3 * Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
4 * by Larry Wall and others
6 * You may distribute under the terms of either the GNU General Public
7 * License or the Artistic License, as specified in the README file.
12 * 'What a fix!' said Sam. 'That's the one place in all the lands we've ever
13 * heard of that we don't want to see any closer; and that's the one place
14 * we're trying to get to! And that's just where we can't get, nohow.'
16 * [p.603 of _The Lord of the Rings_, IV/I: "The Taming of Sméagol"]
18 * 'Well do I understand your speech,' he answered in the same language;
19 * 'yet few strangers do so. Why then do you not speak in the Common Tongue,
20 * as is the custom in the West, if you wish to be answered?'
21 * --Gandalf, addressing Théoden's door wardens
23 * [p.508 of _The Lord of the Rings_, III/vi: "The King of the Golden Hall"]
25 * ...the travellers perceived that the floor was paved with stones of many
26 * hues; branching runes and strange devices intertwined beneath their feet.
28 * [p.512 of _The Lord of the Rings_, III/vi: "The King of the Golden Hall"]
32 #define PERL_IN_UTF8_C
36 /* Separate prototypes needed because in ASCII systems these are
37 * usually macros but they still are compiled as code, too. */
38 PERL_CALLCONV UV Perl_utf8n_to_uvchr(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags);
39 PERL_CALLCONV U8* Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv);
42 static const char unees[] =
43 "Malformed UTF-8 character (unexpected end of string)";
46 =head1 Unicode Support
48 This file contains various utility functions for manipulating UTF8-encoded
49 strings. For the uninitiated, this is a method of representing arbitrary
50 Unicode characters as a variable number of bytes, in such a way that
51 characters in the ASCII range are unmodified, and a zero byte never appears
52 within non-zero characters.
58 =for apidoc is_ascii_string
60 Returns true if the first C<len> bytes of the given string are the same whether
61 or not the string is encoded in UTF-8 (or UTF-EBCDIC on EBCDIC machines). That
62 is, if they are invariant. On ASCII-ish machines, only ASCII characters
63 fit this definition, hence the function's name.
65 See also is_utf8_string(), is_utf8_string_loclen(), and is_utf8_string_loc().
71 Perl_is_ascii_string(const U8 *s, STRLEN len)
73 const U8* const send = s + (len ? len : strlen((const char *)s));
76 PERL_ARGS_ASSERT_IS_ASCII_STRING;
78 for (; x < send; ++x) {
79 if (!UTF8_IS_INVARIANT(*x))
87 =for apidoc uvuni_to_utf8_flags
89 Adds the UTF-8 representation of the Unicode codepoint C<uv> to the end
90 of the string C<d>; C<d> should be have at least C<UTF8_MAXBYTES+1> free
91 bytes available. The return value is the pointer to the byte after the
92 end of the new character. In other words,
94 d = uvuni_to_utf8_flags(d, uv, flags);
98 d = uvuni_to_utf8(d, uv);
100 (which is equivalent to)
102 d = uvuni_to_utf8_flags(d, uv, 0);
104 is the recommended Unicode-aware way of saying
112 Perl_uvuni_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags)
114 PERL_ARGS_ASSERT_UVUNI_TO_UTF8_FLAGS;
116 if (ckWARN(WARN_UTF8)) {
117 if (UNICODE_IS_SURROGATE(uv) &&
118 !(flags & UNICODE_ALLOW_SURROGATE))
119 Perl_warner(aTHX_ packWARN(WARN_UTF8), "UTF-16 surrogate 0x%04"UVxf, uv);
121 ((uv >= 0xFDD0 && uv <= 0xFDEF &&
122 !(flags & UNICODE_ALLOW_FDD0))
124 ((uv & 0xFFFE) == 0xFFFE && /* Either FFFE or FFFF. */
125 !(flags & UNICODE_ALLOW_FFFF))) &&
126 /* UNICODE_ALLOW_SUPER includes
127 * FFFEs and FFFFs beyond 0x10FFFF. */
128 ((uv <= PERL_UNICODE_MAX) ||
129 !(flags & UNICODE_ALLOW_SUPER))
131 Perl_warner(aTHX_ packWARN(WARN_UTF8),
132 "Unicode non-character 0x%04"UVxf" is illegal for interchange", uv);
134 if (UNI_IS_INVARIANT(uv)) {
135 *d++ = (U8)UTF_TO_NATIVE(uv);
140 STRLEN len = UNISKIP(uv);
143 *p-- = (U8)UTF_TO_NATIVE((uv & UTF_CONTINUATION_MASK) | UTF_CONTINUATION_MARK);
144 uv >>= UTF_ACCUMULATION_SHIFT;
146 *p = (U8)UTF_TO_NATIVE((uv & UTF_START_MASK(len)) | UTF_START_MARK(len));
149 #else /* Non loop style */
151 *d++ = (U8)(( uv >> 6) | 0xc0);
152 *d++ = (U8)(( uv & 0x3f) | 0x80);
156 *d++ = (U8)(( uv >> 12) | 0xe0);
157 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
158 *d++ = (U8)(( uv & 0x3f) | 0x80);
162 *d++ = (U8)(( uv >> 18) | 0xf0);
163 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
164 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
165 *d++ = (U8)(( uv & 0x3f) | 0x80);
168 if (uv < 0x4000000) {
169 *d++ = (U8)(( uv >> 24) | 0xf8);
170 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
171 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
172 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
173 *d++ = (U8)(( uv & 0x3f) | 0x80);
176 if (uv < 0x80000000) {
177 *d++ = (U8)(( uv >> 30) | 0xfc);
178 *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80);
179 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
180 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
181 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
182 *d++ = (U8)(( uv & 0x3f) | 0x80);
186 if (uv < UTF8_QUAD_MAX)
189 *d++ = 0xfe; /* Can't match U+FEFF! */
190 *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80);
191 *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80);
192 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
193 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
194 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
195 *d++ = (U8)(( uv & 0x3f) | 0x80);
200 *d++ = 0xff; /* Can't match U+FFFE! */
201 *d++ = 0x80; /* 6 Reserved bits */
202 *d++ = (U8)(((uv >> 60) & 0x0f) | 0x80); /* 2 Reserved bits */
203 *d++ = (U8)(((uv >> 54) & 0x3f) | 0x80);
204 *d++ = (U8)(((uv >> 48) & 0x3f) | 0x80);
205 *d++ = (U8)(((uv >> 42) & 0x3f) | 0x80);
206 *d++ = (U8)(((uv >> 36) & 0x3f) | 0x80);
207 *d++ = (U8)(((uv >> 30) & 0x3f) | 0x80);
208 *d++ = (U8)(((uv >> 24) & 0x3f) | 0x80);
209 *d++ = (U8)(((uv >> 18) & 0x3f) | 0x80);
210 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
211 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
212 *d++ = (U8)(( uv & 0x3f) | 0x80);
216 #endif /* Loop style */
221 Tests if some arbitrary number of bytes begins in a valid UTF-8
222 character. Note that an INVARIANT (i.e. ASCII) character is a valid
223 UTF-8 character. The actual number of bytes in the UTF-8 character
224 will be returned if it is valid, otherwise 0.
226 This is the "slow" version as opposed to the "fast" version which is
227 the "unrolled" IS_UTF8_CHAR(). E.g. for t/uni/class.t the speed
228 difference is a factor of 2 to 3. For lengths (UTF8SKIP(s)) of four
229 or less you should use the IS_UTF8_CHAR(), for lengths of five or more
230 you should use the _slow(). In practice this means that the _slow()
231 will be used very rarely, since the maximum Unicode code point (as of
232 Unicode 4.1) is U+10FFFF, which encodes in UTF-8 to four bytes. Only
233 the "Perl extended UTF-8" (the infamous 'v-strings') will encode into
238 S_is_utf8_char_slow(const U8 *s, const STRLEN len)
244 PERL_ARGS_ASSERT_IS_UTF8_CHAR_SLOW;
246 if (UTF8_IS_INVARIANT(u))
249 if (!UTF8_IS_START(u))
252 if (len < 2 || !UTF8_IS_CONTINUATION(s[1]))
258 u = NATIVE_TO_UTF(u);
260 u &= UTF_START_MASK(len);
264 if (!UTF8_IS_CONTINUATION(*s))
266 uv = UTF8_ACCUMULATE(uv, *s);
273 if ((STRLEN)UNISKIP(uv) < len)
280 =for apidoc is_utf8_char
282 Tests if some arbitrary number of bytes begins in a valid UTF-8
283 character. Note that an INVARIANT (i.e. ASCII on non-EBCDIC machines)
284 character is a valid UTF-8 character. The actual number of bytes in the UTF-8
285 character will be returned if it is valid, otherwise 0.
289 Perl_is_utf8_char(const U8 *s)
291 const STRLEN len = UTF8SKIP(s);
293 PERL_ARGS_ASSERT_IS_UTF8_CHAR;
295 if (IS_UTF8_CHAR_FAST(len))
296 return IS_UTF8_CHAR(s, len) ? len : 0;
297 #endif /* #ifdef IS_UTF8_CHAR */
298 return is_utf8_char_slow(s, len);
303 =for apidoc is_utf8_string
305 Returns true if first C<len> bytes of the given string form a valid
306 UTF-8 string, false otherwise. Note that 'a valid UTF-8 string' does
307 not mean 'a string that contains code points above 0x7F encoded in UTF-8'
308 because a valid ASCII string is a valid UTF-8 string.
310 See also is_ascii_string(), is_utf8_string_loclen(), and is_utf8_string_loc().
316 Perl_is_utf8_string(const U8 *s, STRLEN len)
318 const U8* const send = s + (len ? len : strlen((const char *)s));
321 PERL_ARGS_ASSERT_IS_UTF8_STRING;
325 /* Inline the easy bits of is_utf8_char() here for speed... */
326 if (UTF8_IS_INVARIANT(*x))
328 else if (!UTF8_IS_START(*x))
331 /* ... and call is_utf8_char() only if really needed. */
334 if (IS_UTF8_CHAR_FAST(c)) {
335 if (!IS_UTF8_CHAR(x, c))
339 c = is_utf8_char_slow(x, c);
342 #endif /* #ifdef IS_UTF8_CHAR */
357 Implemented as a macro in utf8.h
359 =for apidoc is_utf8_string_loc
361 Like is_utf8_string() but stores the location of the failure (in the
362 case of "utf8ness failure") or the location s+len (in the case of
363 "utf8ness success") in the C<ep>.
365 See also is_utf8_string_loclen() and is_utf8_string().
367 =for apidoc is_utf8_string_loclen
369 Like is_utf8_string() but stores the location of the failure (in the
370 case of "utf8ness failure") or the location s+len (in the case of
371 "utf8ness success") in the C<ep>, and the number of UTF-8
372 encoded characters in the C<el>.
374 See also is_utf8_string_loc() and is_utf8_string().
380 Perl_is_utf8_string_loclen(const U8 *s, STRLEN len, const U8 **ep, STRLEN *el)
382 const U8* const send = s + (len ? len : strlen((const char *)s));
387 PERL_ARGS_ASSERT_IS_UTF8_STRING_LOCLEN;
390 /* Inline the easy bits of is_utf8_char() here for speed... */
391 if (UTF8_IS_INVARIANT(*x))
393 else if (!UTF8_IS_START(*x))
396 /* ... and call is_utf8_char() only if really needed. */
399 if (IS_UTF8_CHAR_FAST(c)) {
400 if (!IS_UTF8_CHAR(x, c))
403 c = is_utf8_char_slow(x, c);
406 #endif /* #ifdef IS_UTF8_CHAR */
425 =for apidoc utf8n_to_uvuni
427 Bottom level UTF-8 decode routine.
428 Returns the Unicode code point value of the first character in the string C<s>
429 which is assumed to be in UTF-8 encoding and no longer than C<curlen>;
430 C<retlen> will be set to the length, in bytes, of that character.
432 If C<s> does not point to a well-formed UTF-8 character, the behaviour
433 is dependent on the value of C<flags>: if it contains UTF8_CHECK_ONLY,
434 it is assumed that the caller will raise a warning, and this function
435 will silently just set C<retlen> to C<-1> and return zero. If the
436 C<flags> does not contain UTF8_CHECK_ONLY, warnings about
437 malformations will be given, C<retlen> will be set to the expected
438 length of the UTF-8 character in bytes, and zero will be returned.
440 The C<flags> can also contain various flags to allow deviations from
441 the strict UTF-8 encoding (see F<utf8.h>).
443 Most code should use utf8_to_uvchr() rather than call this directly.
449 Perl_utf8n_to_uvuni(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen, U32 flags)
452 const U8 * const s0 = s;
455 const bool dowarn = ckWARN_d(WARN_UTF8);
456 const UV startbyte = *s;
457 STRLEN expectlen = 0;
461 PERL_ARGS_ASSERT_UTF8N_TO_UVUNI;
463 /* This list is a superset of the UTF8_ALLOW_XXX. BUT it isn't, eg SUPER missing XXX */
465 #define UTF8_WARN_EMPTY 1
466 #define UTF8_WARN_CONTINUATION 2
467 #define UTF8_WARN_NON_CONTINUATION 3
468 #define UTF8_WARN_FE_FF 4
469 #define UTF8_WARN_SHORT 5
470 #define UTF8_WARN_OVERFLOW 6
471 #define UTF8_WARN_SURROGATE 7
472 #define UTF8_WARN_LONG 8
473 #define UTF8_WARN_FFFF 9 /* Also FFFE. */
476 !(flags & UTF8_ALLOW_EMPTY)) {
477 warning = UTF8_WARN_EMPTY;
481 if (UTF8_IS_INVARIANT(uv)) {
484 return (UV) (NATIVE_TO_UTF(*s));
487 if (UTF8_IS_CONTINUATION(uv) &&
488 !(flags & UTF8_ALLOW_CONTINUATION)) {
489 warning = UTF8_WARN_CONTINUATION;
493 if (UTF8_IS_START(uv) && curlen > 1 && !UTF8_IS_CONTINUATION(s[1]) &&
494 !(flags & UTF8_ALLOW_NON_CONTINUATION)) {
495 warning = UTF8_WARN_NON_CONTINUATION;
500 uv = NATIVE_TO_UTF(uv);
502 if ((uv == 0xfe || uv == 0xff) &&
503 !(flags & UTF8_ALLOW_FE_FF)) {
504 warning = UTF8_WARN_FE_FF;
509 if (!(uv & 0x20)) { len = 2; uv &= 0x1f; }
510 else if (!(uv & 0x10)) { len = 3; uv &= 0x0f; }
511 else if (!(uv & 0x08)) { len = 4; uv &= 0x07; }
512 else if (!(uv & 0x04)) { len = 5; uv &= 0x03; }
514 else if (!(uv & 0x02)) { len = 6; uv &= 0x01; }
515 else { len = 7; uv &= 0x01; }
517 else if (!(uv & 0x02)) { len = 6; uv &= 0x01; }
518 else if (!(uv & 0x01)) { len = 7; uv = 0; }
519 else { len = 13; uv = 0; } /* whoa! */
527 if ((curlen < expectlen) &&
528 !(flags & UTF8_ALLOW_SHORT)) {
529 warning = UTF8_WARN_SHORT;
538 if (!UTF8_IS_CONTINUATION(*s) &&
539 !(flags & UTF8_ALLOW_NON_CONTINUATION)) {
541 warning = UTF8_WARN_NON_CONTINUATION;
545 uv = UTF8_ACCUMULATE(uv, *s);
547 /* These cannot be allowed. */
549 if (expectlen != 13 && !(flags & UTF8_ALLOW_LONG)) {
550 warning = UTF8_WARN_LONG;
554 else { /* uv < ouv */
555 /* This cannot be allowed. */
556 warning = UTF8_WARN_OVERFLOW;
564 if (UNICODE_IS_SURROGATE(uv) &&
565 !(flags & UTF8_ALLOW_SURROGATE)) {
566 warning = UTF8_WARN_SURROGATE;
568 } else if ((expectlen > (STRLEN)UNISKIP(uv)) &&
569 !(flags & UTF8_ALLOW_LONG)) {
570 warning = UTF8_WARN_LONG;
572 } else if (UNICODE_IS_ILLEGAL(uv) &&
573 !(flags & UTF8_ALLOW_FFFF)) {
574 warning = UTF8_WARN_FFFF;
582 if (flags & UTF8_CHECK_ONLY) {
584 *retlen = ((STRLEN) -1);
589 if (warning == UTF8_WARN_FFFF) {
590 sv = newSVpvs_flags("Unicode non-character ", SVs_TEMP);
591 Perl_sv_catpvf(aTHX_ sv, "0x%04"UVxf" is illegal for interchange", uv);
594 sv = newSVpvs_flags("Malformed UTF-8 character ", SVs_TEMP);
597 case 0: /* Intentionally empty. */ break;
598 case UTF8_WARN_EMPTY:
599 sv_catpvs(sv, "(empty string)");
601 case UTF8_WARN_CONTINUATION:
602 Perl_sv_catpvf(aTHX_ sv, "(unexpected continuation byte 0x%02"UVxf", with no preceding start byte)", uv);
604 case UTF8_WARN_NON_CONTINUATION:
606 Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", immediately after start byte 0x%02"UVxf")",
607 (UV)s[1], startbyte);
609 const int len = (int)(s-s0);
610 Perl_sv_catpvf(aTHX_ sv, "(unexpected non-continuation byte 0x%02"UVxf", %d byte%s after start byte 0x%02"UVxf", expected %d bytes)",
611 (UV)s[1], len, len > 1 ? "s" : "", startbyte, (int)expectlen);
615 case UTF8_WARN_FE_FF:
616 Perl_sv_catpvf(aTHX_ sv, "(byte 0x%02"UVxf")", uv);
618 case UTF8_WARN_SHORT:
619 Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")",
620 (int)curlen, curlen == 1 ? "" : "s", (int)expectlen, startbyte);
621 expectlen = curlen; /* distance for caller to skip */
623 case UTF8_WARN_OVERFLOW:
624 Perl_sv_catpvf(aTHX_ sv, "(overflow at 0x%"UVxf", byte 0x%02x, after start byte 0x%02"UVxf")",
627 case UTF8_WARN_SURROGATE:
628 Perl_sv_catpvf(aTHX_ sv, "(UTF-16 surrogate 0x%04"UVxf")", uv);
631 Perl_sv_catpvf(aTHX_ sv, "(%d byte%s, need %d, after start byte 0x%02"UVxf")",
632 (int)expectlen, expectlen == 1 ? "": "s", UNISKIP(uv), startbyte);
635 sv_catpvs(sv, "(unknown reason)");
641 const char * const s = SvPVX_const(sv);
644 Perl_warner(aTHX_ packWARN(WARN_UTF8),
645 "%s in %s", s, OP_DESC(PL_op));
647 Perl_warner(aTHX_ packWARN(WARN_UTF8), "%s", s);
652 *retlen = expectlen ? expectlen : len;
658 =for apidoc utf8_to_uvchr
660 Returns the native character value of the first character in the string C<s>
661 which is assumed to be in UTF-8 encoding; C<retlen> will be set to the
662 length, in bytes, of that character.
664 If C<s> does not point to a well-formed UTF-8 character, zero is
665 returned and retlen is set, if possible, to -1.
671 Perl_utf8_to_uvchr(pTHX_ const U8 *s, STRLEN *retlen)
673 PERL_ARGS_ASSERT_UTF8_TO_UVCHR;
675 return utf8n_to_uvchr(s, UTF8_MAXBYTES, retlen,
676 ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
680 =for apidoc utf8_to_uvuni
682 Returns the Unicode code point of the first character in the string C<s>
683 which is assumed to be in UTF-8 encoding; C<retlen> will be set to the
684 length, in bytes, of that character.
686 This function should only be used when the returned UV is considered
687 an index into the Unicode semantic tables (e.g. swashes).
689 If C<s> does not point to a well-formed UTF-8 character, zero is
690 returned and retlen is set, if possible, to -1.
696 Perl_utf8_to_uvuni(pTHX_ const U8 *s, STRLEN *retlen)
698 PERL_ARGS_ASSERT_UTF8_TO_UVUNI;
700 /* Call the low level routine asking for checks */
701 return Perl_utf8n_to_uvuni(aTHX_ s, UTF8_MAXBYTES, retlen,
702 ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
706 =for apidoc utf8_length
708 Return the length of the UTF-8 char encoded string C<s> in characters.
709 Stops at C<e> (inclusive). If C<e E<lt> s> or if the scan would end
710 up past C<e>, croaks.
716 Perl_utf8_length(pTHX_ const U8 *s, const U8 *e)
721 PERL_ARGS_ASSERT_UTF8_LENGTH;
723 /* Note: cannot use UTF8_IS_...() too eagerly here since e.g.
724 * the bitops (especially ~) can create illegal UTF-8.
725 * In other words: in Perl UTF-8 is not just for Unicode. */
728 goto warn_and_return;
730 if (!UTF8_IS_INVARIANT(*s))
741 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8),
742 "%s in %s", unees, OP_DESC(PL_op));
744 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8), unees);
751 =for apidoc utf8_distance
753 Returns the number of UTF-8 characters between the UTF-8 pointers C<a>
756 WARNING: use only if you *know* that the pointers point inside the
763 Perl_utf8_distance(pTHX_ const U8 *a, const U8 *b)
765 PERL_ARGS_ASSERT_UTF8_DISTANCE;
767 return (a < b) ? -1 * (IV) utf8_length(a, b) : (IV) utf8_length(b, a);
773 Return the UTF-8 pointer C<s> displaced by C<off> characters, either
776 WARNING: do not use the following unless you *know* C<off> is within
777 the UTF-8 data pointed to by C<s> *and* that on entry C<s> is aligned
778 on the first byte of character or just after the last byte of a character.
784 Perl_utf8_hop(pTHX_ const U8 *s, I32 off)
786 PERL_ARGS_ASSERT_UTF8_HOP;
789 /* Note: cannot use UTF8_IS_...() too eagerly here since e.g
790 * the bitops (especially ~) can create illegal UTF-8.
791 * In other words: in Perl UTF-8 is not just for Unicode. */
800 while (UTF8_IS_CONTINUATION(*s))
808 =for apidoc bytes_cmp_utf8
810 Compares the sequence of characters (stored as octets) in b, blen with the
811 sequence of characters (stored as UTF-8) in u, ulen. Returns 0 if they are
812 equal, -1 or -2 if the first string is less than the second string, +1 or +2
813 if the first string is greater than the second string.
815 -1 or +1 is returned if the shorter string was identical to the start of the
816 longer string. -2 or +2 is returned if the was a difference between characters
823 Perl_bytes_cmp_utf8(pTHX_ const U8 *b, STRLEN blen, const U8 *u, STRLEN ulen)
825 const U8 *const bend = b + blen;
826 const U8 *const uend = u + ulen;
828 PERL_ARGS_ASSERT_BYTES_CMP_UTF8;
832 while (b < bend && u < uend) {
834 if (!UTF8_IS_INVARIANT(c)) {
835 if (UTF8_IS_DOWNGRADEABLE_START(c)) {
838 if (UTF8_IS_CONTINUATION(c1)) {
839 c = UTF8_ACCUMULATE(NATIVE_TO_UTF(c), c1);
840 c = ASCII_TO_NATIVE(c);
842 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8),
843 "Malformed UTF-8 character "
844 "(unexpected non-continuation byte 0x%02x"
845 ", immediately after start byte 0x%02x)"
846 /* Dear diag.t, it's in the pod. */
849 PL_op ? OP_DESC(PL_op) : "");
854 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8),
855 "%s in %s", unees, OP_DESC(PL_op));
857 Perl_ck_warner_d(aTHX_ packWARN(WARN_UTF8), unees);
858 return -2; /* Really want to return undef :-) */
865 return *b < c ? -2 : +2;
870 if (b == bend && u == uend)
873 return b < bend ? +1 : -1;
877 =for apidoc utf8_to_bytes
879 Converts a string C<s> of length C<len> from UTF-8 into native byte encoding.
880 Unlike C<bytes_to_utf8>, this over-writes the original string, and
881 updates len to contain the new length.
882 Returns zero on failure, setting C<len> to -1.
884 If you need a copy of the string, see C<bytes_from_utf8>.
890 Perl_utf8_to_bytes(pTHX_ U8 *s, STRLEN *len)
893 U8 * const send = s + *len;
896 PERL_ARGS_ASSERT_UTF8_TO_BYTES;
898 /* ensure valid UTF-8 and chars < 256 before updating string */
902 if (!UTF8_IS_INVARIANT(c) &&
903 (!UTF8_IS_DOWNGRADEABLE_START(c) || (s >= send)
904 || !(c = *s++) || !UTF8_IS_CONTINUATION(c))) {
905 *len = ((STRLEN) -1);
913 *d++ = (U8)utf8_to_uvchr(s, &ulen);
922 =for apidoc bytes_from_utf8
924 Converts a string C<s> of length C<len> from UTF-8 into native byte encoding.
925 Unlike C<utf8_to_bytes> but like C<bytes_to_utf8>, returns a pointer to
926 the newly-created string, and updates C<len> to contain the new
927 length. Returns the original string if no conversion occurs, C<len>
928 is unchanged. Do nothing if C<is_utf8> points to 0. Sets C<is_utf8> to
929 0 if C<s> is converted or consisted entirely of characters that are invariant
930 in utf8 (i.e., US-ASCII on non-EBCDIC machines).
936 Perl_bytes_from_utf8(pTHX_ const U8 *s, STRLEN *len, bool *is_utf8)
943 PERL_ARGS_ASSERT_BYTES_FROM_UTF8;
949 /* ensure valid UTF-8 and chars < 256 before converting string */
950 for (send = s + *len; s < send;) {
952 if (!UTF8_IS_INVARIANT(c)) {
953 if (UTF8_IS_DOWNGRADEABLE_START(c) && s < send &&
954 (c = *s++) && UTF8_IS_CONTINUATION(c))
963 Newx(d, (*len) - count + 1, U8);
964 s = start; start = d;
967 if (!UTF8_IS_INVARIANT(c)) {
968 /* Then it is two-byte encoded */
969 c = UTF8_ACCUMULATE(NATIVE_TO_UTF(c), *s++);
970 c = ASCII_TO_NATIVE(c);
980 =for apidoc bytes_to_utf8
982 Converts a string C<s> of length C<len> from the native encoding into UTF-8.
983 Returns a pointer to the newly-created string, and sets C<len> to
984 reflect the new length.
986 A NUL character will be written after the end of the string.
988 If you want to convert to UTF-8 from encodings other than
989 the native (Latin1 or EBCDIC),
990 see sv_recode_to_utf8().
996 Perl_bytes_to_utf8(pTHX_ const U8 *s, STRLEN *len)
998 const U8 * const send = s + (*len);
1002 PERL_ARGS_ASSERT_BYTES_TO_UTF8;
1003 PERL_UNUSED_CONTEXT;
1005 Newx(d, (*len) * 2 + 1, U8);
1009 const UV uv = NATIVE_TO_ASCII(*s++);
1010 if (UNI_IS_INVARIANT(uv))
1011 *d++ = (U8)UTF_TO_NATIVE(uv);
1013 *d++ = (U8)UTF8_EIGHT_BIT_HI(uv);
1014 *d++ = (U8)UTF8_EIGHT_BIT_LO(uv);
1023 * Convert native (big-endian) or reversed (little-endian) UTF-16 to UTF-8.
1025 * Destination must be pre-extended to 3/2 source. Do not use in-place.
1026 * We optimize for native, for obvious reasons. */
1029 Perl_utf16_to_utf8(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
1034 PERL_ARGS_ASSERT_UTF16_TO_UTF8;
1037 Perl_croak(aTHX_ "panic: utf16_to_utf8: odd bytelen %"UVuf, (UV)bytelen);
1042 UV uv = (p[0] << 8) + p[1]; /* UTF-16BE */
1046 *d++ = UNI_TO_NATIVE(uv);
1053 *d++ = (U8)(( uv >> 6) | 0xc0);
1054 *d++ = (U8)(( uv & 0x3f) | 0x80);
1057 if (uv >= 0xd800 && uv <= 0xdbff) { /* surrogates */
1059 Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
1061 UV low = (p[0] << 8) + p[1];
1063 if (low < 0xdc00 || low > 0xdfff)
1064 Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
1065 uv = ((uv - 0xd800) << 10) + (low - 0xdc00) + 0x10000;
1067 } else if (uv >= 0xdc00 && uv <= 0xdfff) {
1068 Perl_croak(aTHX_ "Malformed UTF-16 surrogate");
1071 *d++ = (U8)(( uv >> 12) | 0xe0);
1072 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
1073 *d++ = (U8)(( uv & 0x3f) | 0x80);
1077 *d++ = (U8)(( uv >> 18) | 0xf0);
1078 *d++ = (U8)(((uv >> 12) & 0x3f) | 0x80);
1079 *d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
1080 *d++ = (U8)(( uv & 0x3f) | 0x80);
1084 *newlen = d - dstart;
1088 /* Note: this one is slightly destructive of the source. */
1091 Perl_utf16_to_utf8_reversed(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
1094 U8* const send = s + bytelen;
1096 PERL_ARGS_ASSERT_UTF16_TO_UTF8_REVERSED;
1099 Perl_croak(aTHX_ "panic: utf16_to_utf8_reversed: odd bytelen %"UVuf,
1103 const U8 tmp = s[0];
1108 return utf16_to_utf8(p, d, bytelen, newlen);
1111 /* for now these are all defined (inefficiently) in terms of the utf8 versions */
1114 Perl_is_uni_alnum(pTHX_ UV c)
1116 U8 tmpbuf[UTF8_MAXBYTES+1];
1117 uvchr_to_utf8(tmpbuf, c);
1118 return is_utf8_alnum(tmpbuf);
1122 Perl_is_uni_idfirst(pTHX_ UV c)
1124 U8 tmpbuf[UTF8_MAXBYTES+1];
1125 uvchr_to_utf8(tmpbuf, c);
1126 return is_utf8_idfirst(tmpbuf);
1130 Perl_is_uni_alpha(pTHX_ UV c)
1132 U8 tmpbuf[UTF8_MAXBYTES+1];
1133 uvchr_to_utf8(tmpbuf, c);
1134 return is_utf8_alpha(tmpbuf);
1138 Perl_is_uni_ascii(pTHX_ UV c)
1140 U8 tmpbuf[UTF8_MAXBYTES+1];
1141 uvchr_to_utf8(tmpbuf, c);
1142 return is_utf8_ascii(tmpbuf);
1146 Perl_is_uni_space(pTHX_ UV c)
1148 U8 tmpbuf[UTF8_MAXBYTES+1];
1149 uvchr_to_utf8(tmpbuf, c);
1150 return is_utf8_space(tmpbuf);
1154 Perl_is_uni_digit(pTHX_ UV c)
1156 U8 tmpbuf[UTF8_MAXBYTES+1];
1157 uvchr_to_utf8(tmpbuf, c);
1158 return is_utf8_digit(tmpbuf);
1162 Perl_is_uni_upper(pTHX_ UV c)
1164 U8 tmpbuf[UTF8_MAXBYTES+1];
1165 uvchr_to_utf8(tmpbuf, c);
1166 return is_utf8_upper(tmpbuf);
1170 Perl_is_uni_lower(pTHX_ UV c)
1172 U8 tmpbuf[UTF8_MAXBYTES+1];
1173 uvchr_to_utf8(tmpbuf, c);
1174 return is_utf8_lower(tmpbuf);
1178 Perl_is_uni_cntrl(pTHX_ UV c)
1180 U8 tmpbuf[UTF8_MAXBYTES+1];
1181 uvchr_to_utf8(tmpbuf, c);
1182 return is_utf8_cntrl(tmpbuf);
1186 Perl_is_uni_graph(pTHX_ UV c)
1188 U8 tmpbuf[UTF8_MAXBYTES+1];
1189 uvchr_to_utf8(tmpbuf, c);
1190 return is_utf8_graph(tmpbuf);
1194 Perl_is_uni_print(pTHX_ UV c)
1196 U8 tmpbuf[UTF8_MAXBYTES+1];
1197 uvchr_to_utf8(tmpbuf, c);
1198 return is_utf8_print(tmpbuf);
1202 Perl_is_uni_punct(pTHX_ UV c)
1204 U8 tmpbuf[UTF8_MAXBYTES+1];
1205 uvchr_to_utf8(tmpbuf, c);
1206 return is_utf8_punct(tmpbuf);
1210 Perl_is_uni_xdigit(pTHX_ UV c)
1212 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
1213 uvchr_to_utf8(tmpbuf, c);
1214 return is_utf8_xdigit(tmpbuf);
1218 Perl_to_uni_upper(pTHX_ UV c, U8* p, STRLEN *lenp)
1220 PERL_ARGS_ASSERT_TO_UNI_UPPER;
1222 uvchr_to_utf8(p, c);
1223 return to_utf8_upper(p, p, lenp);
1227 Perl_to_uni_title(pTHX_ UV c, U8* p, STRLEN *lenp)
1229 PERL_ARGS_ASSERT_TO_UNI_TITLE;
1231 uvchr_to_utf8(p, c);
1232 return to_utf8_title(p, p, lenp);
1236 Perl_to_uni_lower(pTHX_ UV c, U8* p, STRLEN *lenp)
1238 PERL_ARGS_ASSERT_TO_UNI_LOWER;
1240 uvchr_to_utf8(p, c);
1241 return to_utf8_lower(p, p, lenp);
1245 Perl_to_uni_fold(pTHX_ UV c, U8* p, STRLEN *lenp)
1247 PERL_ARGS_ASSERT_TO_UNI_FOLD;
1249 uvchr_to_utf8(p, c);
1250 return to_utf8_fold(p, p, lenp);
1253 /* for now these all assume no locale info available for Unicode > 255 */
1256 Perl_is_uni_alnum_lc(pTHX_ UV c)
1258 return is_uni_alnum(c); /* XXX no locale support yet */
1262 Perl_is_uni_idfirst_lc(pTHX_ UV c)
1264 return is_uni_idfirst(c); /* XXX no locale support yet */
1268 Perl_is_uni_alpha_lc(pTHX_ UV c)
1270 return is_uni_alpha(c); /* XXX no locale support yet */
1274 Perl_is_uni_ascii_lc(pTHX_ UV c)
1276 return is_uni_ascii(c); /* XXX no locale support yet */
1280 Perl_is_uni_space_lc(pTHX_ UV c)
1282 return is_uni_space(c); /* XXX no locale support yet */
1286 Perl_is_uni_digit_lc(pTHX_ UV c)
1288 return is_uni_digit(c); /* XXX no locale support yet */
1292 Perl_is_uni_upper_lc(pTHX_ UV c)
1294 return is_uni_upper(c); /* XXX no locale support yet */
1298 Perl_is_uni_lower_lc(pTHX_ UV c)
1300 return is_uni_lower(c); /* XXX no locale support yet */
1304 Perl_is_uni_cntrl_lc(pTHX_ UV c)
1306 return is_uni_cntrl(c); /* XXX no locale support yet */
1310 Perl_is_uni_graph_lc(pTHX_ UV c)
1312 return is_uni_graph(c); /* XXX no locale support yet */
1316 Perl_is_uni_print_lc(pTHX_ UV c)
1318 return is_uni_print(c); /* XXX no locale support yet */
1322 Perl_is_uni_punct_lc(pTHX_ UV c)
1324 return is_uni_punct(c); /* XXX no locale support yet */
1328 Perl_is_uni_xdigit_lc(pTHX_ UV c)
1330 return is_uni_xdigit(c); /* XXX no locale support yet */
1334 Perl_to_uni_upper_lc(pTHX_ U32 c)
1336 /* XXX returns only the first character -- do not use XXX */
1337 /* XXX no locale support yet */
1339 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
1340 return (U32)to_uni_upper(c, tmpbuf, &len);
1344 Perl_to_uni_title_lc(pTHX_ U32 c)
1346 /* XXX returns only the first character XXX -- do not use XXX */
1347 /* XXX no locale support yet */
1349 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
1350 return (U32)to_uni_title(c, tmpbuf, &len);
1354 Perl_to_uni_lower_lc(pTHX_ U32 c)
1356 /* XXX returns only the first character -- do not use XXX */
1357 /* XXX no locale support yet */
1359 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
1360 return (U32)to_uni_lower(c, tmpbuf, &len);
1364 S_is_utf8_common(pTHX_ const U8 *const p, SV **swash,
1365 const char *const swashname)
1369 PERL_ARGS_ASSERT_IS_UTF8_COMMON;
1371 if (!is_utf8_char(p))
1374 *swash = swash_init("utf8", swashname, &PL_sv_undef, 1, 0);
1375 return swash_fetch(*swash, p, TRUE) != 0;
1379 Perl_is_utf8_alnum(pTHX_ const U8 *p)
1383 PERL_ARGS_ASSERT_IS_UTF8_ALNUM;
1385 /* NOTE: "IsWord", not "IsAlnum", since Alnum is a true
1386 * descendant of isalnum(3), in other words, it doesn't
1387 * contain the '_'. --jhi */
1388 return is_utf8_common(p, &PL_utf8_alnum, "IsWord");
1392 Perl_is_utf8_idfirst(pTHX_ const U8 *p) /* The naming is historical. */
1396 PERL_ARGS_ASSERT_IS_UTF8_IDFIRST;
1400 /* is_utf8_idstart would be more logical. */
1401 return is_utf8_common(p, &PL_utf8_idstart, "IdStart");
1405 Perl_is_utf8_idcont(pTHX_ const U8 *p)
1409 PERL_ARGS_ASSERT_IS_UTF8_IDCONT;
1413 return is_utf8_common(p, &PL_utf8_idcont, "IdContinue");
1417 Perl_is_utf8_alpha(pTHX_ const U8 *p)
1421 PERL_ARGS_ASSERT_IS_UTF8_ALPHA;
1423 return is_utf8_common(p, &PL_utf8_alpha, "IsAlpha");
1427 Perl_is_utf8_ascii(pTHX_ const U8 *p)
1431 PERL_ARGS_ASSERT_IS_UTF8_ASCII;
1433 return is_utf8_common(p, &PL_utf8_ascii, "IsAscii");
1437 Perl_is_utf8_space(pTHX_ const U8 *p)
1441 PERL_ARGS_ASSERT_IS_UTF8_SPACE;
1443 return is_utf8_common(p, &PL_utf8_space, "IsSpacePerl");
1447 Perl_is_utf8_perl_space(pTHX_ const U8 *p)
1451 PERL_ARGS_ASSERT_IS_UTF8_PERL_SPACE;
1453 return is_utf8_common(p, &PL_utf8_perl_space, "IsPerlSpace");
1457 Perl_is_utf8_perl_word(pTHX_ const U8 *p)
1461 PERL_ARGS_ASSERT_IS_UTF8_PERL_WORD;
1463 return is_utf8_common(p, &PL_utf8_perl_word, "IsPerlWord");
1467 Perl_is_utf8_digit(pTHX_ const U8 *p)
1471 PERL_ARGS_ASSERT_IS_UTF8_DIGIT;
1473 return is_utf8_common(p, &PL_utf8_digit, "IsDigit");
1477 Perl_is_utf8_posix_digit(pTHX_ const U8 *p)
1481 PERL_ARGS_ASSERT_IS_UTF8_POSIX_DIGIT;
1483 return is_utf8_common(p, &PL_utf8_posix_digit, "IsPosixDigit");
1487 Perl_is_utf8_upper(pTHX_ const U8 *p)
1491 PERL_ARGS_ASSERT_IS_UTF8_UPPER;
1493 return is_utf8_common(p, &PL_utf8_upper, "IsUppercase");
1497 Perl_is_utf8_lower(pTHX_ const U8 *p)
1501 PERL_ARGS_ASSERT_IS_UTF8_LOWER;
1503 return is_utf8_common(p, &PL_utf8_lower, "IsLowercase");
1507 Perl_is_utf8_cntrl(pTHX_ const U8 *p)
1511 PERL_ARGS_ASSERT_IS_UTF8_CNTRL;
1513 return is_utf8_common(p, &PL_utf8_cntrl, "IsCntrl");
1517 Perl_is_utf8_graph(pTHX_ const U8 *p)
1521 PERL_ARGS_ASSERT_IS_UTF8_GRAPH;
1523 return is_utf8_common(p, &PL_utf8_graph, "IsGraph");
1527 Perl_is_utf8_print(pTHX_ const U8 *p)
1531 PERL_ARGS_ASSERT_IS_UTF8_PRINT;
1533 return is_utf8_common(p, &PL_utf8_print, "IsPrint");
1537 Perl_is_utf8_punct(pTHX_ const U8 *p)
1541 PERL_ARGS_ASSERT_IS_UTF8_PUNCT;
1543 return is_utf8_common(p, &PL_utf8_punct, "IsPunct");
1547 Perl_is_utf8_xdigit(pTHX_ const U8 *p)
1551 PERL_ARGS_ASSERT_IS_UTF8_XDIGIT;
1553 return is_utf8_common(p, &PL_utf8_xdigit, "IsXDigit");
1557 Perl_is_utf8_mark(pTHX_ const U8 *p)
1561 PERL_ARGS_ASSERT_IS_UTF8_MARK;
1563 return is_utf8_common(p, &PL_utf8_mark, "IsM");
1567 Perl_is_utf8_X_begin(pTHX_ const U8 *p)
1571 PERL_ARGS_ASSERT_IS_UTF8_X_BEGIN;
1573 return is_utf8_common(p, &PL_utf8_X_begin, "_X_Begin");
1577 Perl_is_utf8_X_extend(pTHX_ const U8 *p)
1581 PERL_ARGS_ASSERT_IS_UTF8_X_EXTEND;
1583 return is_utf8_common(p, &PL_utf8_X_extend, "_X_Extend");
1587 Perl_is_utf8_X_prepend(pTHX_ const U8 *p)
1591 PERL_ARGS_ASSERT_IS_UTF8_X_PREPEND;
1593 return is_utf8_common(p, &PL_utf8_X_prepend, "GCB=Prepend");
1597 Perl_is_utf8_X_non_hangul(pTHX_ const U8 *p)
1601 PERL_ARGS_ASSERT_IS_UTF8_X_NON_HANGUL;
1603 return is_utf8_common(p, &PL_utf8_X_non_hangul, "HST=Not_Applicable");
1607 Perl_is_utf8_X_L(pTHX_ const U8 *p)
1611 PERL_ARGS_ASSERT_IS_UTF8_X_L;
1613 return is_utf8_common(p, &PL_utf8_X_L, "GCB=L");
1617 Perl_is_utf8_X_LV(pTHX_ const U8 *p)
1621 PERL_ARGS_ASSERT_IS_UTF8_X_LV;
1623 return is_utf8_common(p, &PL_utf8_X_LV, "GCB=LV");
1627 Perl_is_utf8_X_LVT(pTHX_ const U8 *p)
1631 PERL_ARGS_ASSERT_IS_UTF8_X_LVT;
1633 return is_utf8_common(p, &PL_utf8_X_LVT, "GCB=LVT");
1637 Perl_is_utf8_X_T(pTHX_ const U8 *p)
1641 PERL_ARGS_ASSERT_IS_UTF8_X_T;
1643 return is_utf8_common(p, &PL_utf8_X_T, "GCB=T");
1647 Perl_is_utf8_X_V(pTHX_ const U8 *p)
1651 PERL_ARGS_ASSERT_IS_UTF8_X_V;
1653 return is_utf8_common(p, &PL_utf8_X_V, "GCB=V");
1657 Perl_is_utf8_X_LV_LVT_V(pTHX_ const U8 *p)
1661 PERL_ARGS_ASSERT_IS_UTF8_X_LV_LVT_V;
1663 return is_utf8_common(p, &PL_utf8_X_LV_LVT_V, "_X_LV_LVT_V");
1667 =for apidoc to_utf8_case
1669 The "p" contains the pointer to the UTF-8 string encoding
1670 the character that is being converted.
1672 The "ustrp" is a pointer to the character buffer to put the
1673 conversion result to. The "lenp" is a pointer to the length
1676 The "swashp" is a pointer to the swash to use.
1678 Both the special and normal mappings are stored lib/unicore/To/Foo.pl,
1679 and loaded by SWASHNEW, using lib/utf8_heavy.pl. The special (usually,
1680 but not always, a multicharacter mapping), is tried first.
1682 The "special" is a string like "utf8::ToSpecLower", which means the
1683 hash %utf8::ToSpecLower. The access to the hash is through
1684 Perl_to_utf8_case().
1686 The "normal" is a string like "ToLower" which means the swash
1692 Perl_to_utf8_case(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp,
1693 SV **swashp, const char *normal, const char *special)
1696 U8 tmpbuf[UTF8_MAXBYTES_CASE+1];
1698 const UV uv0 = utf8_to_uvchr(p, NULL);
1699 /* The NATIVE_TO_UNI() and UNI_TO_NATIVE() mappings
1700 * are necessary in EBCDIC, they are redundant no-ops
1701 * in ASCII-ish platforms, and hopefully optimized away. */
1702 const UV uv1 = NATIVE_TO_UNI(uv0);
1704 PERL_ARGS_ASSERT_TO_UTF8_CASE;
1706 uvuni_to_utf8(tmpbuf, uv1);
1708 if (!*swashp) /* load on-demand */
1709 *swashp = swash_init("utf8", normal, &PL_sv_undef, 4, 0);
1710 /* This is the beginnings of a skeleton of code to read the info section
1711 * that is in all the swashes in case we ever want to do that, so one can
1712 * read things whose maps aren't code points, and whose default if missing
1713 * is not to the code point itself. This was just to see if it actually
1714 * worked. Details on what the possibilities are are in perluniprops.pod
1715 HV * const hv = get_hv("utf8::SwashInfo", 0);
1718 svp = hv_fetch(hv, (const char*)normal, strlen(normal), FALSE);
1721 HV * const this_hash = SvRV(*svp);
1722 svp = hv_fetch(this_hash, "type", strlen("type"), FALSE);
1723 s = SvPV_const(*svp, len);
1728 /* It might be "special" (sometimes, but not always,
1729 * a multicharacter mapping) */
1730 HV * const hv = get_hv(special, 0);
1734 (svp = hv_fetch(hv, (const char*)tmpbuf, UNISKIP(uv1), FALSE)) &&
1738 s = SvPV_const(*svp, len);
1740 len = uvuni_to_utf8(ustrp, NATIVE_TO_UNI(*(U8*)s)) - ustrp;
1743 /* If we have EBCDIC we need to remap the characters
1744 * since any characters in the low 256 are Unicode
1745 * code points, not EBCDIC. */
1746 U8 *t = (U8*)s, *tend = t + len, *d;
1753 const UV c = utf8_to_uvchr(t, &tlen);
1755 d = uvchr_to_utf8(d, UNI_TO_NATIVE(c));
1764 d = uvchr_to_utf8(d, UNI_TO_NATIVE(*t));
1769 Copy(tmpbuf, ustrp, len, U8);
1771 Copy(s, ustrp, len, U8);
1777 if (!len && *swashp) {
1778 const UV uv2 = swash_fetch(*swashp, tmpbuf, TRUE);
1781 /* It was "normal" (a single character mapping). */
1782 const UV uv3 = UNI_TO_NATIVE(uv2);
1783 len = uvchr_to_utf8(ustrp, uv3) - ustrp;
1787 if (!len) /* Neither: just copy. In other words, there was no mapping
1788 defined, which means that the code point maps to itself */
1789 len = uvchr_to_utf8(ustrp, uv0) - ustrp;
1794 return len ? utf8_to_uvchr(ustrp, 0) : 0;
1798 =for apidoc to_utf8_upper
1800 Convert the UTF-8 encoded character at p to its uppercase version and
1801 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1802 that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since
1803 the uppercase version may be longer than the original character.
1805 The first character of the uppercased version is returned
1806 (but note, as explained above, that there may be more.)
1811 Perl_to_utf8_upper(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp)
1815 PERL_ARGS_ASSERT_TO_UTF8_UPPER;
1817 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1818 &PL_utf8_toupper, "ToUpper", "utf8::ToSpecUpper");
1822 =for apidoc to_utf8_title
1824 Convert the UTF-8 encoded character at p to its titlecase version and
1825 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1826 that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the
1827 titlecase version may be longer than the original character.
1829 The first character of the titlecased version is returned
1830 (but note, as explained above, that there may be more.)
1835 Perl_to_utf8_title(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp)
1839 PERL_ARGS_ASSERT_TO_UTF8_TITLE;
1841 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1842 &PL_utf8_totitle, "ToTitle", "utf8::ToSpecTitle");
1846 =for apidoc to_utf8_lower
1848 Convert the UTF-8 encoded character at p to its lowercase version and
1849 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1850 that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the
1851 lowercase version may be longer than the original character.
1853 The first character of the lowercased version is returned
1854 (but note, as explained above, that there may be more.)
1859 Perl_to_utf8_lower(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp)
1863 PERL_ARGS_ASSERT_TO_UTF8_LOWER;
1865 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1866 &PL_utf8_tolower, "ToLower", "utf8::ToSpecLower");
1870 =for apidoc to_utf8_fold
1872 Convert the UTF-8 encoded character at p to its foldcase version and
1873 store that in UTF-8 in ustrp and its length in bytes in lenp. Note
1874 that the ustrp needs to be at least UTF8_MAXBYTES_CASE+1 bytes since the
1875 foldcase version may be longer than the original character (up to
1878 The first character of the foldcased version is returned
1879 (but note, as explained above, that there may be more.)
1884 Perl_to_utf8_fold(pTHX_ const U8 *p, U8* ustrp, STRLEN *lenp)
1888 PERL_ARGS_ASSERT_TO_UTF8_FOLD;
1890 return Perl_to_utf8_case(aTHX_ p, ustrp, lenp,
1891 &PL_utf8_tofold, "ToFold", "utf8::ToSpecFold");
1895 * A "swash" is a swatch hash.
1896 * A "swatch" is a bit vector generated by utf8.c:S_swash_get().
1897 * C<pkg> is a pointer to a package name for SWASHNEW, should be "utf8".
1898 * For other parameters, see utf8::SWASHNEW in lib/utf8_heavy.pl.
1901 Perl_swash_init(pTHX_ const char* pkg, const char* name, SV *listsv, I32 minbits, I32 none)
1906 const size_t pkg_len = strlen(pkg);
1907 const size_t name_len = strlen(name);
1908 HV * const stash = gv_stashpvn(pkg, pkg_len, 0);
1911 PERL_ARGS_ASSERT_SWASH_INIT;
1913 PUSHSTACKi(PERLSI_MAGIC);
1917 if (!gv_fetchmeth(stash, "SWASHNEW", 8, -1)) { /* demand load utf8 */
1919 errsv_save = newSVsv(ERRSV);
1920 /* It is assumed that callers of this routine are not passing in any
1921 user derived data. */
1922 /* Need to do this after save_re_context() as it will set PL_tainted to
1923 1 while saving $1 etc (see the code after getrx: in Perl_magic_get).
1924 Even line to create errsv_save can turn on PL_tainted. */
1925 SAVEBOOL(PL_tainted);
1927 Perl_load_module(aTHX_ PERL_LOADMOD_NOIMPORT, newSVpvn(pkg,pkg_len),
1930 sv_setsv(ERRSV, errsv_save);
1931 SvREFCNT_dec(errsv_save);
1937 mPUSHp(pkg, pkg_len);
1938 mPUSHp(name, name_len);
1943 errsv_save = newSVsv(ERRSV);
1944 if (call_method("SWASHNEW", G_SCALAR))
1945 retval = newSVsv(*PL_stack_sp--);
1947 retval = &PL_sv_undef;
1949 sv_setsv(ERRSV, errsv_save);
1950 SvREFCNT_dec(errsv_save);
1953 if (IN_PERL_COMPILETIME) {
1954 CopHINTS_set(PL_curcop, PL_hints);
1956 if (!SvROK(retval) || SvTYPE(SvRV(retval)) != SVt_PVHV) {
1958 Perl_croak(aTHX_ "Can't find Unicode property definition \"%"SVf"\"",
1960 Perl_croak(aTHX_ "SWASHNEW didn't return an HV ref");
1966 /* This API is wrong for special case conversions since we may need to
1967 * return several Unicode characters for a single Unicode character
1968 * (see lib/unicore/SpecCase.txt) The SWASHGET in lib/utf8_heavy.pl is
1969 * the lower-level routine, and it is similarly broken for returning
1970 * multiple values. --jhi */
1971 /* Now SWASHGET is recasted into S_swash_get in this file. */
1974 * Returns the value of property/mapping C<swash> for the first character
1975 * of the string C<ptr>. If C<do_utf8> is true, the string C<ptr> is
1976 * assumed to be in utf8. If C<do_utf8> is false, the string C<ptr> is
1977 * assumed to be in native 8-bit encoding. Caches the swatch in C<swash>.
1980 Perl_swash_fetch(pTHX_ SV *swash, const U8 *ptr, bool do_utf8)
1983 HV *const hv = MUTABLE_HV(SvRV(swash));
1988 const U8 *tmps = NULL;
1992 const UV c = NATIVE_TO_ASCII(*ptr);
1994 PERL_ARGS_ASSERT_SWASH_FETCH;
1996 if (!do_utf8 && !UNI_IS_INVARIANT(c)) {
1997 tmputf8[0] = (U8)UTF8_EIGHT_BIT_HI(c);
1998 tmputf8[1] = (U8)UTF8_EIGHT_BIT_LO(c);
2001 /* Given a UTF-X encoded char 0xAA..0xYY,0xZZ
2002 * then the "swatch" is a vec() for all the chars which start
2004 * So the key in the hash (klen) is length of encoded char -1
2006 klen = UTF8SKIP(ptr) - 1;
2010 /* If char is invariant then swatch is for all the invariant chars
2011 * In both UTF-8 and UTF-8-MOD that happens to be UTF_CONTINUATION_MARK
2013 needents = UTF_CONTINUATION_MARK;
2014 off = NATIVE_TO_UTF(ptr[klen]);
2017 /* If char is encoded then swatch is for the prefix */
2018 needents = (1 << UTF_ACCUMULATION_SHIFT);
2019 off = NATIVE_TO_UTF(ptr[klen]) & UTF_CONTINUATION_MASK;
2023 * This single-entry cache saves about 1/3 of the utf8 overhead in test
2024 * suite. (That is, only 7-8% overall over just a hash cache. Still,
2025 * it's nothing to sniff at.) Pity we usually come through at least
2026 * two function calls to get here...
2028 * NB: this code assumes that swatches are never modified, once generated!
2031 if (hv == PL_last_swash_hv &&
2032 klen == PL_last_swash_klen &&
2033 (!klen || memEQ((char *)ptr, (char *)PL_last_swash_key, klen)) )
2035 tmps = PL_last_swash_tmps;
2036 slen = PL_last_swash_slen;
2039 /* Try our second-level swatch cache, kept in a hash. */
2040 SV** svp = hv_fetch(hv, (const char*)ptr, klen, FALSE);
2042 /* If not cached, generate it via swash_get */
2043 if (!svp || !SvPOK(*svp)
2044 || !(tmps = (const U8*)SvPV_const(*svp, slen))) {
2045 /* We use utf8n_to_uvuni() as we want an index into
2046 Unicode tables, not a native character number.
2048 const UV code_point = utf8n_to_uvuni(ptr, UTF8_MAXBYTES, 0,
2050 0 : UTF8_ALLOW_ANY);
2051 swatch = swash_get(swash,
2052 /* On EBCDIC & ~(0xA0-1) isn't a useful thing to do */
2053 (klen) ? (code_point & ~(needents - 1)) : 0,
2056 if (IN_PERL_COMPILETIME)
2057 CopHINTS_set(PL_curcop, PL_hints);
2059 svp = hv_store(hv, (const char *)ptr, klen, swatch, 0);
2061 if (!svp || !(tmps = (U8*)SvPV(*svp, slen))
2062 || (slen << 3) < needents)
2063 Perl_croak(aTHX_ "panic: swash_fetch got improper swatch");
2066 PL_last_swash_hv = hv;
2067 assert(klen <= sizeof(PL_last_swash_key));
2068 PL_last_swash_klen = (U8)klen;
2069 /* FIXME change interpvar.h? */
2070 PL_last_swash_tmps = (U8 *) tmps;
2071 PL_last_swash_slen = slen;
2073 Copy(ptr, PL_last_swash_key, klen, U8);
2076 switch ((int)((slen << 3) / needents)) {
2078 bit = 1 << (off & 7);
2080 return (tmps[off] & bit) != 0;
2085 return (tmps[off] << 8) + tmps[off + 1] ;
2088 return (tmps[off] << 24) + (tmps[off+1] << 16) + (tmps[off+2] << 8) + tmps[off + 3] ;
2090 Perl_croak(aTHX_ "panic: swash_fetch got swatch of unexpected bit width");
2091 NORETURN_FUNCTION_END;
2094 /* Read a single line of the main body of the swash input text. These are of
2097 * where each number is hex. The first two numbers form the minimum and
2098 * maximum of a range, and the third is the value associated with the range.
2099 * Not all swashes should have a third number
2101 * On input: l points to the beginning of the line to be examined; it points
2102 * to somewhere in the string of the whole input text, and is
2103 * terminated by a \n or the null string terminator.
2104 * lend points to the null terminator of that string
2105 * wants_value is non-zero if the swash expects a third number
2106 * typestr is the name of the swash's mapping, like 'ToLower'
2107 * On output: *min, *max, and *val are set to the values read from the line.
2108 * returns a pointer just beyond the line examined. If there was no
2109 * valid min number on the line, returns lend+1
2113 S_swash_scan_list_line(pTHX_ U8* l, U8* const lend, UV* min, UV* max, UV* val,
2114 const bool wants_value, const U8* const typestr)
2116 const int typeto = typestr[0] == 'T' && typestr[1] == 'o';
2117 STRLEN numlen; /* Length of the number */
2118 I32 flags = PERL_SCAN_SILENT_ILLDIGIT | PERL_SCAN_DISALLOW_PREFIX;
2120 /* nl points to the next \n in the scan */
2121 U8* const nl = (U8*)memchr(l, '\n', lend - l);
2123 /* Get the first number on the line: the range minimum */
2125 *min = grok_hex((char *)l, &numlen, &flags, NULL);
2126 if (numlen) /* If found a hex number, position past it */
2128 else if (nl) { /* Else, go handle next line, if any */
2129 return nl + 1; /* 1 is length of "\n" */
2131 else { /* Else, no next line */
2132 return lend + 1; /* to LIST's end at which \n is not found */
2135 /* The max range value follows, separated by a BLANK */
2138 flags = PERL_SCAN_SILENT_ILLDIGIT | PERL_SCAN_DISALLOW_PREFIX;
2140 *max = grok_hex((char *)l, &numlen, &flags, NULL);
2143 else /* If no value here, it is a single element range */
2146 /* Non-binary tables have a third entry: what the first element of the
2151 flags = PERL_SCAN_SILENT_ILLDIGIT |
2152 PERL_SCAN_DISALLOW_PREFIX;
2154 *val = grok_hex((char *)l, &numlen, &flags, NULL);
2163 Perl_croak(aTHX_ "%s: illegal mapping '%s'",
2169 *val = 0; /* bits == 1, then any val should be ignored */
2171 else { /* Nothing following range min, should be single element with no
2177 Perl_croak(aTHX_ "%s: illegal mapping '%s'", typestr, l);
2181 *val = 0; /* bits == 1, then val should be ignored */
2184 /* Position to next line if any, or EOF */
2194 * Returns a swatch (a bit vector string) for a code point sequence
2195 * that starts from the value C<start> and comprises the number C<span>.
2196 * A C<swash> must be an object created by SWASHNEW (see lib/utf8_heavy.pl).
2197 * Should be used via swash_fetch, which will cache the swatch in C<swash>.
2200 S_swash_get(pTHX_ SV* swash, UV start, UV span)
2203 U8 *l, *lend, *x, *xend, *s;
2204 STRLEN lcur, xcur, scur;
2205 HV *const hv = MUTABLE_HV(SvRV(swash));
2207 /* The string containing the main body of the table */
2208 SV** const listsvp = hv_fetchs(hv, "LIST", FALSE);
2210 SV** const typesvp = hv_fetchs(hv, "TYPE", FALSE);
2211 SV** const bitssvp = hv_fetchs(hv, "BITS", FALSE);
2212 SV** const nonesvp = hv_fetchs(hv, "NONE", FALSE);
2213 SV** const extssvp = hv_fetchs(hv, "EXTRAS", FALSE);
2214 const U8* const typestr = (U8*)SvPV_nolen(*typesvp);
2215 const STRLEN bits = SvUV(*bitssvp);
2216 const STRLEN octets = bits >> 3; /* if bits == 1, then octets == 0 */
2217 const UV none = SvUV(*nonesvp);
2218 const UV end = start + span;
2220 PERL_ARGS_ASSERT_SWASH_GET;
2222 if (bits != 1 && bits != 8 && bits != 16 && bits != 32) {
2223 Perl_croak(aTHX_ "panic: swash_get doesn't expect bits %"UVuf,
2227 /* create and initialize $swatch */
2228 scur = octets ? (span * octets) : (span + 7) / 8;
2229 swatch = newSV(scur);
2231 s = (U8*)SvPVX(swatch);
2232 if (octets && none) {
2233 const U8* const e = s + scur;
2236 *s++ = (U8)(none & 0xff);
2237 else if (bits == 16) {
2238 *s++ = (U8)((none >> 8) & 0xff);
2239 *s++ = (U8)( none & 0xff);
2241 else if (bits == 32) {
2242 *s++ = (U8)((none >> 24) & 0xff);
2243 *s++ = (U8)((none >> 16) & 0xff);
2244 *s++ = (U8)((none >> 8) & 0xff);
2245 *s++ = (U8)( none & 0xff);
2251 (void)memzero((U8*)s, scur + 1);
2253 SvCUR_set(swatch, scur);
2254 s = (U8*)SvPVX(swatch);
2256 /* read $swash->{LIST} */
2257 l = (U8*)SvPV(*listsvp, lcur);
2261 l = S_swash_scan_list_line(aTHX_ l, lend, &min, &max, &val,
2262 cBOOL(octets), typestr);
2267 /* If looking for something beyond this range, go try the next one */
2274 if (!none || val < none) {
2279 for (key = min; key <= max; key++) {
2283 /* offset must be non-negative (start <= min <= key < end) */
2284 offset = octets * (key - start);
2286 s[offset] = (U8)(val & 0xff);
2287 else if (bits == 16) {
2288 s[offset ] = (U8)((val >> 8) & 0xff);
2289 s[offset + 1] = (U8)( val & 0xff);
2291 else if (bits == 32) {
2292 s[offset ] = (U8)((val >> 24) & 0xff);
2293 s[offset + 1] = (U8)((val >> 16) & 0xff);
2294 s[offset + 2] = (U8)((val >> 8) & 0xff);
2295 s[offset + 3] = (U8)( val & 0xff);
2298 if (!none || val < none)
2302 else { /* bits == 1, then val should be ignored */
2306 for (key = min; key <= max; key++) {
2307 const STRLEN offset = (STRLEN)(key - start);
2310 s[offset >> 3] |= 1 << (offset & 7);
2316 /* read $swash->{EXTRAS} */
2317 x = (U8*)SvPV(*extssvp, xcur);
2325 SV **otherbitssvp, *other;
2329 const U8 opc = *x++;
2333 nl = (U8*)memchr(x, '\n', xend - x);
2335 if (opc != '-' && opc != '+' && opc != '!' && opc != '&') {
2337 x = nl + 1; /* 1 is length of "\n" */
2341 x = xend; /* to EXTRAS' end at which \n is not found */
2348 namelen = nl - namestr;
2352 namelen = xend - namestr;
2356 othersvp = hv_fetch(hv, (char *)namestr, namelen, FALSE);
2357 otherhv = MUTABLE_HV(SvRV(*othersvp));
2358 otherbitssvp = hv_fetchs(otherhv, "BITS", FALSE);
2359 otherbits = (STRLEN)SvUV(*otherbitssvp);
2360 if (bits < otherbits)
2361 Perl_croak(aTHX_ "panic: swash_get found swatch size mismatch");
2363 /* The "other" swatch must be destroyed after. */
2364 other = swash_get(*othersvp, start, span);
2365 o = (U8*)SvPV(other, olen);
2368 Perl_croak(aTHX_ "panic: swash_get got improper swatch");
2370 s = (U8*)SvPV(swatch, slen);
2371 if (bits == 1 && otherbits == 1) {
2373 Perl_croak(aTHX_ "panic: swash_get found swatch length mismatch");
2397 STRLEN otheroctets = otherbits >> 3;
2399 U8* const send = s + slen;
2404 if (otherbits == 1) {
2405 otherval = (o[offset >> 3] >> (offset & 7)) & 1;
2409 STRLEN vlen = otheroctets;
2417 if (opc == '+' && otherval)
2418 NOOP; /* replace with otherval */
2419 else if (opc == '!' && !otherval)
2421 else if (opc == '-' && otherval)
2423 else if (opc == '&' && !otherval)
2426 s += octets; /* no replacement */
2431 *s++ = (U8)( otherval & 0xff);
2432 else if (bits == 16) {
2433 *s++ = (U8)((otherval >> 8) & 0xff);
2434 *s++ = (U8)( otherval & 0xff);
2436 else if (bits == 32) {
2437 *s++ = (U8)((otherval >> 24) & 0xff);
2438 *s++ = (U8)((otherval >> 16) & 0xff);
2439 *s++ = (U8)((otherval >> 8) & 0xff);
2440 *s++ = (U8)( otherval & 0xff);
2444 sv_free(other); /* through with it! */
2450 Perl__swash_inversion_hash(pTHX_ SV* swash)
2453 /* Subject to change or removal. For use only in one place in regexec.c
2455 * Returns a hash which is the inversion and closure of a swash mapping.
2456 * For example, consider the input lines:
2461 * The returned hash would have two keys, the utf8 for 006B and the utf8 for
2462 * 006C. The value for each key is an array. For 006C, the array would
2463 * have a two elements, the utf8 for itself, and for 004C. For 006B, there
2464 * would be three elements in its array, the utf8 for 006B, 004B and 212A.
2466 * Essentially, for any code point, it gives all the code points that map to
2467 * it, or the list of 'froms' for that point.
2469 * Currently it only looks at the main body of the swash, and ignores any
2470 * additions or deletions from other swashes */
2474 HV *const hv = MUTABLE_HV(SvRV(swash));
2476 /* The string containing the main body of the table */
2477 SV** const listsvp = hv_fetchs(hv, "LIST", FALSE);
2479 SV** const typesvp = hv_fetchs(hv, "TYPE", FALSE);
2480 SV** const bitssvp = hv_fetchs(hv, "BITS", FALSE);
2481 SV** const nonesvp = hv_fetchs(hv, "NONE", FALSE);
2482 /*SV** const extssvp = hv_fetchs(hv, "EXTRAS", FALSE);*/
2483 const U8* const typestr = (U8*)SvPV_nolen(*typesvp);
2484 const STRLEN bits = SvUV(*bitssvp);
2485 const STRLEN octets = bits >> 3; /* if bits == 1, then octets == 0 */
2486 const UV none = SvUV(*nonesvp);
2490 PERL_ARGS_ASSERT__SWASH_INVERSION_HASH;
2492 /* Must have at least 8 bits to get the mappings */
2493 if (bits != 8 && bits != 16 && bits != 32) {
2494 Perl_croak(aTHX_ "panic: swash_inversion_hash doesn't expect bits %"UVuf,
2498 /* read $swash->{LIST} */
2499 l = (U8*)SvPV(*listsvp, lcur);
2502 /* Go through each input line */
2506 l = S_swash_scan_list_line(aTHX_ l, lend, &min, &max, &val,
2507 cBOOL(octets), typestr);
2512 /* Each element in the range is to be inverted */
2513 for (inverse = min; inverse <= max; inverse++) {
2518 bool found_key = FALSE;
2520 /* The key is the inverse mapping */
2521 char key[UTF8_MAXBYTES+1];
2522 char* key_end = (char *) uvuni_to_utf8((U8*) key, val);
2523 STRLEN key_len = key_end - key;
2525 /* And the value is what the forward mapping is from. */
2526 char utf8_inverse[UTF8_MAXBYTES+1];
2527 char *utf8_inverse_end = (char *) uvuni_to_utf8((U8*) utf8_inverse, inverse);
2529 /* Get the list for the map */
2530 if ((listp = hv_fetch(ret, key, key_len, FALSE))) {
2531 list = (AV*) *listp;
2533 else { /* No entry yet for it: create one */
2535 if (! hv_store(ret, key, key_len, (SV*) list, FALSE)) {
2536 Perl_croak(aTHX_ "panic: hv_store() unexpectedly failed");
2540 for (i = 0; i < av_len(list); i++) {
2541 SV** entryp = av_fetch(list, i, FALSE);
2543 if (entryp == NULL) {
2544 Perl_croak(aTHX_ "panic: av_fetch() unexpectedly failed");
2547 if (SvCUR(entry) != key_len) {
2550 if (memEQ(key, SvPVX(entry), key_len)) {
2556 element = newSVpvn_flags(key, key_len, SVf_UTF8);
2557 av_push(list, element);
2561 /* Simply add the value to the list */
2562 element = newSVpvn_flags(utf8_inverse, utf8_inverse_end - utf8_inverse, SVf_UTF8);
2563 av_push(list, element);
2565 /* swash_get() increments the value of val for each element in the
2566 * range. That makes more compact tables possible. You can
2567 * express the capitalization, for example, of all consecutive
2568 * letters with a single line: 0061\t007A\t0041 This maps 0061 to
2569 * 0041, 0062 to 0042, etc. I (khw) have never understood 'none',
2570 * and it's not documented, and perhaps not even currently used,
2571 * but I copied the semantics from swash_get(), just in case */
2572 if (!none || val < none) {
2582 =for apidoc uvchr_to_utf8
2584 Adds the UTF-8 representation of the Native codepoint C<uv> to the end
2585 of the string C<d>; C<d> should be have at least C<UTF8_MAXBYTES+1> free
2586 bytes available. The return value is the pointer to the byte after the
2587 end of the new character. In other words,
2589 d = uvchr_to_utf8(d, uv);
2591 is the recommended wide native character-aware way of saying
2598 /* On ASCII machines this is normally a macro but we want a
2599 real function in case XS code wants it
2602 Perl_uvchr_to_utf8(pTHX_ U8 *d, UV uv)
2604 PERL_ARGS_ASSERT_UVCHR_TO_UTF8;
2606 return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), 0);
2610 Perl_uvchr_to_utf8_flags(pTHX_ U8 *d, UV uv, UV flags)
2612 PERL_ARGS_ASSERT_UVCHR_TO_UTF8_FLAGS;
2614 return Perl_uvuni_to_utf8_flags(aTHX_ d, NATIVE_TO_UNI(uv), flags);
2618 =for apidoc utf8n_to_uvchr
2621 Returns the native character value of the first character in the string
2623 which is assumed to be in UTF-8 encoding; C<retlen> will be set to the
2624 length, in bytes, of that character.
2626 Allows length and flags to be passed to low level routine.
2630 /* On ASCII machines this is normally a macro but we want
2631 a real function in case XS code wants it
2634 Perl_utf8n_to_uvchr(pTHX_ const U8 *s, STRLEN curlen, STRLEN *retlen,
2637 const UV uv = Perl_utf8n_to_uvuni(aTHX_ s, curlen, retlen, flags);
2639 PERL_ARGS_ASSERT_UTF8N_TO_UVCHR;
2641 return UNI_TO_NATIVE(uv);
2645 =for apidoc pv_uni_display
2647 Build to the scalar dsv a displayable version of the string spv,
2648 length len, the displayable version being at most pvlim bytes long
2649 (if longer, the rest is truncated and "..." will be appended).
2651 The flags argument can have UNI_DISPLAY_ISPRINT set to display
2652 isPRINT()able characters as themselves, UNI_DISPLAY_BACKSLASH
2653 to display the \\[nrfta\\] as the backslashed versions (like '\n')
2654 (UNI_DISPLAY_BACKSLASH is preferred over UNI_DISPLAY_ISPRINT for \\).
2655 UNI_DISPLAY_QQ (and its alias UNI_DISPLAY_REGEX) have both
2656 UNI_DISPLAY_BACKSLASH and UNI_DISPLAY_ISPRINT turned on.
2658 The pointer to the PV of the dsv is returned.
2662 Perl_pv_uni_display(pTHX_ SV *dsv, const U8 *spv, STRLEN len, STRLEN pvlim, UV flags)
2667 PERL_ARGS_ASSERT_PV_UNI_DISPLAY;
2671 for (s = (const char *)spv, e = s + len; s < e; s += UTF8SKIP(s)) {
2673 /* This serves double duty as a flag and a character to print after
2674 a \ when flags & UNI_DISPLAY_BACKSLASH is true.
2678 if (pvlim && SvCUR(dsv) >= pvlim) {
2682 u = utf8_to_uvchr((U8*)s, 0);
2684 const unsigned char c = (unsigned char)u & 0xFF;
2685 if (flags & UNI_DISPLAY_BACKSLASH) {
2702 const char string = ok;
2703 sv_catpvs(dsv, "\\");
2704 sv_catpvn(dsv, &string, 1);
2707 /* isPRINT() is the locale-blind version. */
2708 if (!ok && (flags & UNI_DISPLAY_ISPRINT) && isPRINT(c)) {
2709 const char string = c;
2710 sv_catpvn(dsv, &string, 1);
2715 Perl_sv_catpvf(aTHX_ dsv, "\\x{%"UVxf"}", u);
2718 sv_catpvs(dsv, "...");
2724 =for apidoc sv_uni_display
2726 Build to the scalar dsv a displayable version of the scalar sv,
2727 the displayable version being at most pvlim bytes long
2728 (if longer, the rest is truncated and "..." will be appended).
2730 The flags argument is as in pv_uni_display().
2732 The pointer to the PV of the dsv is returned.
2737 Perl_sv_uni_display(pTHX_ SV *dsv, SV *ssv, STRLEN pvlim, UV flags)
2739 PERL_ARGS_ASSERT_SV_UNI_DISPLAY;
2741 return Perl_pv_uni_display(aTHX_ dsv, (const U8*)SvPVX_const(ssv),
2742 SvCUR(ssv), pvlim, flags);
2746 =for apidoc foldEQ_utf8
2748 Returns true if the leading portions of the strings s1 and s2 (either or both
2749 of which may be in UTF-8) are the same case-insensitively; false otherwise.
2750 How far into the strings to compare is determined by other input parameters.
2752 If u1 is true, the string s1 is assumed to be in UTF-8-encoded Unicode;
2753 otherwise it is assumed to be in native 8-bit encoding. Correspondingly for u2
2756 If the byte length l1 is non-zero, it says how far into s1 to check for fold
2757 equality. In other words, s1+l1 will be used as a goal to reach. The
2758 scan will not be considered to be a match unless the goal is reached, and
2759 scanning won't continue past that goal. Correspondingly for l2 with respect to
2762 If pe1 is non-NULL and the pointer it points to is not NULL, that pointer is
2763 considered an end pointer beyond which scanning of s1 will not continue under
2764 any circumstances. This means that if both l1 and pe1 are specified, and pe1
2765 is less than s1+l1, the match will never be successful because it can never
2766 get as far as its goal (and in fact is asserted against). Correspondingly for
2767 pe2 with respect to s2.
2769 At least one of s1 and s2 must have a goal (at least one of l1 and l2 must be
2770 non-zero), and if both do, both have to be
2771 reached for a successful match. Also, if the fold of a character is multiple
2772 characters, all of them must be matched (see tr21 reference below for
2775 Upon a successful match, if pe1 is non-NULL,
2776 it will be set to point to the beginning of the I<next> character of s1 beyond
2777 what was matched. Correspondingly for pe2 and s2.
2779 For case-insensitiveness, the "casefolding" of Unicode is used
2780 instead of upper/lowercasing both the characters, see
2781 http://www.unicode.org/unicode/reports/tr21/ (Case Mappings).
2785 Perl_foldEQ_utf8(pTHX_ const char *s1, char **pe1, register UV l1, bool u1, const char *s2, char **pe2, register UV l2, bool u2)
2788 register const U8 *p1 = (const U8*)s1; /* Point to current char */
2789 register const U8 *p2 = (const U8*)s2;
2790 register const U8 *g1 = NULL; /* goal for s1 */
2791 register const U8 *g2 = NULL;
2792 register const U8 *e1 = NULL; /* Don't scan s1 past this */
2793 register U8 *f1 = NULL; /* Point to current folded */
2794 register const U8 *e2 = NULL;
2795 register U8 *f2 = NULL;
2796 STRLEN n1 = 0, n2 = 0; /* Number of bytes in current char */
2797 U8 foldbuf1[UTF8_MAXBYTES_CASE+1];
2798 U8 foldbuf2[UTF8_MAXBYTES_CASE+1];
2799 U8 natbuf[2]; /* Holds native 8-bit char converted to utf8;
2800 these always fit in 2 bytes */
2802 PERL_ARGS_ASSERT_FOLDEQ_UTF8;
2809 g1 = (const U8*)s1 + l1;
2817 g2 = (const U8*)s2 + l2;
2820 /* Must have at least one goal */
2825 /* Will never match if goal is out-of-bounds */
2826 assert(! e1 || e1 >= g1);
2828 /* Here, there isn't an end pointer, or it is beyond the goal. We
2829 * only go as far as the goal */
2833 assert(e1); /* Must have an end for looking at s1 */
2836 /* Same for goal for s2 */
2838 assert(! e2 || e2 >= g2);
2845 /* Look through both strings, a character at a time */
2846 while (p1 < e1 && p2 < e2) {
2848 /* If at the beginning of a new character in s1, get its fold to use
2849 * and the length of the fold */
2852 to_utf8_fold(p1, foldbuf1, &n1);
2854 else { /* Not utf8, convert to it first and then get fold */
2855 uvuni_to_utf8(natbuf, (UV) NATIVE_TO_UNI(((UV)*p1)));
2856 to_utf8_fold(natbuf, foldbuf1, &n1);
2861 if (n2 == 0) { /* Same for s2 */
2863 to_utf8_fold(p2, foldbuf2, &n2);
2866 uvuni_to_utf8(natbuf, (UV) NATIVE_TO_UNI(((UV)*p2)));
2867 to_utf8_fold(natbuf, foldbuf2, &n2);
2872 /* While there is more to look for in both folds, see if they
2873 * continue to match */
2875 U8 fold_length = UTF8SKIP(f1);
2876 if (fold_length != UTF8SKIP(f2)
2877 || (fold_length == 1 && *f1 != *f2) /* Short circuit memNE
2878 function call for single
2880 || memNE((char*)f1, (char*)f2, fold_length))
2882 return 0; /* mismatch */
2885 /* Here, they matched, advance past them */
2892 /* When reach the end of any fold, advance the input past it */
2894 p1 += u1 ? UTF8SKIP(p1) : 1;
2897 p2 += u2 ? UTF8SKIP(p2) : 1;
2899 } /* End of loop through both strings */
2901 /* A match is defined by each scan that specified an explicit length
2902 * reaching its final goal, and the other not having matched a partial
2903 * character (which can happen when the fold of a character is more than one
2905 if (! ((g1 == 0 || p1 == g1) && (g2 == 0 || p2 == g2)) || n1 || n2) {
2909 /* Successful match. Set output pointers */
2921 * c-indentation-style: bsd
2923 * indent-tabs-mode: t
2926 * ex: set ts=8 sts=4 sw=4 noet: