-
Notifications
You must be signed in to change notification settings - Fork 576
/
Copy pathutf8.h
1353 lines (1109 loc) · 62.9 KB
/
utf8.h
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/* utf8.h
*
* This file contains definitions for use with the UTF-8 encoding. It
* actually also works with the variant UTF-8 encoding called UTF-EBCDIC, and
* hides almost all of the differences between these from the caller. In other
* words, someone should #include this file, and if the code is being compiled
* on an EBCDIC platform, things should mostly just work.
*
* Copyright (C) 2000, 2001, 2002, 2005, 2006, 2007, 2009,
* 2010, 2011 by Larry Wall and others
*
* You may distribute under the terms of either the GNU General Public
* License or the Artistic License, as specified in the README file.
*
* A note on nomenclature: The term UTF-8 is used loosely and inconsistently
* in Perl documentation. For one, perl uses an extension of UTF-8 to
* represent code points that Unicode considers illegal. For another, ASCII
* platform UTF-8 is usually conflated with EBCDIC platform UTF-EBCDIC, because
* outside some of the macros in this this file, the differences are hopefully
* invisible at the semantic level.
*
* UTF-EBCDIC has an isomorphic translation named I8 (for "Intermediate eight")
* which differs from UTF-8 only in a few details. It is often useful to
* translate UTF-EBCDIC into this form for processing. In general, macros and
* functions that are expecting their inputs to be either in I8 or UTF-8 are
* named UTF_foo (without an '8'), to indicate this. khw thinks it would be
* clearer if these were renamed to be I8, because UTF is the prefix for UTF16,
* U32, etc., and we use it only for 8 bit quantities; though on ASCII machines
* these are final, not intermediate, values. U8 would be more accurate, but
* bears too much resemblence to the ubiquitous U8 declaration. 'I8' stands
* out as very different from 'UTFn'.
*
* Unfortunately there are inconsistencies.
*
*/
#ifndef PERL_UTF8_H_ /* Guard against recursive inclusion */
#define PERL_UTF8_H_ 1
/*
=for apidoc Ay||utf8ness_t
This typedef is used by several core functions that return PV strings, to
indicate the UTF-8ness of those strings.
(If you write a new function, you probably should instead return the PV in an
SV with the UTF-8 flag of the SV properly set, rather than use this mechanism.)
The possible values this can be are:
=over
=item C<UTF8NESS_YES>
This means the string definitely should be treated as a sequence of
UTF-8-encoded characters.
Most code that needs to handle this typedef should be of the form:
if (utf8ness_flag == UTF8NESS_YES) {
treat as utf8; // like turning on an SV UTF-8 flag
}
=item C<UTF8NESS_NO>
This means the string definitely should be treated as a sequence of bytes, not
encoded as UTF-8.
=item C<UTF8NESS_IMMATERIAL>
This means it is equally valid to treat the string as bytes, or as UTF-8
characters; use whichever way you want. This happens when the string consists
entirely of characters which have the same representation whether encoded in
UTF-8 or not.
=item C<UTF8NESS_UNKNOWN>
This means it is unknown how the string should be treated. No core function
will ever return this value to a non-core caller. Instead, it is used by the
caller to initialize a variable to a non-legal value. A typical call will look like:
utf8ness_t string_is_utf8 = UTF8NESS_UNKNOWN
const char * string = foo(arg1, arg2, ..., &string_is_utf8);
if (string_is_utf8 == UTF8NESS_YES) {
do something for UTF-8;
}
=back
The following relationships hold between the enum values:
=over
=item S<C<0 E<lt>= I<enum value> E<lt>= UTF8NESS_IMMATERIAL>>
the string may be treated in code as non-UTF8
=item S<C<UTF8NESS_IMMATERIAL E<lt>= <I<enum value>>>
the string may be treated in code as encoded in UTF-8
=back
=cut
*/
typedef enum {
UTF8NESS_NO = 0, /* Definitely not UTF-8 */
UTF8NESS_IMMATERIAL = 1, /* Representation is the same in UTF-8 as
not, so the UTF8ness doesn't actually
matter */
UTF8NESS_YES = 2, /* Defintely is UTF-8, wideness
unspecified */
UTF8NESS_UNKNOWN = -1, /* Undetermined so far */
} utf8ness_t;
/* Use UTF-8 as the default script encoding?
* Turning this on will break scripts having non-UTF-8 binary
* data (such as Latin-1) in string literals. */
#ifdef USE_UTF8_SCRIPTS
# define USE_UTF8_IN_NAMES (!IN_BYTES)
#else
# define USE_UTF8_IN_NAMES (PL_hints & HINT_UTF8)
#endif
#include "regcharclass.h"
#include "unicode_constants.h"
/* For to_utf8_fold_flags, q.v. */
#define FOLD_FLAGS_LOCALE 0x1
#define FOLD_FLAGS_FULL 0x2
#define FOLD_FLAGS_NOMIX_ASCII 0x4
/*
=for apidoc_defn APRTdm|bool|is_ascii_string|NN const U8 * const s|STRLEN len
=for apidoc_defn APRTdm|bool|is_invariant_string|NN const U8 * const s|STRLEN len
=cut
*/
#define is_ascii_string(s, len) is_utf8_invariant_string(s, len)
#define is_invariant_string(s, len) is_utf8_invariant_string(s, len)
#define uvoffuni_to_utf8_flags(d,uv,flags) \
uvoffuni_to_utf8_flags_msgs(d, uv, flags, 0)
#define uvchr_to_utf8 uv_to_utf8
#define uvchr_to_utf8_flags uv_to_utf8_flags
#define uvchr_to_utf8_flags_msgs uv_to_utf8_msgs
#define Perl_uvchr_to_utf8 Perl_uv_to_utf8
#define Perl_uvchr_to_utf8_flags Perl_uv_to_utf8_flags
#define Perl_uvchr_to_utf8_flags_msgs Perl_uv_to_utf8_msgs
/* This is needed to cast the parameters for all those calls that had them
* improperly as chars */
#define utf8_to_uvchr_buf(s, e, lenp) \
Perl_utf8_to_uvchr_buf(aTHX_ (const U8 *) (s), (const U8 *) e, lenp)
#define Perl_utf8n_to_uvchr(s, len, lenp, flags) \
Perl_utf8n_to_uvchr_error(s, len, lenp, flags, 0)
#define Perl_utf8n_to_uvchr_error(s, len, lenp, flags, errors) \
Perl_utf8n_to_uvchr_msgs(s, len, lenp, flags, errors, 0)
#define Perl_utf8_to_uv( s, e, cp_p, advance_p) \
Perl_utf8_to_uv_flags( s, e, cp_p, advance_p, 0)
#define Perl_utf8_to_uv_flags( s, e, cp_p, advance_p, flags) \
Perl_utf8_to_uv_errors( s, e, cp_p, advance_p, flags, 0)
#define Perl_utf8_to_uv_errors( s, e, cp_p, advance_p, flags, errors) \
Perl_utf8_to_uv_msgs( s, e, cp_p, advance_p, flags, errors, 0)
#define Perl_extended_utf8_to_uv(s, e, cp_p, advance_p) \
Perl_utf8_to_uv(s, e, cp_p, advance_p)
#define Perl_strict_utf8_to_uv( s, e, cp_p, advance_p) \
Perl_utf8_to_uv_flags( s, e, cp_p, advance_p, \
UTF8_DISALLOW_ILLEGAL_INTERCHANGE)
#define Perl_c9strict_utf8_to_uv(s, e, cp_p, advance_p) \
Perl_utf8_to_uv_flags( s, e, cp_p, advance_p, \
UTF8_DISALLOW_ILLEGAL_C9_INTERCHANGE)
#define utf16_to_utf8(p, d, bytelen, newlen) \
utf16_to_utf8_base(p, d, bytelen, newlen, 0, 1)
#define utf16_to_utf8_reversed(p, d, bytelen, newlen) \
utf16_to_utf8_base(p, d, bytelen, newlen, 1, 0)
#define utf8_to_utf16(p, d, bytelen, newlen) \
utf8_to_utf16_base(p, d, bytelen, newlen, 0, 1)
#define utf8_to_utf16_reversed(p, d, bytelen, newlen) \
utf8_to_utf16_base(p, d, bytelen, newlen, 1, 0)
#define to_uni_fold(c, p, lenp) _to_uni_fold_flags(c, p, lenp, FOLD_FLAGS_FULL)
#define foldEQ_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2) \
foldEQ_utf8_flags(s1, pe1, l1, u1, s2, pe2, l2, u2, 0)
#define FOLDEQ_UTF8_NOMIX_ASCII (1 << 0)
#define FOLDEQ_LOCALE (1 << 1)
#define FOLDEQ_S1_ALREADY_FOLDED (1 << 2)
#define FOLDEQ_S2_ALREADY_FOLDED (1 << 3)
#define FOLDEQ_S1_FOLDS_SANE (1 << 4)
#define FOLDEQ_S2_FOLDS_SANE (1 << 5)
/* This will be described more fully below, but it turns out that the
* fundamental difference between UTF-8 and UTF-EBCDIC is that the former has
* the upper 2 bits of a continuation byte be '10', and the latter has the
* upper 3 bits be '101', leaving 6 and 5 significant bits respectively.
*
* It is helpful to know the EBCDIC value on ASCII platforms, mainly to avoid
* some #ifdef's */
#define UTF_EBCDIC_CONTINUATION_BYTE_INFO_BITS 5
/* See explanation below at 'UTF8_MAXBYTES' */
#define ASCII_PLATFORM_UTF8_MAXBYTES 13
#ifdef EBCDIC
/* The equivalent of the next few macros but implementing UTF-EBCDIC are in the
* following header file: */
# include "utfebcdic.h"
# else /* ! EBCDIC */
START_EXTERN_C
# ifndef DOINIT
EXTCONST unsigned char PL_utf8skip[];
# else
EXTCONST unsigned char PL_utf8skip[] = {
/* 0x00 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x10 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x20 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x30 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x40 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x50 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x60 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x70 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* ascii */
/* 0x80 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
/* 0x90 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
/* 0xA0 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
/* 0xB0 */ 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, /* bogus: continuation byte */
/* 0xC0 */ 2,2, /* overlong */
/* 0xC2 */ 2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* U+0080 to U+03FF */
/* 0xD0 */ 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /* U+0400 to U+07FF */
/* 0xE0 */ 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, /* U+0800 to U+FFFF */
/* 0xF0 */ 4,4,4,4,4,4,4,4,5,5,5,5,6,6, /* above BMP to 2**31 - 1 */
/* Perl extended (never was official UTF-8). Up to 36 bit */
/* 0xFE */ 7,
/* More extended, Up to 72 bits (64-bit + reserved) */
/* 0xFF */ ASCII_PLATFORM_UTF8_MAXBYTES
};
# endif
END_EXTERN_C
/*
=for apidoc Am|U8|NATIVE_TO_LATIN1|U8 ch
Returns the Latin-1 (including ASCII and control characters) equivalent of the
input native code point given by C<ch>. Thus, C<NATIVE_TO_LATIN1(193)> on
EBCDIC platforms returns 65. These each represent the character C<"A"> on
their respective platforms. On ASCII platforms no conversion is needed, so
this macro expands to just its input, adding no time nor space requirements to
the implementation.
For conversion of code points potentially larger than will fit in a character,
use L</NATIVE_TO_UNI>.
=for apidoc Am|U8|LATIN1_TO_NATIVE|U8 ch
Returns the native equivalent of the input Latin-1 code point (including ASCII
and control characters) given by C<ch>. Thus, C<LATIN1_TO_NATIVE(66)> on
EBCDIC platforms returns 194. These each represent the character C<"B"> on
their respective platforms. On ASCII platforms no conversion is needed, so
this macro expands to just its input, adding no time nor space requirements to
the implementation.
For conversion of code points potentially larger than will fit in a character,
use L</UNI_TO_NATIVE>.
=for apidoc Am|UV|NATIVE_TO_UNI|UV ch
Returns the Unicode equivalent of the input native code point given by C<ch>.
Thus, C<NATIVE_TO_UNI(195)> on EBCDIC platforms returns 67. These each
represent the character C<"C"> on their respective platforms. On ASCII
platforms no conversion is needed, so this macro expands to just its input,
adding no time nor space requirements to the implementation.
=for apidoc Am|UV|UNI_TO_NATIVE|UV ch
Returns the native equivalent of the input Unicode code point given by C<ch>.
Thus, C<UNI_TO_NATIVE(68)> on EBCDIC platforms returns 196. These each
represent the character C<"D"> on their respective platforms. On ASCII
platforms no conversion is needed, so this macro expands to just its input,
adding no time nor space requirements to the implementation.
=cut
*/
#define NATIVE_TO_LATIN1(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch)))
#define LATIN1_TO_NATIVE(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch)))
/* I8 is an intermediate version of UTF-8 used only in UTF-EBCDIC. We thus
* consider it to be identical to UTF-8 on ASCII platforms. Strictly speaking
* UTF-8 and UTF-EBCDIC are two different things, but we often conflate them
* because they are 8-bit encodings that serve the same purpose in Perl, and
* rarely do we need to distinguish them. The term "NATIVE_UTF8" applies to
* whichever one is applicable on the current platform */
#define NATIVE_UTF8_TO_I8(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch)))
#define I8_TO_NATIVE_UTF8(ch) (__ASSERT_(FITS_IN_8_BITS(ch)) ((U8) (ch)))
#define UNI_TO_NATIVE(ch) ((UV) ASSERT_NOT_PTR(ch))
#define NATIVE_TO_UNI(ch) ((UV) ASSERT_NOT_PTR(ch))
/*
The following table is from Unicode 3.2, plus the Perl extensions for above
U+10FFFF
Code Points 1st Byte 2nd Byte 3rd 4th 5th 6th 7th 8th-13th
U+0000..U+007F 00..7F
U+0080..U+07FF * C2..DF 80..BF
U+0800..U+0FFF E0 * A0..BF 80..BF
U+1000..U+CFFF E1..EC 80..BF 80..BF
U+D000..U+D7FF ED 80..9F 80..BF
U+D800..U+DFFF ED A0..BF 80..BF (surrogates)
U+E000..U+FFFF EE..EF 80..BF 80..BF
U+10000..U+3FFFF F0 * 90..BF 80..BF 80..BF
U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
Below are above-Unicode code points
U+110000..U+13FFFF F4 90..BF 80..BF 80..BF
U+110000..U+1FFFFF F5..F7 80..BF 80..BF 80..BF
U+200000..U+FFFFFF F8 * 88..BF 80..BF 80..BF 80..BF
U+1000000..U+3FFFFFF F9..FB 80..BF 80..BF 80..BF 80..BF
U+4000000..U+3FFFFFFF FC * 84..BF 80..BF 80..BF 80..BF 80..BF
U+40000000..U+7FFFFFFF FD 80..BF 80..BF 80..BF 80..BF 80..BF
U+80000000..U+FFFFFFFFF FE * 82..BF 80..BF 80..BF 80..BF 80..BF 80..BF
U+1000000000.. FF 80..BF 80..BF 80..BF 80..BF 80..BF * 81..BF 80..BF
Note the gaps before several of the byte entries above marked by '*'. These are
caused by legal UTF-8 avoiding non-shortest encodings: it is technically
possible to UTF-8-encode a single code point in different ways, but that is
explicitly forbidden, and the shortest possible encoding should always be used
(and that is what Perl does). The non-shortest ones are called 'overlongs'.
Another way to look at it, as bits:
Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
0aaa aaaa 0aaa aaaa
0000 0bbb bbaa aaaa 110b bbbb 10aa aaaa
cccc bbbb bbaa aaaa 1110 cccc 10bb bbbb 10aa aaaa
00 000d ddcc cccc bbbb bbaa aaaa 1111 0ddd 10cc cccc 10bb bbbb 10aa aaaa
As you can see, the continuation bytes all begin with C<10>, and the
leading bits of the start byte tell how many bytes there are in the
encoded character.
Perl's extended UTF-8 means we can have start bytes up through FF, though any
beginning with FF yields a code point that is too large for 32-bit ASCII
platforms. FF signals to use 13 bytes for the encoded character. This breaks
the paradigm that the number of leading bits gives how many total bytes there
are in the character. */
/* A continuation byte in a UTF-8 encoded sequence contributes this number of
* low-order bits to the specification of the code point. In the bit
* maps above, you see that the first 2 bits are a constant '10', leaving 6 of
* real information. (If you're really curious, the only two numbers that work
* out for this on an 8-bit byte are 5 and 6. Since the first two bits are
* already taken, a maximum of 6 are available for anything else. If 6 is
* used, there are 64 possible continuations 80-BF. With 5, there are 32,
* A0-BF. And with 4 there would be 0 continuations possible; an
* impossibility. So 5 is the minimum. UTF-EBCDIC I8 (Intermediate 8) is just
* setting this to 5. We could have a UTF-8 encoding that is based on ASCII,
* but uses just 5 bits of payload per continuation byte. The reason someone
* might want to do this is to extend the set of characters that occupy a
* single byte when encoded in this hypothetical UTF-8 to additionally include
* the C1 controls.) */
# define UTF_CONTINUATION_BYTE_INFO_BITS 6
/* ^? is defined to be DEL on ASCII systems. See the definition of toCTRL()
* for more */
# define QUESTION_MARK_CTRL DEL_NATIVE
#endif /* EBCDIC vs ASCII */
/* It turns out that in a number of cases, that handling ASCII vs EBCDIC is a
* matter of being off-by-one. So this is a convenience macro, used to avoid
* some #ifdefs. */
#define ONE_IF_EBCDIC_ZERO_IF_NOT \
(UTF_CONTINUATION_BYTE_INFO_BITS == UTF_EBCDIC_CONTINUATION_BYTE_INFO_BITS)
/* Since the significant bits in a continuation byte are stored in the
* least-significant positions, we often find ourselves shifting by that
* amount. This is a clearer name in such situations */
#define UTF_ACCUMULATION_SHIFT UTF_CONTINUATION_BYTE_INFO_BITS
/* 2**info_bits - 1. This masks out all but the bits that carry real
* information in a continuation byte. This turns out to be 0x3F in UTF-8,
* 0x1F in UTF-EBCDIC. */
#define UTF_CONTINUATION_MASK \
((U8) nBIT_MASK(UTF_CONTINUATION_BYTE_INFO_BITS))
/* For use in UTF8_IS_CONTINUATION(). This turns out to be 0xC0 in UTF-8,
* E0 in UTF-EBCDIC */
#define UTF_IS_CONTINUATION_MASK \
((U8) ((0xFF << UTF_ACCUMULATION_SHIFT) & 0xFF))
/* This defines the bits that mark a byte in a multi-byte UTF-8 encoded
* character as being a continuation byte. A MASK clears the bits you don't
* want, using a binary '&'; and a MARK sets the ones you do want, using a
* binary '|'. As stated earlier, the fundamental difference between UTF-8 and
* UTF-EBCDIC is that the former has the upper 2 bits of a continuation byte be
* '10', and the latter has the upper 3 bits be '101', leaving 6 and 5 bits
* respectively in which to store information. This is equivalent to "All bits
* are 1 except those that store information (which vary) plus the bit that is
* required to be 0". This yields 1000 0000 (0x80) for ASCII, and 1010 0000
* (0xA0) for UTF-EBCDIC. */
#define UTF_CONTINUATION_MARK (~(0x40 | UTF_CONTINUATION_MASK) & 0xff)
/* These values are clearer in some contexts; still apply to UTF, not UTF-8 */
#define UTF_MIN_CONTINUATION_BYTE UTF_CONTINUATION_MARK
#define MIN_OFFUNI_VARIANT_CP UTF_MIN_CONTINUATION_BYTE
/* This is the name to use if you are working in UTF-8, as opposed to plain
* 'UTF', which is I8 on EBCDIC platforms */
#define UTF8_MIN_CONTINUATION_BYTE I8_TO_NATIVE_UTF8(UTF_MIN_CONTINUATION_BYTE)
/* Is the byte 'c' part of a multi-byte UTF8-8 encoded sequence, and not the
* first byte thereof? */
#define UTF8_IS_CONTINUATION(c) (__ASSERT_(FITS_IN_8_BITS(c)) \
(((NATIVE_UTF8_TO_I8(c) & UTF_IS_CONTINUATION_MASK) \
== UTF_CONTINUATION_MARK)))
/* Is the representation of the Unicode code point 'cp' the same regardless of
* being encoded in UTF-8 or not? This is a fundamental property of
* UTF-8,EBCDIC */
#define OFFUNI_IS_INVARIANT(c) (((WIDEST_UTYPE)(c)) < MIN_OFFUNI_VARIANT_CP)
/*
=for apidoc Am|bool|UVCHR_IS_INVARIANT|UV cp
Evaluates to 1 if the representation of code point C<cp> is the same whether or
not it is encoded in UTF-8; otherwise evaluates to 0. UTF-8 invariant
characters can be copied as-is when converting to/from UTF-8, saving time.
C<cp> is Unicode if above 255; otherwise is platform-native.
=cut
*/
#define UVCHR_IS_INVARIANT(cp) (OFFUNI_IS_INVARIANT(NATIVE_TO_UNI(cp)))
/* This defines the 1-bits that are to be in the first byte of a multi-byte
* UTF-8 encoded character that mark it as a start byte and give the number of
* bytes that comprise the character. 'len' is that number.
*
* To illustrate: len = 2 => ((U8) ~ 0b0011_1111) or 1100_0000
* 7 => ((U8) ~ 0b0000_0001) or 1111_1110
* > 7 => 0xFF
*
* This is not to be used on a single-byte character. As in many places in
* perl, U8 must be 8 bits
*/
#define UTF_START_MARK(len) ((U8) ~(0xFF >> (len)))
/* Masks out the initial one bits in a start byte, leaving the following 0 bit
* and the real data bits. 'len' is the number of bytes in the multi-byte
* sequence that comprises the character.
*
* To illustrate: len = 2 => 0b0011_1111 works on start byte 110xxxxx
* 6 => 0b0000_0011 works on start byte 1111110x
* >= 7 => There are no data bits in the start byte
* Note that on ASCII platforms, this can be passed a len=1 byte; and all the
* real data bits will be returned:
len = 1 => 0b0111_1111
* This isn't true on EBCDIC platforms, where some len=1 bytes are of the form
* 0b101x_xxxx, so this can't be used there on single-byte characters. */
#define UTF_START_MASK(len) (0xFF >> (len))
/*
=for apidoc AmnU|STRLEN|UTF8_MAXBYTES
The maximum width of a single UTF-8 encoded character, in bytes.
NOTE: Strictly speaking Perl's UTF-8 should not be called UTF-8 since UTF-8
is an encoding of Unicode, and Unicode's upper limit, 0x10FFFF, can be
expressed with 4 bytes. However, Perl thinks of UTF-8 as a way to encode
non-negative integers in a binary format, even those above Unicode.
=cut
The start byte 0xFE, never used in any ASCII platform UTF-8 specification, has
an obvious meaning, namely it has its upper 7 bits set, so it should start a
sequence of 7 bytes. And in fact, this is exactly what standard UTF-EBCDIC
does.
The start byte FF, on the other hand could have several different plausible
meanings:
1) The meaning in standard UTF-EBCDIC, namely as an FE start byte, with the
bottom bit that should be a fixed '0' to form FE, instead acting as an
info bit, 0 or 1.
2) That the sequence should have exactly 8 bytes.
3) That the next byte is to be treated as a sort of extended start byte,
which in combination with this one gives the total length of the sequence.
There are published UTF-8 extensions that do this, some string together
multiple initial FF start bytes to achieve arbitrary precision.
4) That the sequence has exactly n bytes, where n is what the implementation
chooses.
Perl has chosen 4).
The goal is to be able to represent 64-bit values in UTF-8 or UTF-EBCDIC. That
rules out items 1) and 2). Item 3) has the deal-breaking disadvantage of
requiring one to read more than one byte to determine the total length of the
sequence. So in Perl, a start byte of FF indicates a UTF-8 string consisting
of the start byte, plus enough continuation bytes to encode a 64 bit value.
This turns out to be 13 total bytes in UTF-8 and 14 in UTF-EBCDIC. This is
because we get zero info bits from the start byte, plus
12 * 6 bits of info per continuation byte (could encode 72-bit numbers) on
UTF-8 (khw knows not why 11, which would encode 66 bits wasn't
chosen instead); and
13 * 5 bits of info per byte (could encode 65-bit numbers) on UTF-EBCDIC
The disadvantages of this method are:
1) There's potentially a lot of wasted bytes for all but the largest values.
For example, something that could be represented by 7 continuation bytes,
instead requires the full 12 or 13.
2) There would be problems should larger values, 128-bit say, ever need to be
represented.
WARNING: This number must be in sync with the value in
regen/charset_translations.pl. */
#define UTF8_MAXBYTES \
(ASCII_PLATFORM_UTF8_MAXBYTES + ONE_IF_EBCDIC_ZERO_IF_NOT)
/* Calculate how many bytes are necessary to represent a value whose most
* significant 1 bit is in bit position 'pos' of the word. For 0x1, 'pos would
* be 0; and for 0x400, 'pos' would be 10, and the result would be:
* EBCDIC floor((-1 + (10 + 5 - 1 - 1)) / (5 - 1))
* = floor((-1 + (13)) / 4)
* = floor(12 / 4)
* = 3
* ASCII floor(( 0 + (10 + 6 - 1 - 1)) / (6 - 1))
* = floor(14 / 5)
* = 2
* The reason this works is because the number of bits needed to represent a
* value is proportional to (UTF_CONTINUATION_BYTE_INFO_BITS - 1). The -1 is
* because each new continuation byte removes one bit of information from the
* start byte.
*
* This is a step function (we need to allocate a full extra byte if we
* overflow by just a single bit)
*
* The caller is responsible for making sure 'pos' is at least 8 (occupies 9
* bits), as it breaks down at the lower edge. At the high end, if it returns
* 8 or more, Perl instead anomalously uses MAX_BYTES, so this would be wrong.
* */
#define UNISKIP_BY_MSB_(pos) \
( ( -ONE_IF_EBCDIC_ZERO_IF_NOT /* platform break pos's are off-by-one */ \
+ (pos) + ((UTF_CONTINUATION_BYTE_INFO_BITS - 1) - 1)) /* Step fcn */ \
/ (UTF_CONTINUATION_BYTE_INFO_BITS - 1)) /* take floor of */
/* Compute the number of UTF-8 bytes required for representing the input uv,
* which must be a Unicode, not native value.
*
* This uses msbit_pos() which doesn't work on NUL, and UNISKIP_BY_MSB_ breaks
* down for small code points. So first check if the input is invariant to get
* around that, and use a helper for high code points to accommodate the fact
* that above 7 btyes, the value is anomalous. The helper is empty on
* platforms that don't go that high */
#define OFFUNISKIP(uv) \
((OFFUNI_IS_INVARIANT(uv)) \
? 1 \
: (OFFUNISKIP_helper_(uv) UNISKIP_BY_MSB_(msbit_pos(uv))))
/* We need to go to MAX_BYTES when we can't represent 'uv' by the number of
* information bits in 6 continuation bytes (when we get to 6, the start byte
* has no information bits to add to the total). But on 32-bit ASCII
* platforms, that doesn't happen until 6*6 bits, so on those platforms, this
* will always be false */
#if UVSIZE * CHARBITS > (6 * UTF_CONTINUATION_BYTE_INFO_BITS)
# define HAS_EXTRA_LONG_UTF8
# define OFFUNISKIP_helper_(uv) \
UNLIKELY(uv > nBIT_UMAX(6 * UTF_CONTINUATION_BYTE_INFO_BITS)) \
? UTF8_MAXBYTES :
#else
# define OFFUNISKIP_helper_(uv)
#endif
/*
=for apidoc Am|STRLEN|UVCHR_SKIP|UV cp
returns the number of bytes required to represent the code point C<cp> when
encoded as UTF-8. C<cp> is a native (ASCII or EBCDIC) code point if less than
255; a Unicode code point otherwise.
=cut
*/
#define UVCHR_SKIP(uv) OFFUNISKIP(NATIVE_TO_UNI(uv))
#define NATIVE_SKIP(uv) UVCHR_SKIP(uv) /* Old terminology */
/* Most code which says UNISKIP is really thinking in terms of native code
* points (0-255) plus all those beyond. This is an imprecise term, but having
* it means existing code continues to work. For precision, use UVCHR_SKIP,
* NATIVE_SKIP, or OFFUNISKIP */
#define UNISKIP(uv) UVCHR_SKIP(uv)
/* Compute the start byte for a given code point. This requires the log2 of
* the code point, which is hard to compute at compile time, which this macro
* wants to be. (Perhaps deBruijn sequences could be used.) So a parameter
* for the number of bits the value occupies is passed in, which the programmer
* has had to figure out to get compile-time effect. And asserts are used to
* make sure the value is correct.
*
* Since we are interested only in the start byte, we ignore the lower bits
* accounted for by the continuation bytes. Each continuation byte eats up
* UTF_CONTINUATION_BYTE_INFO_BITS bits, so the number of continuation bytes
* needed is floor(bits / UTF_CONTINUATION_BYTE_INFO_BITS). That number is fed
* to UTF_START_MARK() to get the upper part of the start byte. The left over
* bits form the lower part which is OR'd with the mark
*
* Note that on EBCDIC platforms, this is actually the I8 */
#define UTF_START_BYTE(uv, bits) \
(__ASSERT_((uv) >> ((bits) - 1)) /* At least 'bits' */ \
__ASSERT_(((uv) & ~nBIT_MASK(bits)) == 0) /* No extra bits */ \
UTF_START_MARK(UNISKIP_BY_MSB_((bits) - 1)) \
| ((uv) >> (((bits) / UTF_CONTINUATION_BYTE_INFO_BITS) \
* UTF_CONTINUATION_BYTE_INFO_BITS)))
/* Compute the first continuation byte for a given code point. This is mostly
* for compile-time, so how many bits it occupies is also passed in).
*
* We are interested in the first continuation byte, so we ignore the lower
* bits accounted for by the rest of the continuation bytes by right shifting
* out their info bit, and mask out the higher bits that will go into the start
* byte.
*
* Note that on EBCDIC platforms, this is actually the I8 */
#define UTF_FIRST_CONT_BYTE(uv, bits) \
(__ASSERT_((uv) >> ((bits) - 1)) /* At least 'bits' */ \
__ASSERT_(((uv) & ~nBIT_MASK(bits)) == 0) /* No extra bits */ \
UTF_CONTINUATION_MARK \
| ( UTF_CONTINUATION_MASK \
& ((uv) >> ((((bits) / UTF_CONTINUATION_BYTE_INFO_BITS) - 1) \
* UTF_CONTINUATION_BYTE_INFO_BITS))))
#define UTF_MIN_START_BYTE UTF_START_BYTE(UTF_MIN_CONTINUATION_BYTE, 8)
/* Is the byte 'c' the first byte of a multi-byte UTF8-8 encoded sequence?
* This excludes invariants (they are single-byte). It also excludes the
* illegal overlong sequences that begin with C0 and C1 on ASCII platforms, and
* C0-C4 I8 start bytes on EBCDIC ones. On EBCDIC E0 can't start a
* non-overlong sequence, so we define a base macro and for those platforms,
* extend it to also exclude E0 */
#define UTF8_IS_START_base(c) (__ASSERT_(FITS_IN_8_BITS(c)) \
(NATIVE_UTF8_TO_I8(c) >= UTF_MIN_START_BYTE))
#ifdef EBCDIC
# define UTF8_IS_START(c) \
(UTF8_IS_START_base(c) && (c) != I8_TO_NATIVE_UTF8(0xE0))
#else
# define UTF8_IS_START(c) UTF8_IS_START_base(c)
#endif
#define UTF_MIN_ABOVE_LATIN1_BYTE UTF_START_BYTE(0x100, 9)
/* Is the UTF8-encoded byte 'c' the first byte of a sequence of bytes that
* represent a code point > 255? */
#define UTF8_IS_ABOVE_LATIN1(c) (__ASSERT_(FITS_IN_8_BITS(c)) \
(NATIVE_UTF8_TO_I8(c) >= UTF_MIN_ABOVE_LATIN1_BYTE))
/* Is the UTF8-encoded byte 'c' the first byte of a two byte sequence? Use
* UTF8_IS_NEXT_CHAR_DOWNGRADEABLE() instead if the input isn't known to
* be well-formed. */
#define UTF8_IS_DOWNGRADEABLE_START(c) (__ASSERT_(FITS_IN_8_BITS(c)) \
inRANGE_helper_(U8, NATIVE_UTF8_TO_I8(c), \
UTF_MIN_START_BYTE, UTF_MIN_ABOVE_LATIN1_BYTE - 1))
/* The largest code point representable by two UTF-8 bytes on this platform.
* The binary for that code point is:
* 1101_1111 10xx_xxxx in UTF-8, and
* 1101_1111 101y_yyyy in UTF-EBCDIC I8.
* where both x and y are 1, and shown this way to indicate there is one more x
* than there is y. The number of x and y bits are their platform's respective
* UTF_CONTINUATION_BYTE_INFO_BITS. Squeezing out the bits that don't
* contribute to the value, these evaluate to:
* 1_1111 xx_xxxx in UTF-8, and
* 1_1111 y_yyyy in UTF-EBCDIC I8.
* or, the maximum value of an unsigned with (5 + info_bit_count) bits */
#define MAX_UTF8_TWO_BYTE nBIT_UMAX(5 + UTF_CONTINUATION_BYTE_INFO_BITS)
/* The largest code point representable by two UTF-8 bytes on any platform that
* Perl runs on. */
#define MAX_PORTABLE_UTF8_TWO_BYTE \
nBIT_UMAX(5 + MIN( UTF_CONTINUATION_BYTE_INFO_BITS, \
UTF_EBCDIC_CONTINUATION_BYTE_INFO_BITS))
/*
=for apidoc AmnU|STRLEN|UTF8_MAXBYTES_CASE
The maximum number of UTF-8 bytes a single Unicode character can
uppercase/lowercase/titlecase/fold into.
=cut
* Unicode guarantees that the maximum expansion is UTF8_MAX_FOLD_CHAR_EXPAND
* characters, but any above-Unicode code point will fold to itself, so we only
* have to look at the expansion of the maximum Unicode code point. But this
* number may be less than the space occupied by a very large code point under
* Perl's extended UTF-8. We have to make it large enough to fit any single
* character. (It turns out that ASCII and EBCDIC differ in which is larger)
*
=cut
*/
#define UTF8_MAXBYTES_CASE \
MAX(UTF8_MAXBYTES, UTF8_MAX_FOLD_CHAR_EXPAND * UNISKIP_BY_MSB_(20))
/* Rest of these are attributes of Unicode and perl's internals rather than the
* encoding, or happen to be the same in both ASCII and EBCDIC (at least at
* this level; the macros that some of these call may have different
* definitions in the two encodings */
/* In domain restricted to ASCII, these may make more sense to the reader than
* the ones with Latin1 in the name */
#define NATIVE_TO_ASCII(ch) NATIVE_TO_LATIN1(ch)
#define ASCII_TO_NATIVE(ch) LATIN1_TO_NATIVE(ch)
/* More or less misleadingly-named defines, retained for back compat */
#define NATIVE_TO_UTF(ch) NATIVE_UTF8_TO_I8(ch)
#define NATIVE_TO_I8(ch) NATIVE_UTF8_TO_I8(ch)
#define UTF_TO_NATIVE(ch) I8_TO_NATIVE_UTF8(ch)
#define I8_TO_NATIVE(ch) I8_TO_NATIVE_UTF8(ch)
#define NATIVE8_TO_UNI(ch) NATIVE_TO_LATIN1(ch)
/* Adds a UTF8 continuation byte 'new' of information to a running total code
* point 'old' of all the continuation bytes so far. This is designed to be
* used in a loop to convert from UTF-8 to the code point represented. Note
* that this is asymmetric on EBCDIC platforms, in that the 'new' parameter is
* the UTF-EBCDIC byte, whereas the 'old' parameter is a Unicode (not EBCDIC)
* code point in process of being generated */
#define UTF8_ACCUMULATE(old, new) (__ASSERT_(FITS_IN_8_BITS(new)) \
((old) << UTF_ACCUMULATION_SHIFT) \
| ((NATIVE_UTF8_TO_I8(new)) \
& UTF_CONTINUATION_MASK))
/* This works in the face of malformed UTF-8. */
#define UTF8_IS_NEXT_CHAR_DOWNGRADEABLE(s, e) \
( ( (e) - (s) > 1) \
&& UTF8_IS_DOWNGRADEABLE_START(*(s)) \
&& UTF8_IS_CONTINUATION(*((s)+1)))
/* Longer, but more accurate name */
#define UTF8_IS_ABOVE_LATIN1_START(c) UTF8_IS_ABOVE_LATIN1(c)
/* Convert a UTF-8 variant Latin1 character to a native code point value.
* Needs just one iteration of accumulate. Should be used only if it is known
* that the code point is < 256, and is not UTF-8 invariant. Use the slower
* but more general TWO_BYTE_UTF8_TO_NATIVE() which handles any code point
* representable by two bytes (which turns out to be up through
* MAX_PORTABLE_UTF8_TWO_BYTE). The two parameters are:
* HI: a downgradable start byte;
* LO: continuation.
* */
#define EIGHT_BIT_UTF8_TO_NATIVE(HI, LO) \
( __ASSERT_(UTF8_IS_DOWNGRADEABLE_START(HI)) \
__ASSERT_(UTF8_IS_CONTINUATION(LO)) \
LATIN1_TO_NATIVE(UTF8_ACCUMULATE(( \
NATIVE_UTF8_TO_I8(HI) & UTF_START_MASK(2)), (LO))))
/* Convert a two (not one) byte utf8 character to a native code point value.
* Needs just one iteration of accumulate. Should not be used unless it is
* known that the two bytes are legal: 1) two-byte start, and 2) continuation.
* Note that the result can be larger than 255 if the input character is not
* downgradable */
#define TWO_BYTE_UTF8_TO_NATIVE(HI, LO) \
(__ASSERT_(FITS_IN_8_BITS(HI)) \
__ASSERT_(FITS_IN_8_BITS(LO)) \
__ASSERT_(PL_utf8skip[(U8) HI] == 2) \
__ASSERT_(UTF8_IS_CONTINUATION(LO)) \
UNI_TO_NATIVE(UTF8_ACCUMULATE((NATIVE_UTF8_TO_I8(HI) & UTF_START_MASK(2)), \
(LO))))
/* Should never be used, and be deprecated */
#define TWO_BYTE_UTF8_TO_UNI(HI, LO) NATIVE_TO_UNI(TWO_BYTE_UTF8_TO_NATIVE(HI, LO))
/*
=for apidoc Am|STRLEN|UTF8SKIP|const char* s
=for apidoc_item |STRLEN|UTF8_SKIP|const char* s
=for apidoc_item |STRLEN|UTF8_CHK_SKIP|const char* s
=for apidoc_item |STRLEN|UTF8_SAFE_SKIP|const char* s|const char* e
Each of these returns the number of bytes in the UTF-8 encoded character
whose first (perhaps only) byte is pointed to by C<s>.
C<UTF8SKIP> and C<UTF8_SKIP> are synonyms. Use them when there is no
possibility that the character pointed to by C<s> is malformed.
If there is a possibility of malformed input, use instead:
=over
=item C<L</UTF8_SAFE_SKIP>> if you know the maximum ending pointer in the buffer pointed to by C<s>
If the buffer has enough bytes to hold the character, it returns the
same value as C<UTF8SKIP> and C<UTF8_SKIP> would. If the buffer has
fewer bytes than can fit, it returns the number of bytes available in
the buffer, which could be 0 if S<C<s E<gt>= e>>. On DEBUGGING builds,
it asserts that S<C<s E<lt>= e>>.
=item C<L</UTF8_CHK_SKIP>> if you don't know the maximum ending pointer
This version doesn't blindly assume that the input string pointed to by
C<s> is well-formed, but verifies that there isn't a NUL terminating
character before the expected end of the next character in C<s>. The
length C<UTF8_CHK_SKIP> returns stops just before any such NUL.
Perl tends to add NULs, as an insurance policy, after the end of strings in
SV's, so it is likely that using this macro on an SV string will prevent
inadvertent reading beyond the end of the input buffer, even if it is
malformed UTF-8.
This macro is intended to be used by XS modules where the inputs could be
malformed, and it isn't feasible to restructure to use the safer
C<L</UTF8_SAFE_SKIP>>, for example when interfacing with a C library.
=back
=cut
*/
#define UTF8SKIP(s) PL_utf8skip[*(const U8*)(ASSERT_IS_PTR(s))]
#define UTF8_SKIP(s) UTF8SKIP(s)
#define UTF8_CHK_SKIP(s) \
(UNLIKELY(s[0] == '\0') ? 1 : my_strnlen((const char *) (s), UTF8SKIP(s)))
#define UTF8_SAFE_SKIP(s, e) (__ASSERT_((e) >= (s)) \
UNLIKELY(((e) - (s)) <= 0) \
? 0 \
: MIN(((e) - (s)), UTF8_SKIP(s)))
/* Most code that says 'UNI_' really means the native value for code points up
* through 255 */
#define UNI_IS_INVARIANT(cp) UVCHR_IS_INVARIANT(cp)
/*
=for apidoc Am|bool|UTF8_IS_INVARIANT|char c
Evaluates to 1 if the byte C<c> represents the same character when encoded in
UTF-8 as when not; otherwise evaluates to 0. UTF-8 invariant characters can be
copied as-is when converting to/from UTF-8, saving time.
In spite of the name, this macro gives the correct result if the input string
from which C<c> comes is not encoded in UTF-8.
See C<L</UVCHR_IS_INVARIANT>> for checking if a UV is invariant.
=cut
The reason it works on both UTF-8 encoded strings and non-UTF-8 encoded, is
that it returns TRUE in each for the exact same set of bit patterns. It is
valid on a subset of what UVCHR_IS_INVARIANT is valid on, so can just use that;
and the compiler should optimize out anything extraneous given the
implementation of the latter. */
#define UTF8_IS_INVARIANT(c) UVCHR_IS_INVARIANT(ASSERT_NOT_PTR(c))
/* Like the above, but its name implies a non-UTF8 input, which as the comments
* above show, doesn't matter as to its implementation */
#define NATIVE_BYTE_IS_INVARIANT(c) UVCHR_IS_INVARIANT(c)
/* Misleadingly named: is the UTF8-encoded byte 'c' part of a variant sequence
* in UTF-8? This is the inverse of UTF8_IS_INVARIANT. */
#define UTF8_IS_CONTINUED(c) (__ASSERT_(FITS_IN_8_BITS(c)) \
(! UTF8_IS_INVARIANT(c)))
/* The macros in the next 4 sets are used to generate the two utf8 or utfebcdic
* bytes from an ordinal that is known to fit into exactly two (not one) bytes;
* it must be less than 0x3FF to work across both encodings. */
/* These two are helper macros for the other three sets, and should not be used
* directly anywhere else. 'translate_function' is either NATIVE_TO_LATIN1
* (which works for code points up through 0xFF) or NATIVE_TO_UNI which works
* for any code point */
#define __BASE_TWO_BYTE_HI(c, translate_function) \
(__ASSERT_(! UVCHR_IS_INVARIANT(c)) \
I8_TO_NATIVE_UTF8((translate_function(c) >> UTF_ACCUMULATION_SHIFT) \
| UTF_START_MARK(2)))
#define __BASE_TWO_BYTE_LO(c, translate_function) \
(__ASSERT_(! UVCHR_IS_INVARIANT(c)) \
I8_TO_NATIVE_UTF8((translate_function(c) & UTF_CONTINUATION_MASK) \
| UTF_CONTINUATION_MARK))
/* The next two macros should not be used. They were designed to be usable as
* the case label of a switch statement, but this doesn't work for EBCDIC. Use
* regen/unicode_constants.pl instead */
#define UTF8_TWO_BYTE_HI_nocast(c) __BASE_TWO_BYTE_HI(c, NATIVE_TO_UNI)
#define UTF8_TWO_BYTE_LO_nocast(c) __BASE_TWO_BYTE_LO(c, NATIVE_TO_UNI)
/* The next two macros are used when the source should be a single byte
* character; checked for under DEBUGGING */
#define UTF8_EIGHT_BIT_HI(c) (__ASSERT_(FITS_IN_8_BITS(c)) \
( __BASE_TWO_BYTE_HI(c, NATIVE_TO_LATIN1)))
#define UTF8_EIGHT_BIT_LO(c) (__ASSERT_(FITS_IN_8_BITS(c)) \
(__BASE_TWO_BYTE_LO(c, NATIVE_TO_LATIN1)))
/* These final two macros in the series are used when the source can be any
* code point whose UTF-8 is known to occupy 2 bytes; they are less efficient
* than the EIGHT_BIT versions on EBCDIC platforms. We use the logical '~'
* operator instead of "<=" to avoid getting compiler warnings.
* MAX_UTF8_TWO_BYTE should be exactly all one bits in the lower few
* places, so the ~ works */
#define UTF8_TWO_BYTE_HI(c) \
(__ASSERT_((sizeof(c) == 1) \
|| !(((WIDEST_UTYPE)(c)) & ~MAX_UTF8_TWO_BYTE)) \
(__BASE_TWO_BYTE_HI(c, NATIVE_TO_UNI)))
#define UTF8_TWO_BYTE_LO(c) \
(__ASSERT_((sizeof(c) == 1) \
|| !(((WIDEST_UTYPE)(c)) & ~MAX_UTF8_TWO_BYTE)) \
(__BASE_TWO_BYTE_LO(c, NATIVE_TO_UNI)))
/* This is illegal in any well-formed UTF-8 in both EBCDIC and ASCII
* as it is only in overlongs. */
#define ILLEGAL_UTF8_BYTE I8_TO_NATIVE_UTF8(0xC1)
/*
* 'UTF' is whether or not p is encoded in UTF8. The names 'foo_lazy_if' stem
* from an earlier version of these macros in which they didn't call the
* foo_utf8() macros (i.e. were 'lazy') unless they decided that *p is the
* beginning of a utf8 character. Now that foo_utf8() determines that itself,
* no need to do it again here
*/
#define isIDFIRST_lazy_if_safe(p, e, UTF) \
((IN_BYTES || !UTF) \
? isIDFIRST(*(p)) \
: isIDFIRST_utf8_safe(p, e))
#define isWORDCHAR_lazy_if_safe(p, e, UTF) \
((IN_BYTES || !UTF) \
? isWORDCHAR(*(p)) \
: isWORDCHAR_utf8_safe((U8 *) p, (U8 *) e))
#define isALNUM_lazy_if_safe(p, e, UTF) isWORDCHAR_lazy_if_safe(p, e, UTF)
#define UTF8_MAXLEN UTF8_MAXBYTES
/* A Unicode character can fold to up to 3 characters */
#define UTF8_MAX_FOLD_CHAR_EXPAND 3
#define IN_BYTES UNLIKELY(CopHINTS_get(PL_curcop) & HINT_BYTES)
/*
=for apidoc Am|bool|DO_UTF8|SV* sv
Returns a bool giving whether or not the PV in C<sv> is to be treated as being
encoded in UTF-8.
You should use this I<after> a call to C<SvPV()> or one of its variants, in
case any call to string overloading updates the internal UTF-8 encoding flag.
=cut
*/
#define DO_UTF8(sv) (SvUTF8(sv) && !IN_BYTES)
/* Should all strings be treated as Unicode, and not just UTF-8 encoded ones?
* Is so within 'feature unicode_strings' or 'locale :not_characters', and not
* within 'use bytes'. UTF-8 locales are not tested for here, because it gets
* complicated by the probability of having categories in different locales. */
#define IN_UNI_8_BIT \
(( ( (CopHINTS_get(PL_curcop) & HINT_UNI_8_BIT)) \
|| ( CopHINTS_get(PL_curcop) & HINT_LOCALE \
/* -1 below is for :not_characters */ \
&& is_in_locale_category_(FALSE, -1))) \
&& (! IN_BYTES))
#define UNICODE_SURROGATE_FIRST 0xD800
#define UNICODE_SURROGATE_LAST 0xDFFF
/*
=for apidoc Am|bool|UNICODE_IS_SURROGATE|const UV uv
Returns a boolean as to whether or not C<uv> is one of the Unicode surrogate
code points
=for apidoc Am|bool|UTF8_IS_SURROGATE|const U8 *s|const U8 *e
Evaluates to non-zero if the first few bytes of the string starting at C<s> and
looking no further than S<C<e - 1>> are well-formed UTF-8 that represents one
of the Unicode surrogate code points; otherwise it evaluates to 0. If
non-zero, the value gives how many bytes starting at C<s> comprise the code
point's representation.
=cut
*/
#define UNICODE_IS_SURROGATE(uv) UNLIKELY(inRANGE(uv, UNICODE_SURROGATE_FIRST, \
UNICODE_SURROGATE_LAST))
#define UTF8_IS_SURROGATE(s, e) is_SURROGATE_utf8_safe(s, e)
/*
=for apidoc AmnU|UV|UNICODE_REPLACEMENT
Evaluates to 0xFFFD, the code point of the Unicode REPLACEMENT CHARACTER
=for apidoc Am|bool|UNICODE_IS_REPLACEMENT|const UV uv
Returns a boolean as to whether or not C<uv> is the Unicode REPLACEMENT
CHARACTER