forked from svaarala/duktape
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathduk_bi_string.c
More file actions
1615 lines (1353 loc) · 45 KB
/
duk_bi_string.c
File metadata and controls
1615 lines (1353 loc) · 45 KB
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
/*
* String built-ins
*
* Most String built-ins must only accept strings (or String objects).
* Symbols, represented internally as strings, must be generally rejected.
* The duk_push_this_coercible_to_string() helper does this automatically.
*/
/* XXX: There are several limitations in the current implementation for
* strings with >= 0x80000000UL characters. In some cases one would need
* to be able to represent the range [-0xffffffff,0xffffffff] and so on.
* Generally character and byte length are assumed to fit into signed 32
* bits (< 0x80000000UL). Places with issues are not marked explicitly
* below in all cases, look for signed type usage (duk_int_t etc) for
* offsets/lengths.
*/
#include "duk_internal.h"
#if defined(DUK_USE_STRING_BUILTIN)
/*
* Helpers
*/
DUK_LOCAL duk_hstring *duk__str_tostring_notregexp(duk_hthread *thr, duk_idx_t idx) {
duk_hstring *h;
if (duk_get_class_number(thr, idx) == DUK_HOBJECT_CLASS_REGEXP) {
DUK_ERROR_TYPE_INVALID_ARGS(thr);
DUK_WO_NORETURN(return NULL;);
}
h = duk_to_hstring(thr, idx);
DUK_ASSERT(h != NULL);
return h;
}
DUK_LOCAL duk_int_t
duk__str_search_shared(duk_hthread *thr, duk_hstring *h_this, duk_hstring *h_search, duk_int_t start_cpos, duk_bool_t backwards) {
duk_int_t cpos;
duk_int_t bpos;
const duk_uint8_t *p_start, *p_end, *p;
const duk_uint8_t *q_start;
duk_int_t q_blen;
duk_uint8_t firstbyte;
duk_uint8_t t;
cpos = start_cpos;
/* Empty searchstring always matches; cpos must be clamped here.
* (If q_blen were < 0 due to clamped coercion, it would also be
* caught here.)
*/
q_start = DUK_HSTRING_GET_DATA(h_search);
q_blen = (duk_int_t) DUK_HSTRING_GET_BYTELEN(h_search);
if (q_blen <= 0) {
return cpos;
}
DUK_ASSERT(q_blen > 0);
bpos = (duk_int_t) duk_heap_strcache_offset_char2byte(thr, h_this, (duk_uint32_t) cpos);
p_start = DUK_HSTRING_GET_DATA(h_this);
p_end = p_start + DUK_HSTRING_GET_BYTELEN(h_this);
p = p_start + bpos;
/* This loop is optimized for size. For speed, there should be
* two separate loops, and we should ensure that memcmp() can be
* used without an extra "will searchstring fit" check. Doing
* the preconditioning for 'p' and 'p_end' is easy but cpos
* must be updated if 'p' is wound back (backward scanning).
*/
firstbyte = q_start[0]; /* leading byte of match string */
while (p <= p_end && p >= p_start) {
t = *p;
/* For ECMAScript strings, this check can only match for
* initial UTF-8 bytes (not continuation bytes). For other
* strings all bets are off.
*/
if ((t == firstbyte) && ((duk_size_t) (p_end - p) >= (duk_size_t) q_blen)) {
DUK_ASSERT(q_blen > 0);
if (duk_memcmp((const void *) p, (const void *) q_start, (size_t) q_blen) == 0) {
return cpos;
}
}
/* track cpos while scanning */
if (backwards) {
/* when going backwards, we decrement cpos 'early';
* 'p' may point to a continuation byte of the char
* at offset 'cpos', but that's OK because we'll
* backtrack all the way to the initial byte.
*/
if ((t & 0xc0) != 0x80) {
cpos--;
}
p--;
} else {
if ((t & 0xc0) != 0x80) {
cpos++;
}
p++;
}
}
/* Not found. Empty string case is handled specially above. */
return -1;
}
/*
* Constructor
*/
DUK_INTERNAL duk_ret_t duk_bi_string_constructor(duk_hthread *thr) {
duk_hstring *h;
duk_uint_t flags;
/* String constructor needs to distinguish between an argument not given at all
* vs. given as 'undefined'. We're a vararg function to handle this properly.
*/
/* XXX: copy current activation flags to thr, including current magic,
* is_constructor_call etc. This takes a few bytes in duk_hthread but
* makes call sites smaller (there are >30 is_constructor_call and get
* current magic call sites.
*/
if (duk_get_top(thr) == 0) {
duk_push_hstring_empty(thr);
} else {
h = duk_to_hstring_acceptsymbol(thr, 0);
if (DUK_UNLIKELY(DUK_HSTRING_HAS_SYMBOL(h) && !duk_is_constructor_call(thr))) {
duk_push_symbol_descriptive_string(thr, h);
duk_replace(thr, 0);
}
}
duk_to_string(thr, 0); /* catches symbol argument for constructor call */
DUK_ASSERT(duk_is_string(thr, 0));
duk_set_top(thr, 1); /* Top may be 1 or larger. */
if (duk_is_constructor_call(thr)) {
/* String object internal value is immutable */
flags = DUK_HOBJECT_FLAG_EXTENSIBLE | DUK_HOBJECT_FLAG_FASTREFS | DUK_HOBJECT_FLAG_EXOTIC_STRINGOBJ |
DUK_HOBJECT_CLASS_AS_FLAGS(DUK_HOBJECT_CLASS_STRING);
duk_push_object_helper(thr, flags, DUK_BIDX_STRING_PROTOTYPE);
duk_dup_0(thr);
duk_xdef_prop_stridx_short(thr, -2, DUK_STRIDX_INT_VALUE, DUK_PROPDESC_FLAGS_NONE);
}
/* Note: unbalanced stack on purpose */
return 1;
}
DUK_LOCAL duk_ret_t duk__construct_from_codepoints(duk_hthread *thr, duk_bool_t nonbmp) {
duk_bufwriter_ctx bw_alloc;
duk_bufwriter_ctx *bw;
duk_idx_t i, n;
duk_ucodepoint_t cp;
/* XXX: It would be nice to build the string directly but ToUint16()
* coercion is needed so a generic helper would not be very
* helpful (perhaps coerce the value stack first here and then
* build a string from a duk_tval number sequence in one go?).
*/
n = duk_get_top(thr);
bw = &bw_alloc;
DUK_BW_INIT_PUSHBUF(thr, bw, (duk_size_t) n); /* initial estimate for ASCII only codepoints */
for (i = 0; i < n; i++) {
/* XXX: could improve bufwriter handling to write multiple codepoints
* with one ensure call but the relative benefit would be quite small.
*/
if (nonbmp) {
/* ES2015 requires that (1) SameValue(cp, ToInteger(cp)) and
* (2) cp >= 0 and cp <= 0x10ffff. This check does not
* implement the steps exactly but the outcome should be
* the same.
*/
duk_int32_t i32 = 0;
if (!duk_is_whole_get_int32(duk_to_number(thr, i), &i32) || i32 < 0 || i32 > 0x10ffffL) {
DUK_DCERROR_RANGE_INVALID_ARGS(thr);
}
DUK_ASSERT(i32 >= 0 && i32 <= 0x10ffffL);
cp = (duk_ucodepoint_t) i32;
DUK_BW_WRITE_ENSURE_CESU8(thr, bw, cp);
} else {
#if defined(DUK_USE_NONSTD_STRING_FROMCHARCODE_32BIT)
/* ToUint16() coercion is mandatory in the E5.1 specification, but
* this non-compliant behavior makes more sense because we support
* non-BMP codepoints. Don't use CESU-8 because that'd create
* surrogate pairs.
*/
cp = (duk_ucodepoint_t) duk_to_uint32(thr, i);
DUK_BW_WRITE_ENSURE_XUTF8(thr, bw, cp);
#else
cp = (duk_ucodepoint_t) duk_to_uint16(thr, i);
DUK_ASSERT(cp >= 0 && cp <= 0x10ffffL);
DUK_BW_WRITE_ENSURE_CESU8(thr, bw, cp);
#endif
}
}
DUK_BW_COMPACT(thr, bw);
(void) duk_buffer_to_string(thr, -1); /* Safe, extended UTF-8 or CESU-8 encoded. */
return 1;
}
DUK_INTERNAL duk_ret_t duk_bi_string_constructor_from_char_code(duk_hthread *thr) {
return duk__construct_from_codepoints(thr, 0 /*nonbmp*/);
}
#if defined(DUK_USE_ES6)
DUK_INTERNAL duk_ret_t duk_bi_string_constructor_from_code_point(duk_hthread *thr) {
return duk__construct_from_codepoints(thr, 1 /*nonbmp*/);
}
#endif
/*
* toString(), valueOf()
*/
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_to_string(duk_hthread *thr) {
duk_tval *tv;
duk_push_this(thr);
tv = duk_require_tval(thr, -1);
DUK_ASSERT(tv != NULL);
if (DUK_TVAL_IS_STRING(tv)) {
/* return as is */
} else if (DUK_TVAL_IS_OBJECT(tv)) {
duk_hobject *h = DUK_TVAL_GET_OBJECT(tv);
DUK_ASSERT(h != NULL);
/* Must be a "string object", i.e. class "String" */
if (DUK_HOBJECT_GET_CLASS_NUMBER(h) != DUK_HOBJECT_CLASS_STRING) {
goto type_error;
}
duk_xget_owndataprop_stridx_short(thr, -1, DUK_STRIDX_INT_VALUE);
DUK_ASSERT(duk_is_string(thr, -1));
} else {
goto type_error;
}
(void) duk_require_hstring_notsymbol(thr, -1); /* Reject symbols (and wrapped symbols). */
return 1;
type_error:
DUK_DCERROR_TYPE_INVALID_ARGS(thr);
}
/*
* Character and charcode access
*/
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_char_at(duk_hthread *thr) {
duk_hstring *h;
duk_int_t pos;
/* XXX: faster implementation */
h = duk_push_this_coercible_to_string(thr);
DUK_ASSERT(h != NULL);
pos = duk_to_int(thr, 0);
if (sizeof(duk_size_t) >= sizeof(duk_uint_t)) {
/* Cast to duk_size_t works in this case:
* - If pos < 0, (duk_size_t) pos will always be
* >= max_charlen, and result will be the empty string
* (see duk_substring()).
* - If pos >= 0, pos + 1 cannot wrap.
*/
DUK_ASSERT((duk_size_t) DUK_INT_MIN >= DUK_HSTRING_MAX_BYTELEN);
DUK_ASSERT((duk_size_t) DUK_INT_MAX + 1U > (duk_size_t) DUK_INT_MAX);
duk_substring(thr, -1, (duk_size_t) pos, (duk_size_t) pos + 1U);
} else {
/* If size_t is smaller than int, explicit bounds checks
* are needed because an int may wrap multiple times.
*/
if (DUK_UNLIKELY(pos < 0 || (duk_uint_t) pos >= (duk_uint_t) DUK_HSTRING_GET_CHARLEN(h))) {
duk_push_hstring_empty(thr);
} else {
duk_substring(thr, -1, (duk_size_t) pos, (duk_size_t) pos + 1U);
}
}
return 1;
}
/* Magic: 0=charCodeAt, 1=codePointAt */
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_char_code_at(duk_hthread *thr) {
duk_int_t pos;
duk_hstring *h;
duk_bool_t clamped;
duk_uint32_t cp;
duk_int_t magic;
/* XXX: faster implementation */
DUK_DDD(DUK_DDDPRINT("arg=%!T", (duk_tval *) duk_get_tval(thr, 0)));
h = duk_push_this_coercible_to_string(thr);
DUK_ASSERT(h != NULL);
pos = duk_to_int_clamped_raw(thr,
0 /*index*/,
0 /*min(incl)*/,
(duk_int_t) DUK_HSTRING_GET_CHARLEN(h) - 1 /*max(incl)*/,
&clamped /*out_clamped*/);
#if defined(DUK_USE_ES6)
magic = duk_get_current_magic(thr);
#else
DUK_ASSERT(duk_get_current_magic(thr) == 0);
magic = 0;
#endif
if (clamped) {
/* For out-of-bounds indices .charCodeAt() returns NaN and
* .codePointAt() returns undefined.
*/
if (magic != 0) {
return 0;
}
duk_push_nan(thr);
} else {
DUK_ASSERT(pos >= 0);
cp = (duk_uint32_t) duk_hstring_char_code_at_raw(thr, h, (duk_uint_t) pos, (duk_bool_t) magic /*surrogate_aware*/);
duk_push_u32(thr, cp);
}
return 1;
}
/*
* substring(), substr(), slice()
*/
/* XXX: any chance of merging these three similar but still slightly
* different algorithms so that footprint would be reduced?
*/
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_substring(duk_hthread *thr) {
duk_hstring *h;
duk_int_t start_pos, end_pos;
duk_int_t len;
h = duk_push_this_coercible_to_string(thr);
DUK_ASSERT(h != NULL);
len = (duk_int_t) DUK_HSTRING_GET_CHARLEN(h);
/* [ start end str ] */
start_pos = duk_to_int_clamped(thr, 0, 0, len);
if (duk_is_undefined(thr, 1)) {
end_pos = len;
} else {
end_pos = duk_to_int_clamped(thr, 1, 0, len);
}
DUK_ASSERT(start_pos >= 0 && start_pos <= len);
DUK_ASSERT(end_pos >= 0 && end_pos <= len);
if (start_pos > end_pos) {
duk_int_t tmp = start_pos;
start_pos = end_pos;
end_pos = tmp;
}
DUK_ASSERT(end_pos >= start_pos);
duk_substring(thr, -1, (duk_size_t) start_pos, (duk_size_t) end_pos);
return 1;
}
#if defined(DUK_USE_SECTION_B)
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_substr(duk_hthread *thr) {
duk_hstring *h;
duk_int_t start_pos, end_pos;
duk_int_t len;
/* Unlike non-obsolete String calls, substr() algorithm in E5.1
* specification will happily coerce undefined and null to strings
* ("undefined" and "null").
*/
duk_push_this(thr);
h = duk_to_hstring_m1(thr); /* Reject Symbols. */
DUK_ASSERT(h != NULL);
len = (duk_int_t) DUK_HSTRING_GET_CHARLEN(h);
/* [ start length str ] */
/* The implementation for computing of start_pos and end_pos differs
* from the standard algorithm, but is intended to result in the exactly
* same behavior. This is not always obvious.
*/
/* combines steps 2 and 5; -len ensures max() not needed for step 5 */
start_pos = duk_to_int_clamped(thr, 0, -len, len);
if (start_pos < 0) {
start_pos = len + start_pos;
}
DUK_ASSERT(start_pos >= 0 && start_pos <= len);
/* combines steps 3, 6; step 7 is not needed */
if (duk_is_undefined(thr, 1)) {
end_pos = len;
} else {
DUK_ASSERT(start_pos <= len);
end_pos = start_pos + duk_to_int_clamped(thr, 1, 0, len - start_pos);
}
DUK_ASSERT(start_pos >= 0 && start_pos <= len);
DUK_ASSERT(end_pos >= 0 && end_pos <= len);
DUK_ASSERT(end_pos >= start_pos);
duk_substring(thr, -1, (duk_size_t) start_pos, (duk_size_t) end_pos);
return 1;
}
#endif /* DUK_USE_SECTION_B */
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_slice(duk_hthread *thr) {
duk_hstring *h;
duk_int_t start_pos, end_pos;
duk_int_t len;
h = duk_push_this_coercible_to_string(thr);
DUK_ASSERT(h != NULL);
len = (duk_int_t) DUK_HSTRING_GET_CHARLEN(h);
/* [ start end str ] */
start_pos = duk_to_int_clamped(thr, 0, -len, len);
if (start_pos < 0) {
start_pos = len + start_pos;
}
if (duk_is_undefined(thr, 1)) {
end_pos = len;
} else {
end_pos = duk_to_int_clamped(thr, 1, -len, len);
if (end_pos < 0) {
end_pos = len + end_pos;
}
}
DUK_ASSERT(start_pos >= 0 && start_pos <= len);
DUK_ASSERT(end_pos >= 0 && end_pos <= len);
if (end_pos < start_pos) {
end_pos = start_pos;
}
DUK_ASSERT(end_pos >= start_pos);
duk_substring(thr, -1, (duk_size_t) start_pos, (duk_size_t) end_pos);
return 1;
}
/*
* Case conversion
*/
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_caseconv_shared(duk_hthread *thr) {
duk_small_int_t uppercase = duk_get_current_magic(thr);
(void) duk_push_this_coercible_to_string(thr);
duk_unicode_case_convert_string(thr, (duk_bool_t) uppercase);
return 1;
}
/*
* indexOf() and lastIndexOf()
*/
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_indexof_shared(duk_hthread *thr) {
duk_hstring *h_this;
duk_hstring *h_search;
duk_int_t clen_this;
duk_int_t cpos;
duk_small_uint_t is_lastindexof = (duk_small_uint_t) duk_get_current_magic(thr); /* 0=indexOf, 1=lastIndexOf */
h_this = duk_push_this_coercible_to_string(thr);
DUK_ASSERT(h_this != NULL);
clen_this = (duk_int_t) DUK_HSTRING_GET_CHARLEN(h_this);
h_search = duk_to_hstring(thr, 0);
DUK_ASSERT(h_search != NULL);
duk_to_number(thr, 1);
if (duk_is_nan(thr, 1) && is_lastindexof) {
/* indexOf: NaN should cause pos to be zero.
* lastIndexOf: NaN should cause pos to be +Infinity
* (and later be clamped to len).
*/
cpos = clen_this;
} else {
cpos = duk_to_int_clamped(thr, 1, 0, clen_this);
}
cpos = duk__str_search_shared(thr, h_this, h_search, cpos, is_lastindexof /*backwards*/);
duk_push_int(thr, cpos);
return 1;
}
/*
* replace()
*/
/* XXX: the current implementation works but is quite clunky; it compiles
* to almost 1,4kB of x86 code so it needs to be simplified (better approach,
* shared helpers, etc). Some ideas for refactoring:
*
* - a primitive to convert a string into a regexp matcher (reduces matching
* code at the cost of making matching much slower)
* - use replace() as a basic helper for match() and split(), which are both
* much simpler
* - API call to get_prop and to_boolean
*/
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_replace(duk_hthread *thr) {
duk_hstring *h_input;
duk_hstring *h_match;
duk_hstring *h_search;
duk_hobject *h_re;
duk_bufwriter_ctx bw_alloc;
duk_bufwriter_ctx *bw;
#if defined(DUK_USE_REGEXP_SUPPORT)
duk_bool_t is_regexp;
duk_bool_t is_global;
#endif
duk_bool_t is_repl_func;
duk_uint32_t match_start_coff, match_start_boff;
#if defined(DUK_USE_REGEXP_SUPPORT)
duk_int_t match_caps;
#endif
duk_uint32_t prev_match_end_boff;
const duk_uint8_t *r_start, *r_end, *r; /* repl string scan */
duk_size_t tmp_sz;
DUK_ASSERT_TOP(thr, 2);
h_input = duk_push_this_coercible_to_string(thr);
DUK_ASSERT(h_input != NULL);
bw = &bw_alloc;
DUK_BW_INIT_PUSHBUF(thr, bw, DUK_HSTRING_GET_BYTELEN(h_input)); /* input size is good output starting point */
DUK_ASSERT_TOP(thr, 4);
/* stack[0] = search value
* stack[1] = replace value
* stack[2] = input string
* stack[3] = result buffer
*/
h_re = duk_get_hobject_with_class(thr, 0, DUK_HOBJECT_CLASS_REGEXP);
if (h_re) {
#if defined(DUK_USE_REGEXP_SUPPORT)
is_regexp = 1;
is_global = duk_get_prop_stridx_boolean(thr, 0, DUK_STRIDX_GLOBAL, NULL);
if (is_global) {
/* start match from beginning */
duk_push_int(thr, 0);
duk_put_prop_stridx_short(thr, 0, DUK_STRIDX_LAST_INDEX);
}
#else /* DUK_USE_REGEXP_SUPPORT */
DUK_DCERROR_UNSUPPORTED(thr);
#endif /* DUK_USE_REGEXP_SUPPORT */
} else {
duk_to_string(thr, 0); /* rejects symbols */
#if defined(DUK_USE_REGEXP_SUPPORT)
is_regexp = 0;
is_global = 0;
#endif
}
if (duk_is_function(thr, 1)) {
is_repl_func = 1;
r_start = NULL;
r_end = NULL;
} else {
duk_hstring *h_repl;
is_repl_func = 0;
h_repl = duk_to_hstring(thr, 1); /* reject symbols */
DUK_ASSERT(h_repl != NULL);
r_start = DUK_HSTRING_GET_DATA(h_repl);
r_end = r_start + DUK_HSTRING_GET_BYTELEN(h_repl);
}
prev_match_end_boff = 0;
for (;;) {
/*
* If matching with a regexp:
* - non-global RegExp: lastIndex not touched on a match, zeroed
* on a non-match
* - global RegExp: on match, lastIndex will be updated by regexp
* executor to point to next char after the matching part (so that
* characters in the matching part are not matched again)
*
* If matching with a string:
* - always non-global match, find first occurrence
*
* We need:
* - The character offset of start-of-match for the replacer function
* - The byte offsets for start-of-match and end-of-match to implement
* the replacement values $&, $`, and $', and to copy non-matching
* input string portions (including header and trailer) verbatim.
*
* NOTE: the E5.1 specification is a bit vague how the RegExp should
* behave in the replacement process; e.g. is matching done first for
* all matches (in the global RegExp case) before any replacer calls
* are made? See: test-bi-string-proto-replace.js for discussion.
*/
DUK_ASSERT_TOP(thr, 4);
#if defined(DUK_USE_REGEXP_SUPPORT)
if (is_regexp) {
duk_dup_0(thr);
duk_dup_2(thr);
duk_regexp_match(thr); /* [ ... regexp input ] -> [ res_obj ] */
if (!duk_is_object(thr, -1)) {
duk_pop(thr);
break;
}
duk_get_prop_stridx_short(thr, -1, DUK_STRIDX_INDEX);
DUK_ASSERT(duk_is_number(thr, -1));
match_start_coff = duk_get_uint(thr, -1);
duk_pop(thr);
duk_get_prop_index(thr, -1, 0);
DUK_ASSERT(duk_is_string(thr, -1));
h_match = duk_known_hstring(thr, -1);
duk_pop(thr); /* h_match is borrowed, remains reachable through match_obj */
if (DUK_HSTRING_GET_BYTELEN(h_match) == 0) {
/* This should be equivalent to match() algorithm step 8.f.iii.2:
* detect an empty match and allow it, but don't allow it twice.
*/
duk_uint32_t last_index;
duk_get_prop_stridx_short(thr, 0, DUK_STRIDX_LAST_INDEX);
last_index = (duk_uint32_t) duk_get_uint(thr, -1);
DUK_DDD(DUK_DDDPRINT("empty match, bump lastIndex: %ld -> %ld",
(long) last_index,
(long) (last_index + 1)));
duk_pop(thr);
duk_push_uint(thr, (duk_uint_t) (last_index + 1));
duk_put_prop_stridx_short(thr, 0, DUK_STRIDX_LAST_INDEX);
}
DUK_ASSERT(duk_get_length(thr, -1) <= DUK_INT_MAX); /* string limits */
match_caps = (duk_int_t) duk_get_length(thr, -1);
} else {
#else /* DUK_USE_REGEXP_SUPPORT */
{ /* unconditionally */
#endif /* DUK_USE_REGEXP_SUPPORT */
const duk_uint8_t *p_start, *p_end, *p; /* input string scan */
const duk_uint8_t *q_start; /* match string */
duk_size_t p_blen;
duk_size_t q_blen;
#if defined(DUK_USE_REGEXP_SUPPORT)
DUK_ASSERT(!is_global); /* single match always */
#endif
p_start = DUK_HSTRING_GET_DATA(h_input);
p_end = p_start + DUK_HSTRING_GET_BYTELEN(h_input);
p_blen = (duk_size_t) DUK_HSTRING_GET_BYTELEN(h_input);
p = p_start;
h_search = duk_known_hstring(thr, 0);
q_start = DUK_HSTRING_GET_DATA(h_search);
q_blen = (duk_size_t) DUK_HSTRING_GET_BYTELEN(h_search);
if (q_blen > p_blen) {
break; /* no match */
}
p_end -= q_blen; /* ensure full memcmp() fits in while */
DUK_ASSERT(p_end >= p);
match_start_coff = 0;
while (p <= p_end) {
DUK_ASSERT(p + q_blen <= DUK_HSTRING_GET_DATA(h_input) + DUK_HSTRING_GET_BYTELEN(h_input));
if (duk_memcmp((const void *) p, (const void *) q_start, (size_t) q_blen) == 0) {
duk_dup_0(thr);
h_match = duk_known_hstring(thr, -1);
#if defined(DUK_USE_REGEXP_SUPPORT)
match_caps = 0;
#endif
goto found;
}
/* track utf-8 non-continuation bytes */
if ((p[0] & 0xc0) != 0x80) {
match_start_coff++;
}
p++;
}
/* not found */
break;
}
found:
/* stack[0] = search value
* stack[1] = replace value
* stack[2] = input string
* stack[3] = result buffer
* stack[4] = regexp match OR match string
*/
match_start_boff = (duk_uint32_t) duk_heap_strcache_offset_char2byte(thr, h_input, match_start_coff);
tmp_sz = (duk_size_t) (match_start_boff - prev_match_end_boff);
DUK_BW_WRITE_ENSURE_BYTES(thr, bw, DUK_HSTRING_GET_DATA(h_input) + prev_match_end_boff, tmp_sz);
prev_match_end_boff = match_start_boff + DUK_HSTRING_GET_BYTELEN(h_match);
if (is_repl_func) {
duk_idx_t idx_args;
duk_hstring *h_repl;
/* regexp res_obj is at index 4 */
duk_dup_1(thr);
idx_args = duk_get_top(thr);
#if defined(DUK_USE_REGEXP_SUPPORT)
if (is_regexp) {
duk_int_t idx;
duk_require_stack(thr, match_caps + 2);
for (idx = 0; idx < match_caps; idx++) {
/* match followed by capture(s) */
duk_get_prop_index(thr, 4, (duk_uarridx_t) idx);
}
} else {
#else /* DUK_USE_REGEXP_SUPPORT */
{ /* unconditionally */
#endif /* DUK_USE_REGEXP_SUPPORT */
/* match == search string, by definition */
duk_dup_0(thr);
}
duk_push_uint(thr, (duk_uint_t) match_start_coff);
duk_dup_2(thr);
/* [ ... replacer match [captures] match_char_offset input ] */
duk_call(thr, duk_get_top(thr) - idx_args);
h_repl = duk_to_hstring_m1(thr); /* -> [ ... repl_value ] */
DUK_ASSERT(h_repl != NULL);
DUK_BW_WRITE_ENSURE_HSTRING(thr, bw, h_repl);
duk_pop(thr); /* repl_value */
} else {
r = r_start;
while (r < r_end) {
duk_int_t ch1;
duk_int_t ch2;
#if defined(DUK_USE_REGEXP_SUPPORT)
duk_int_t ch3;
#endif
duk_size_t left;
ch1 = *r++;
if (ch1 != DUK_ASC_DOLLAR) {
goto repl_write;
}
DUK_ASSERT(r <= r_end);
left = (duk_size_t) (r_end - r);
if (left <= 0) {
goto repl_write;
}
ch2 = r[0];
switch (ch2) {
case DUK_ASC_DOLLAR: {
ch1 = (1 << 8) + DUK_ASC_DOLLAR;
goto repl_write;
}
case DUK_ASC_AMP: {
DUK_BW_WRITE_ENSURE_HSTRING(thr, bw, h_match);
r++;
continue;
}
case DUK_ASC_GRAVE: {
tmp_sz = (duk_size_t) match_start_boff;
DUK_BW_WRITE_ENSURE_BYTES(thr, bw, DUK_HSTRING_GET_DATA(h_input), tmp_sz);
r++;
continue;
}
case DUK_ASC_SINGLEQUOTE: {
duk_uint32_t match_end_boff;
/* Use match charlen instead of bytelen, just in case the input and
* match codepoint encodings would have different lengths.
*/
/* XXX: charlen computed here, and also in char2byte helper. */
match_end_boff = (duk_uint32_t) duk_heap_strcache_offset_char2byte(
thr,
h_input,
match_start_coff + (duk_uint_fast32_t) DUK_HSTRING_GET_CHARLEN(h_match));
tmp_sz = (duk_size_t) (DUK_HSTRING_GET_BYTELEN(h_input) - match_end_boff);
DUK_BW_WRITE_ENSURE_BYTES(thr, bw, DUK_HSTRING_GET_DATA(h_input) + match_end_boff, tmp_sz);
r++;
continue;
}
default: {
#if defined(DUK_USE_REGEXP_SUPPORT)
duk_int_t capnum, captmp, capadv;
/* XXX: optional check, match_caps is zero if no regexp,
* so dollar will be interpreted literally anyway.
*/
if (!is_regexp) {
goto repl_write;
}
if (!(ch2 >= DUK_ASC_0 && ch2 <= DUK_ASC_9)) {
goto repl_write;
}
capnum = ch2 - DUK_ASC_0;
capadv = 1;
if (left >= 2) {
ch3 = r[1];
if (ch3 >= DUK_ASC_0 && ch3 <= DUK_ASC_9) {
captmp = capnum * 10 + (ch3 - DUK_ASC_0);
if (captmp < match_caps) {
capnum = captmp;
capadv = 2;
}
}
}
if (capnum > 0 && capnum < match_caps) {
DUK_ASSERT(is_regexp != 0); /* match_caps == 0 without regexps */
/* regexp res_obj is at offset 4 */
duk_get_prop_index(thr, 4, (duk_uarridx_t) capnum);
if (duk_is_string(thr, -1)) {
duk_hstring *h_tmp_str;
h_tmp_str = duk_known_hstring(thr, -1);
DUK_BW_WRITE_ENSURE_HSTRING(thr, bw, h_tmp_str);
} else {
/* undefined -> skip (replaced with empty) */
}
duk_pop(thr);
r += capadv;
continue;
} else {
goto repl_write;
}
#else /* DUK_USE_REGEXP_SUPPORT */
goto repl_write; /* unconditionally */
#endif /* DUK_USE_REGEXP_SUPPORT */
} /* default case */
} /* switch (ch2) */
repl_write:
/* ch1 = (r_increment << 8) + byte */
DUK_BW_WRITE_ENSURE_U8(thr, bw, (duk_uint8_t) (ch1 & 0xff));
r += ch1 >> 8;
} /* while repl */
} /* if (is_repl_func) */
duk_pop(thr); /* pop regexp res_obj or match string */
#if defined(DUK_USE_REGEXP_SUPPORT)
if (!is_global) {
#else
{ /* unconditionally; is_global==0 */
#endif
break;
}
}
/* trailer */
tmp_sz = (duk_size_t) (DUK_HSTRING_GET_BYTELEN(h_input) - prev_match_end_boff);
DUK_BW_WRITE_ENSURE_BYTES(thr, bw, DUK_HSTRING_GET_DATA(h_input) + prev_match_end_boff, tmp_sz);
DUK_ASSERT_TOP(thr, 4);
DUK_BW_COMPACT(thr, bw);
(void) duk_buffer_to_string(thr, -1); /* Safe if inputs are safe. */
return 1;
}
/*
* split()
*/
/* XXX: very messy now, but works; clean up, remove unused variables (nomimally
* used so compiler doesn't complain).
*/
DUK_INTERNAL duk_ret_t duk_bi_string_prototype_split(duk_hthread *thr) {
duk_hstring *h_input;
duk_hstring *h_sep;
duk_uint32_t limit;
duk_uint32_t arr_idx;
#if defined(DUK_USE_REGEXP_SUPPORT)
duk_bool_t is_regexp;
#endif
duk_bool_t matched; /* set to 1 if any match exists (needed for empty input special case) */
duk_uint32_t prev_match_end_coff, prev_match_end_boff;
duk_uint32_t match_start_boff, match_start_coff;
duk_uint32_t match_end_boff, match_end_coff;
h_input = duk_push_this_coercible_to_string(thr);
DUK_ASSERT(h_input != NULL);
duk_push_array(thr);
if (duk_is_undefined(thr, 1)) {
limit = 0xffffffffUL;
} else {
limit = duk_to_uint32(thr, 1);
}
if (limit == 0) {
return 1;
}
/* If the separator is a RegExp, make a "clone" of it. The specification
* algorithm calls [[Match]] directly for specific indices; we emulate this
* by tweaking lastIndex and using a "force global" variant of duk_regexp_match()
* which will use global-style matching even when the RegExp itself is non-global.
*/
if (duk_is_undefined(thr, 0)) {
/* The spec algorithm first does "R = ToString(separator)" before checking
* whether separator is undefined. Since this is side effect free, we can
* skip the ToString() here.
*/
duk_dup_2(thr);
duk_put_prop_index(thr, 3, 0);
return 1;
} else if (duk_get_hobject_with_class(thr, 0, DUK_HOBJECT_CLASS_REGEXP) != NULL) {
#if defined(DUK_USE_REGEXP_SUPPORT)
duk_push_hobject_bidx(thr, DUK_BIDX_REGEXP_CONSTRUCTOR);
duk_dup_0(thr);
duk_new(thr, 1); /* [ ... RegExp val ] -> [ ... res ] */
duk_replace(thr, 0);
/* lastIndex is initialized to zero by new RegExp() */
is_regexp = 1;
#else
DUK_DCERROR_UNSUPPORTED(thr);
#endif
} else {
duk_to_string(thr, 0);
#if defined(DUK_USE_REGEXP_SUPPORT)
is_regexp = 0;
#endif
}
/* stack[0] = separator (string or regexp)
* stack[1] = limit
* stack[2] = input string
* stack[3] = result array
*/
prev_match_end_boff = 0;
prev_match_end_coff = 0;
arr_idx = 0;
matched = 0;
for (;;) {
/*
* The specification uses RegExp [[Match]] to attempt match at specific
* offsets. We don't have such a primitive, so we use an actual RegExp
* and tweak lastIndex. Since the RegExp may be non-global, we use a
* special variant which forces global-like behavior for matching.
*/
DUK_ASSERT_TOP(thr, 4);
#if defined(DUK_USE_REGEXP_SUPPORT)
if (is_regexp) {
duk_dup_0(thr);
duk_dup_2(thr);
duk_regexp_match_force_global(thr); /* [ ... regexp input ] -> [ res_obj ] */
if (!duk_is_object(thr, -1)) {
duk_pop(thr);
break;
}