);
873 Perl_is_uni_ascii(pTHX_ UV c)
875 U8 tmpbuf[UTF8_MAXLEN+1];
876 uvchr_to_utf8(tmpbuf, c);
877 return is_utf8_ascii(tmpbuf);
881 Perl_is_uni_space(pTHX_ UV c)
883 U8 tmpbuf[UTF8_MAXLEN+1];
884 uvchr_to_utf8(tmpbuf, c);
885 return is_utf8_space(tmpbuf);
889 Perl_is_uni_digit(pTHX_ UV c)
891 U8 tmpbuf[UTF8_MAXLEN+1];
892 uvchr_to_utf8(tmpbuf, c);
893 return is_utf8_digit(tmpbuf);
897 Perl_is_uni_upper(pTHX_ UV c)
899 U8 tmpbuf[UTF8_MAXLEN+1];
900 uvchr_to_utf8(tmpbuf, c);
901 return is_utf8_upper(tmpbuf);
905 Perl_is_uni_lower(pTHX_ UV c)
907 U8 tmpbuf[UTF8_MAXLEN+1];
908 uvchr_to_utf8(tmpbuf, c);
909 return is_utf8_lower(tmpbuf);
913 Perl_is_uni_cntrl(pTHX_ UV c)
915 U8 tmpbuf[UTF8_MAXLEN+1];
916 uvchr_to_utf8(tmpbuf, c);
917 return is_utf8_cntrl(tmpbuf);
921 Perl_is_uni_graph(pTHX_ UV c)
923 U8 tmpbuf[UTF8_MAXLEN+1];
924 uvchr_to_utf8(tmpbuf, c);
925 return is_utf8_graph(tmpbuf);
929 Perl_is_uni_print(pTHX_ UV c)
931 U8 tmpbuf[UTF8_MAXLEN+1];
932 uvchr_to_utf8(tmpbuf, c);
933 return is_utf8_print(tmpbuf);
937 Perl_is_uni_punct(pTHX_ UV c)
939 U8 tmpbuf[UTF8_MAXLEN+1];
940 uvchr_to_utf8(tmpbuf, c);
941 return is_utf8_punct(tmpbuf);
945 Perl_is_uni_xdigit(pTHX_ UV c)
947 U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
948 uvchr_to_utf8(tmpbuf, c);
949 return is_utf8_xdigit(tmpbuf);
953 Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp)
955 U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
956 uvchr_to_utf8(tmpbuf, c);
957 return to_utf8_upper(tmpbuf, p, lenp);
961 Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp)
963 U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
964 uvchr_to_utf8(tmpbuf, c);
965 return to_utf8_title(tmpbuf, p, lenp);
969 Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp)
971 U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
972 uvchr_to_utf8(tmpbuf, c);
973 return to_utf8_lower(tmpbuf, p, lenp);
977 Perl_to_uni_fold(pTHX_ UV c, U8* p, STRLEN *lenp)
979 U8 tmpbuf[UTF8_MAXLEN_FOLD+1];
980 uvchr_to_utf8(tmpbuf, c);
981 return to_utf8_fold(tmpbuf, p, lenp);
984 /* for now these all assume no locale info available for Unicode > 255 */
987 Perl_is_uni_alnum_lc(pTHX_ UV c)
989 return is_uni_alnum(c); /* XXX no locale support yet */
993 Perl_is_uni_alnumc_lc(pTHX_ UV c)
995 return is_uni_alnumc(c); /* XXX no locale support yet */
999 Perl_is_uni_idfirst_lc(pTHX_ UV c)
1001 return is_uni_idfirst(c); /* XXX no locale support yet */
1005 Perl_is_uni_alpha_lc(pTHX_ UV c)
1007 return is_uni_alpha(c); /* XXX no locale support yet */
1011 Perl_is_uni_ascii_lc(pTHX_ UV c)
1013 return is_uni_ascii(c); /* XXX no locale support yet */
1017 Perl_is_uni_space_lc(pTHX_ UV c)
1019 return is_uni_space(c); /* XXX no locale support yet */
1023 Perl_is_uni_digit_lc(pTHX_ UV c)
1025 return is_uni_digit(c); /* XXX no locale support yet */
1029 Perl_is_uni_upper_lc(pTHX_ UV c)
1031 return is_uni_upper(c); /* XXX no locale support yet */
1035 Perl_is_uni_lower_lc(pTHX_ UV c)
1037 return is_uni_lower(c); /* XXX no locale support yet */
1041 Perl_is_uni_cntrl_lc(pTHX_ UV c)
1043 return is_uni_cntrl(c); /* XXX no locale support yet */
1047 Perl_is_uni_graph_lc(pTHX_ UV c)
1049 return is_uni_graph(c); /* XXX no locale support yet */
1053 Perl_is_uni_print_lc(pTHX_ UV c)
1055 return is_uni_print(c); /* XXX no locale support yet */
1059 Perl_is_uni_punct_lc(pTHX_ UV c)
1061 return is_uni_punct(c); /* XXX no locale support yet */
1065 Perl_is_uni_xdigit_lc(pTHX_ UV c)
1067 return is_uni_xdigit(c); /* XXX no locale support yet */
1071 Perl_to_uni_upper_lc(pTHX_ U32 c)
1073 /* XXX returns only the first character -- do not use XXX */
1074 /* XXX no locale support yet */
1076 U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
1077 return (U32)to_uni_upper(c, tmpbuf, &len);
1081 Perl_to_uni_title_lc(pTHX_ U32 c)
1083 /* XXX returns only the first character XXX -- do not use XXX */
1084 /* XXX no locale support yet */
1086 U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
1087 return (U32)to_uni_title(c, tmpbuf, &len);
1091 Perl_to_uni_lower_lc(pTHX_ U32 c)
1093 /* XXX returns only the first character -- do not use XXX */
1094 /* XXX no locale support yet */
1096 U8 tmpbuf[UTF8_MAXLEN_UCLC+1];
1097 return (U32)to_uni_lower(c, tmpbuf, &len);
1101 Perl_is_utf8_alnum(pTHX_ U8 *p)
1103 if (!is_utf8_char(p))
1106 /* NOTE: "IsWord", not "IsAlnum", since Alnum is a true
1107 * descendant of isalnum(3), in other words, it doesn't
1108 * contain the '_'. --jhi */
1109 PL_utf8_alnum = swash_init("utf8", "IsWord", &PL_sv_undef, 0, 0);
1110 return swash_fetch(PL_utf8_alnum, p, TRUE);
1111 /* return *p == '_' || is_utf8_alpha(p) || is_utf8_digit(p); */
1112 #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */
1114 PL_utf8_alnum = swash_init("utf8", "",
1115 sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0);
1116 return swash_fetch(PL_utf8_alnum, p, TRUE);
1121 Perl_is_utf8_alnumc(pTHX_ U8 *p)
1123 if (!is_utf8_char(p))
1126 PL_utf8_alnum = swash_init("utf8", "IsAlnumC", &PL_sv_undef, 0, 0);
1127 return swash_fetch(PL_utf8_alnum, p, TRUE);
1128 /* return is_utf8_alpha(p) || is_utf8_digit(p); */
1129 #ifdef SURPRISINGLY_SLOWER /* probably because alpha is usually true */
1131 PL_utf8_alnum = swash_init("utf8", "",
1132 sv_2mortal(newSVpv("+utf8::IsAlpha\n+utf8::IsDigit\n005F\n",0)), 0, 0);
1133 return swash_fetch(PL_utf8_alnum, p, TRUE);
1138 Perl_is_utf8_idfirst(pTHX_ U8 *p)
1140 return *p == '_' || is_utf8_alpha(p);
1144 Perl_is_utf8_alpha(pTHX_ U8 *p)
1146 if (!is_utf8_char(p))
1149 PL_utf8_alpha = swash_init("utf8", "IsAlpha", &PL_sv_undef, 0, 0);
1150 return swash_fetch(PL_utf8_alpha, p, TRUE);
1154 Perl_is_utf8_ascii(pTHX_ U8 *p)
1156 if (!is_utf8_char(p))
1159 PL_utf8_ascii = swash_init("utf8", "IsAscii", &PL_sv_undef, 0, 0);
1160 return swash_fetch(PL_utf8_ascii, p, TRUE);
1164 Perl_is_utf8_space(pTHX_ U8 *p)
1166 if (!is_utf8_char(p))
1169 PL_utf8_space = swash_init("utf8", "IsSpacePerl", &PL_sv_undef, 0, 0);
1170 return swash_fetch(PL_utf8_space, p, TRUE);
1174 Perl_is_utf8_digit(pTHX_ U8 *p)
1176 if (!is_utf8_char(p))
1179 PL_utf8_digit = swash_init("utf8", "IsDigit", &PL_sv_undef, 0, 0);
1180 return swash_fetch(PL_utf8_digit, p, TRUE);
1184 Perl_is_utf8_upper(pTHX_ U8 *p)
1186 if (!is_utf8_char(p))
1189 PL_utf8_upper = swash_init("utf8", "IsUpper", &PL_sv_undef, 0, 0);
1190 return swash_fetch(PL_utf8_upper, p, TRUE);
1194 Perl_is_utf8_lower(pTHX_ U8 *p)
1196 if (!is_utf8_char(p))
1199 PL_utf8_lower = swash_init("utf8", "IsLower", &PL_sv_undef, 0, 0);
1200 return swash_fetch(PL_utf8_lower, p, TRUE);
1204 Perl_is_utf8_cntrl(pTHX_ U8 *p)
1206 if (!is_utf8_char(p))
1209 PL_utf8_cntrl = swash_init("utf8", "IsCntrl", &PL_sv_undef, 0, 0);
1210 return swash_fetch(PL_utf8_cntrl, p, TRUE);
1214 Perl_is_utf8_graph(pTHX_ U8 *p)
1216 if (!is_utf8_char(p))
1219 PL_utf8_graph = swash_init("utf8", "IsGraph", &PL_sv_undef, 0, 0);
1220 return swash_fetch(PL_utf8_graph, p, TRUE);
1224 Perl_is_utf8_print(pTHX_ U8 *p)
1226 if (!is_utf8_char(p))
1229 PL_utf8_print = swash_init("utf8", "IsPrint", &PL_sv_undef, 0, 0);
1230 return swash_fetch(PL_utf8_print, p, TRUE);
1234 Perl_is_utf8_punct(pTHX_ U8 *p)
1236 if (!is_utf8_char(p))
1239 PL_utf8_punct = swash_init("utf8", "IsPunct", &PL_sv_undef, 0, 0);
1240 return swash_fetch(PL_utf8_punct, p, TRUE);
1244 Perl_is_utf8_xdigit(pTHX_ U8 *p)
1246 if (!is_utf8_char(p))
1248 if (!PL_utf8_xdigit)
1249 PL_utf8_xdigit = swash_init("utf8", "IsXDigit", &PL_sv_undef, 0, 0);
1250 return swash_fetch(PL_utf8_xdigit, p, TRUE);
1254 Perl_is_utf8_mark(pTHX_ U8 *p)
1256 if (!is_utf8_char(p))
1259 PL_utf8_mark = swash_init("utf8", "IsM", &PL_sv_undef, 0, 0);
1260 return swash_fetch(PL_utf8_mark, p, TRUE);
1264 =for apidoc A|UV|to_utf8_case|U8 *p|U8* ustrp|STRLEN *lenp|SV **swash|char *normal|char *special
1266 The "p" contains the pointer to the UTF-8 string encoding
1267 the character that is being converted.
1269 The "ustrp" is a pointer to the character buffer to put the
1270 conversion result to. The "lenp" is a pointer to the length
1273 The "swashp" is a pointer to the swash to use.
1275 Both the special and normal mappings are stored lib/unicore/To/Foo.pl,
1276 and loaded by SWASHGET, using lib/utf8_heavy.pl. The special (usually,
1277 but not always, a multicharacter mapping), is tried first.
1279 The "special" is a string like "utf8::ToSpecLower", which means the
1280 hash %utf8::ToSpecLower. The access to the hash is through
1281 Perl_to_utf8_case().
1283 The "normal" is a string like "ToLower" which means the swash
1289 Perl_to_utf8_case(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp, SV **swashp, char *normal, char *special)
1292 U8 tmpbuf[UTF8_MAXLEN_FOLD+1];
1295 uv0 = utf8_to_uvchr(p, 0);
1296 /* The NATIVE_TO_UNI() and UNI_TO_NATIVE() mappings
1297 * are necessary in EBCDIC, they are redundant no-ops
1298 * in ASCII-ish platforms, and hopefully optimized away. */
1299 uv1 = NATIVE_TO_UNI(uv0);
1300 uvuni_to_utf8(tmpbuf, uv1);
1302 if (!*swashp) /* load on-demand */
1303 *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0);
1306 /* It might be "special" (sometimes, but not always,
1307 * a multicharacter mapping) */
1313 if ((hv = get_hv(special, FALSE)) &&
1314 (keysv = sv_2mortal(Perl_newSVpvf(aTHX_ "%04"UVXf, uv1))) &&
1315 (he = hv_fetch_ent(hv, keysv, FALSE, 0)) &&
1316 (val = HeVAL(he))) {
1321 len = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s)) - ustrp;
1324 /* If we have EBCDIC we need to remap the characters
1325 * since any characters in the low 256 are Unicode
1326 * code points, not EBCDIC. */
1327 U8 *t = (U8*)s, *tend = t + len, *d;
1334 UV c = utf8_to_uvchr(t, &tlen);
1336 d = uvchr_to_utf8(d, UNI_TO_NATIVE(c));
1345 d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t));
1350 Copy(tmpbuf, ustrp, len, U8);
1352 Copy(s, ustrp, len, U8);
1358 if (!len && *swashp) {
1359 UV uv2 = swash_fetch(*swashp, tmpbuf, TRUE);
1362 /* It was "normal" (a single character mapping). */
1363 UV uv3 = UNI_TO_NATIVE(uv2);
1365 len = uvchr_to_utf8(ustrp, uv3) - ustrp;
1369 if (!len) /* Neither: just copy. */
1370 len = uvchr_to_utf8(ustrp, uv0) - ustrp;
1375 return len ? utf8_to_uvchr(ustrp, 0) : 0;
1379 =for apidoc A|UV|to_utf8_upper|U8 *p|U8 *ustrp|STRLEN *lenp
1381 Convert the UTF-8 encoded character at p to its uppercase version and
1382 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1383 that the ustrp needs to be at least UTF8_MAXLEN_UCLC+1 bytes since the
1384 uppercase version may be longer than the original character (up to two
1387 The first character of the uppercased version is returned
1388 (but note, as explained above, that there may be more.)
1393 Perl_to_utf8_upper(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp)
1395 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1396 &PL_utf8_toupper, "ToUpper", "utf8::ToSpecUpper");
1400 =for apidoc A|UV|to_utf8_title|U8 *p|U8 *ustrp|STRLEN *lenp
1402 Convert the UTF-8 encoded character at p to its titlecase version and
1403 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1404 that the ustrp needs to be at least UTF8_MAXLEN_UCLC+1 bytes since the
1405 titlecase version may be longer than the original character (up to two
1408 The first character of the titlecased version is returned
1409 (but note, as explained above, that there may be more.)
1414 Perl_to_utf8_title(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp)
1416 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1417 &PL_utf8_totitle, "ToTitle", "utf8::ToSpecTitle");
1421 =for apidoc A|UV|to_utf8_lower|U8 *p|U8 *ustrp|STRLEN *lenp
1423 Convert the UTF-8 encoded character at p to its lowercase version and
1424 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1425 that the ustrp needs to be at least UTF8_MAXLEN_UCLC+1 bytes since the
1426 lowercase version may be longer than the original character (up to two
1429 The first character of the lowercased version is returned
1430 (but note, as explained above, that there may be more.)
1435 Perl_to_utf8_lower(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp)
1437 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1438 &PL_utf8_tolower, "ToLower", "utf8::ToSpecLower");
1442 =for apidoc A|UV|to_utf8_fold|U8 *p|U8 *ustrp|STRLEN *lenp
1444 Convert the UTF-8 encoded character at p to its foldcase version and
1445 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1446 that the ustrp needs to be at least UTF8_MAXLEN_FOLD+1 bytes since the
1447 foldcase version may be longer than the original character (up to
1450 The first character of the foldcased version is returned
1451 (but note, as explained above, that there may be more.)
1456 Perl_to_utf8_fold(pTHX_ U8 *p, U8* ustrp, STRLEN *lenp)
1458 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1459 &PL_utf8_tofold, "ToFold", "utf8::ToSpecFold");
1462 /* a "swash" is a swatch hash */
1465 Perl_swash_init(pTHX_ char* pkg, char* name, SV *listsv, I32 minbits, I32 none)
1468 SV* tokenbufsv = sv_2mortal(NEWSV(0,0));
1470 HV *stash = gv_stashpvn(pkg, strlen(pkg), FALSE);
1473 if (!gv_fetchmeth(stash, "SWASHNEW", 8, -1)) { /* demand load utf8 */
1475 errsv_save = newSVsv(ERRSV);
1476 Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpv(pkg,0), Nullsv);
1478 sv_setsv(ERRSV, errsv_save);
1479 SvREFCNT_dec(errsv_save);
1483 PUSHSTACKi(PERLSI_MAGIC);
1486 PUSHs(sv_2mortal(newSVpvn(pkg, strlen(pkg))));
1487 PUSHs(sv_2mortal(newSVpvn(name, strlen(name))));
1489 PUSHs(sv_2mortal(newSViv(minbits)));
1490 PUSHs(sv_2mortal(newSViv(none)));
1496 if (PL_curcop == &PL_compiling)
1497 /* XXX ought to be handled by lex_start */
1498 sv_setpv(tokenbufsv, PL_tokenbuf);
1499 errsv_save = newSVsv(ERRSV);
1500 if (call_method("SWASHNEW", G_SCALAR))
1501 retval = newSVsv(*PL_stack_sp--);
1503 retval = &PL_sv_undef;
1505 sv_setsv(ERRSV, errsv_save);
1506 SvREFCNT_dec(errsv_save);
1509 if (PL_curcop == &PL_compiling) {
1511 char* pv = SvPV(tokenbufsv, len);
1513 Copy(pv, PL_tokenbuf, len+1, char);
1514 PL_curcop->op_private = PL_hints;
1516 if (!SvROK(retval) || SvTYPE(SvRV(retval)) != SVt_PVHV)
1517 Perl_croak(aTHX_ "SWASHNEW didn't return an HV ref");
1522 /* This API is wrong for special case conversions since we may need to
1523 * return several Unicode characters for a single Unicode character
1524 * (see lib/unicore/SpecCase.txt) The SWASHGET in lib/utf8_heavy.pl is
1525 * the lower-level routine, and it is similarly broken for returning
1526 * multiple values. --jhi */
1528 Perl_swash_fetch(pTHX_ SV *sv, U8 *ptr, bool do_utf8)
1530 HV* hv = (HV*)SvRV(sv);
1539 UV c = NATIVE_TO_ASCII(*ptr);
1541 if (!do_utf8 && !UNI_IS_INVARIANT(c)) {
1542 tmputf8[0] = UTF8_EIGHT_BIT_HI(c);
1543 tmputf8[1] = UTF8_EIGHT_BIT_LO(c);
1546 /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ
1547 * then the "swatch" is a vec() for al the chars which start
1549 * So the key in the hash (klen) is length of encoded char -1
1551 klen = UTF8SKIP(ptr) - 1;
1556 /* If char in invariant then swatch is for all the invariant chars
1557 * In both UTF-8 and UTF8-MOD that happens to be UTF_CONTINUATION_MARK
1559 needents = UTF_CONTINUATION_MARK;
1560 off = NATIVE_TO_UTF(ptr[klen]);
1564 /* If char is encoded then swatch is for the prefix */
1565 needents = (1 << UTF_ACCUMULATION_SHIFT);
1566 off = NATIVE_TO_UTF(ptr[klen]) & UTF_CONTINUATION_MASK;
1570 * This single-entry cache saves about 1/3 of the utf8 overhead in test
1571 * suite. (That is, only 7-8% overall over just a hash cache. Still,
1572 * it's nothing to sniff at.) Pity we usually come through at least
1573 * two function calls to get here...
1575 * NB: this code assumes that swatches are never modified, once generated!
1578 if (hv == PL_last_swash_hv &&
1579 klen == PL_last_swash_klen &&
1580 (!klen || memEQ((char *)ptr, (char *)PL_last_swash_key, klen)) )
1582 tmps = PL_last_swash_tmps;
1583 slen = PL_last_swash_slen;
1586 /* Try our second-level swatch cache, kept in a hash. */
1587 SV** svp = hv_fetch(hv, (char*)ptr, klen, FALSE);
1589 /* If not cached, generate it via utf8::SWASHGET */
1590 if (!svp || !SvPOK(*svp) || !(tmps = (U8*)SvPV(*svp, slen))) {
1592 /* We use utf8n_to_uvuni() as we want an index into
1593 Unicode tables, not a native character number.
1595 UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXLEN, NULL, 0);
1600 PUSHSTACKi(PERLSI_MAGIC);
1604 /* On EBCDIC & ~(0xA0-1) isn't a useful thing to do */
1605 PUSHs(sv_2mortal(newSViv((klen) ?
1606 (code_point & ~(needents - 1)) : 0)));
1607 PUSHs(sv_2mortal(newSViv(needents)));
1609 errsv_save = newSVsv(ERRSV);
1610 if (call_method("SWASHGET", G_SCALAR))
1611 retval = newSVsv(*PL_stack_sp--);
1613 retval = &PL_sv_undef;
1615 sv_setsv(ERRSV, errsv_save);
1616 SvREFCNT_dec(errsv_save);
1620 if (PL_curcop == &PL_compiling)
1621 PL_curcop->op_private = PL_hints;
1623 svp = hv_store(hv, (char*)ptr, klen, retval, 0);
1625 if (!svp || !(tmps = (U8*)SvPV(*svp, slen)) || (slen << 3) < needents)
1626 Perl_croak(aTHX_ "SWASHGET didn't return result of proper length");
1629 PL_last_swash_hv = hv;
1630 PL_last_swash_klen = klen;
1631 PL_last_swash_tmps = tmps;
1632 PL_last_swash_slen = slen;
1634 Copy(ptr, PL_last_swash_key, klen, U8);
1637 switch ((int)((slen << 3) / needents)) {
1639 bit = 1 << (off & 7);
1641 return (tmps[off] & bit) != 0;
1646 return (tmps[off] << 8) + tmps[off + 1] ;
1649 return (tmps[off] << 24) + (tmps[off+1] << 16) + (tmps[off+2] << 8) + tmps[off + 3] ;
1651 Perl_croak(aTHX_ "panic: swash_fetch");
1657 =for apidoc A|U8 *|uvchr_to_utf8|U8 *d|UV uv
1659 Adds the UTF8 representation of the Native codepoint C<uv> to the end
1660 of the string C<d>; C<d> should be have at least C<UTF8_MAXLEN+1> free
1661 bytes available. The return value is the pointer to the byte after the
1662 end of the new character. In other words,
1664 d = uvchr_to_utf8(d, uv);
1666 is the recommended wide native character-aware way of saying
1673 /* On ASCII machines this is normally a macro but we want a
1674 real function in case XS code wants it
1676 #undef Perl_uvchr_to_utf8
1678 Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv)
1680 return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), 0);
1684 Perl_uvchr_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags)
1686 return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), flags);
1690 =for apidoc A|UV|utf8n_to_uvchr|U8 *s|STRLEN curlen|STRLEN *retlen|U32 flags
1692 Returns the native character value of the first character in the string C<s>
1693 which is assumed to be in UTF8 encoding; C<retlen> will be set to the
1694 length, in bytes, of that character.
1696 Allows length and flags to be passed to low level routine.
1700 /* On ASCII machines this is normally a macro but we want
1701 a real function in case XS code wants it
1703 #undef Perl_utf8n_to_uvchr
1705 Perl_utf8n_to_uvchr(pTHX_ U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags)
1707 UV uv = Perl_utf8n_to_uvuni(aTHX_ s, curlen, retlen, flags);
1708 return UNI_TO_NATIVE(uv);
1712 =for apidoc A|char *|pv_uni_display|SV *dsv|U8 *spv|STRLEN len|STRLEN pvlim|UV flags
1714 Build to the scalar dsv a displayable version of the string spv,
1715 length len, the displayable version being at most pvlim bytes long
1716 (if longer, the rest is truncated and "..." will be appended).
1718 The flags argument can have UNI_DISPLAY_ISPRINT set to display
1719 isPRINT()able characters as themselves, UNI_DISPLAY_BACKSLASH
1720 to display the \\[nrfta\\] as the backslashed versions (like '\n')
1721 (UNI_DISPLAY_BACKSLASH is preferred over UNI_DISPLAY_ISPRINT for \\).
1722 UNI_DISPLAY_QQ (and its alias UNI_DISPLAY_REGEX) have both
1723 UNI_DISPLAY_BACKSLASH and UNI_DISPLAY_ISPRINT turned on.
1725 The pointer to the PV of the dsv is returned.
1729 Perl_pv_uni_display(pTHX_ SV *dsv, U8 *spv, STRLEN len, STRLEN pvlim, UV flags)
1734 sv_setpvn(dsv, "", 0);
1735 for (s = (char *)spv, e = s + len; s < e; s += UTF8SKIP(s)) {
1739 if (pvlim && SvCUR(dsv) >= pvlim) {
1743 u = utf8_to_uvchr((U8*)s, 0);
1745 if (!ok && (flags & UNI_DISPLAY_BACKSLASH)) {
1748 Perl_sv_catpvf(aTHX_ dsv, "\\n"); ok = TRUE; break;
1750 Perl_sv_catpvf(aTHX_ dsv, "\\r"); ok = TRUE; break;
1752 Perl_sv_catpvf(aTHX_ dsv, "\\t"); ok = TRUE; break;
1754 Perl_sv_catpvf(aTHX_ dsv, "\\f"); ok = TRUE; break;
1756 Perl_sv_catpvf(aTHX_ dsv, "\\a"); ok = TRUE; break;
1758 Perl_sv_catpvf(aTHX_ dsv, "\\" ); ok = TRUE; break;
1762 /* isPRINT() is the locale-blind version. */
1763 if (!ok && (flags & UNI_DISPLAY_ISPRINT) && isPRINT(u & 0xFF)) {
1764 Perl_sv_catpvf(aTHX_ dsv, "%c", (char)(u & 0xFF));
1769 Perl_sv_catpvf(aTHX_ dsv, "\\x{%"UVxf"}", u);
1772 sv_catpvn(dsv, "...", 3);
1778 =for apidoc A|char *|sv_uni_display|SV *dsv|SV *ssv|STRLEN pvlim|UV flags
1780 Build to the scalar dsv a displayable version of the scalar sv,
1781 the displayable version being at most pvlim bytes long
1782 (if longer, the rest is truncated and "..." will be appended).
1784 The flags argument is as in pv_uni_display().
1786 The pointer to the PV of the dsv is returned.
1790 Perl_sv_uni_display(pTHX_ SV *dsv, SV *ssv, STRLEN pvlim, UV flags)
1792 return Perl_pv_uni_display(aTHX_ dsv, (U8*)SvPVX(ssv), SvCUR(ssv),
1797 =for apidoc A|I32|ibcmp_utf8|const char *s1|char **pe1|register UV l1|bool u1|const char *s2|char **pe2|register UV l2|bool u2
1799 Return true if the strings s1 and s2 differ case-insensitively, false
1800 if not (if they are equal case-insensitively). If u1 is true, the
1801 string s1 is assumed to be in UTF-8-encoded Unicode. If u2 is true,
1802 the string s2 is assumed to be in UTF-8-encoded Unicode. If u1 or u2
1803 are false, the respective string is assumed to be in native 8-bit
1806 If the pe1 and pe2 are non-NULL, the scanning pointers will be copied
1807 in there (they will point at the beginning of the I<next> character).
1808 If the pointers behind pe1 or pe2 are non-NULL, they are the end
1809 pointers beyond which scanning will not continue under any
1810 circustances. If the byte lengths l1 and l2 are non-zero, s1+l1 and
1811 s2+l2 will be used as goal end pointers that will also stop the scan,
1812 and which qualify towards defining a successful match: all the scans
1813 that define an explicit length must reach their goal pointers for
1814 a match to succeed).
1816 For case-insensitiveness, the "casefolding" of Unicode is used
1817 instead of upper/lowercasing both the characters, see
1818 http://www.unicode.org/unicode/reports/tr21/ (Case Mappings).
1822 Perl_ibcmp_utf8(pTHX_ const char *s1, char **pe1, register UV l1, bool u1, const char *s2, char **pe2, register UV l2, bool u2)
1824 register U8 *p1 = (U8*)s1;
1825 register U8 *p2 = (U8*)s2;
1826 register U8 *e1 = 0, *f1 = 0, *q1 = 0;
1827 register U8 *e2 = 0, *f2 = 0, *q2 = 0;
1828 STRLEN n1 = 0, n2 = 0;
1829 U8 foldbuf1[UTF8_MAXLEN_FOLD+1];
1830 U8 foldbuf2[UTF8_MAXLEN_FOLD+1];
1832 STRLEN foldlen1, foldlen2;
1837 if (e1 == 0 || (l1 && l1 < e1 - (U8*)s1))
1841 if (e2 == 0 || (l2 && l2 < e2 - (U8*)s2))
1844 if ((e1 == 0 && f1 == 0) || (e2 == 0 && f2 == 0) || (f1 == 0 && f2 == 0))
1845 return 1; /* mismatch; possible infinite loop or false positive */
1848 natbuf[1] = 0; /* Need to terminate the buffer. */
1850 while ((e1 == 0 || p1 < e1) &&
1851 (f1 == 0 || p1 < f1) &&
1852 (e2 == 0 || p2 < e2) &&
1853 (f2 == 0 || p2 < f2)) {
1856 to_utf8_fold(p1, foldbuf1, &foldlen1);
1859 to_utf8_fold(natbuf, foldbuf1, &foldlen1);
1866 to_utf8_fold(p2, foldbuf2, &foldlen2);
1869 to_utf8_fold(natbuf, foldbuf2, &foldlen2);
1875 if ( UTF8SKIP(q1) != UTF8SKIP(q2) ||
1876 (UTF8SKIP(q1) == 1 && *q1 != *q2) ||
1877 memNE((char*)q1, (char*)q2, UTF8SKIP(q1)) )
1878 return 1; /* mismatch */
1885 p1 += u1 ? UTF8SKIP(p1) : 1;
1887 p2 += u2 ? UTF8SKIP(p2) : 1;
1891 /* A match is defined by all the scans that specified
1892 * an explicit length reaching their final goals. */
1893 match = (f1 == 0 || p1 == f1) && (f2 == 0 || p2 == f2);
1902 return match ? 0 : 1; /* 0 match, 1 mismatch */