summaryrefslogtreecommitdiff
path: root/json-perl.c
blob: a39e4ee3dae3ff5c9f4e4a066f8dc393f5217dff (plain)
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
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
/* The C part is broken into three pieces, "json-common.c",
   "json-perl.c", and "json-entry-points.c". This file contains the
   "Perl" stuff, for example if we have a string, the stuff to convert
   it into a Perl hash key or a Perl scalar is in this file. */

/* There are two routes through the code, the PERLING route and the
   non-PERLING route. If we go via the non-PERLING route, we never
   create or alter any Perl-related stuff, we just parse each byte and
   possibly throw an error. This makes validation faster. */

#ifdef PERLING

/* We are creating Perl structures from the JSON. */

#define PREFIX(x) x
#define SVPTR SV *
#define SETVALUE value =

#elif defined(TOKENING)

/* We are just tokenizing the JSON. */

#define PREFIX(x) tokenize_ ## x
#define SVPTR json_token_t *
#define SETVALUE value =

#else /* not def PERLING/TOKENING */

/* Turn off everything to do with creating Perl things. */

#define PREFIX(x) valid_ ## x
#define SVPTR void
#define SETVALUE

#endif /* def PERLING */

/*

This is what INT_MAX_DIGITS is, but #defining it like this causes huge
amounts of unnecessary calculation, so this is commented out.

#define INT_MAX_DIGITS ((int) (log (INT_MAX) / log (10)) - 1)

*/

/* The maximum digits we allow an integer before throwing in the towel
   and returning a Perl string type. */

#define INT_MAX_DIGITS 8

#define USEDIGIT guess = guess * 10 + (c - '0')

static INLINE SVPTR
PREFIX (number) (json_parse_t * parser)
{
    /* End marker for strtod. */

    char * end;

    /* Start marker for strtod. */

    char * start;

    /* A guess for integer numbers. */

    int guess;

    /* The parsed character itself, the cause of our motion. */

    unsigned char c;

    /* If it has exp or dot in it. */

    double d;

    /* Negative number. */

    int minus;

    /* When this is called, it means that a byte indicating a number
       was found. We need to re-examine that byte as a number. */

    parser->end--;
    start = (char *) parser->end;

#define FAILNUMBER(err)				\
    if (STRINGEND &&				\
	parser->top_level_value &&		\
	c == '\0') {				\
	goto exp_number_end;			\
    }						\
    parser->bad_byte = parser->end - 1;		\
    parser->error = json_error_ ## err;		\
    parser->bad_type = json_number;		\
    parser->bad_beginning =			\
	(unsigned char*) start;			\
    failbadinput (parser)

#define NUMBEREND				\
         WHITESPACE:				\
    case ']':					\
    case '}':					\
    case ','

#define XNUMBEREND (XCOMMA|XWHITESPACE|parser->end_expected)

    guess = 0;
    minus = 0;

    switch (NEXTBYTE) {
    case DIGIT19:
	guess = c - '0';
	goto leading_digit19;
    case '0':
	goto leading_zero;
    case '-':
	minus = 1;
	goto leading_minus;
    default:
	parser->expected = XDIGIT | XMINUS;
	FAILNUMBER (unexpected_character);
    }

 leading_digit19:

    switch (NEXTBYTE) {
    case DIGIT:
	USEDIGIT;
	goto leading_digit19;
    case '.':
	goto dot;
    case 'e':
    case 'E':
	goto exp;
    case NUMBEREND:
        goto int_number_end;
    default:
	parser->expected = XDIGIT | XDOT | XEXPONENTIAL | XNUMBEREND;
	if (parser->top_level_value) {
	    parser->expected &= ~XCOMMA;
	}
	FAILNUMBER (unexpected_character);
    }

 leading_zero:
    switch (NEXTBYTE) {
    case '.':
	/* "0." */
	goto dot;
    case 'e':
    case 'E':
	/* "0e" */
	goto exp;
    case NUMBEREND:
	/* "0" */
        goto int_number_end;
    default:
	parser->expected = XDOT | XEXPONENTIAL | XNUMBEREND;
	if (parser->top_level_value) {
	    parser->expected &= ~XCOMMA;
	}
	FAILNUMBER (unexpected_character);
    }

 leading_minus:
    switch (NEXTBYTE) {
    case DIGIT19:
	USEDIGIT;
	goto leading_digit19;
    case '0':
	goto leading_zero;
    default:
	parser->expected = XDIGIT;
	FAILNUMBER (unexpected_character);
    }

    /* Things like "5." are not allowed so there is no NUMBEREND
       here. */

 dot:
    switch (NEXTBYTE) {
    case DIGIT:
	goto dot_digits;
    default:
	parser->expected = XDIGIT;
	FAILNUMBER (unexpected_character);
    }

    /* We have as much as 5.5 so we can stop. */

 dot_digits:
    switch (NEXTBYTE) {
    case DIGIT:
	goto dot_digits;
    case 'e':
    case 'E':
	goto exp;
    case NUMBEREND:
        goto exp_number_end;
    default:
	parser->expected = XDIGIT | XNUMBEREND | XEXPONENTIAL;
	if (parser->top_level_value) {
	    parser->expected &= ~XCOMMA;
	}
	FAILNUMBER (unexpected_character);
    }

    /* Things like "10E" are not allowed so there is no NUMBEREND
       here. */

 exp:
    switch (NEXTBYTE) {
    case '-':
    case '+':
	goto exp_sign;
    case DIGIT:
	goto exp_digits;
    default:
	parser->expected = XDIGIT | XMINUS | XPLUS;
	FAILNUMBER (unexpected_character);
    }

 exp_sign:

    switch (NEXTBYTE) {
    case DIGIT:
	goto exp_digits;
    default:
	parser->expected = XDIGIT;
	FAILNUMBER (unexpected_character);
    }

    /* We have as much as "3.0e1" or similar. */

 exp_digits:
    switch (NEXTBYTE) {
    case DIGIT:
	goto exp_digits;
    case NUMBEREND:
        goto exp_number_end;
    default:
	parser->expected = XDIGIT | XNUMBEREND;
	if (parser->top_level_value) {
	    parser->expected &= ~XCOMMA;
	}
	FAILNUMBER (unexpected_character);
    }

 exp_number_end:
    parser->end--;
#ifdef PERLING
    d = strtod (start, & end);
#else
    strtod (start, & end);
#endif
    if ((unsigned char *) end == parser->end) {
	/* Success, strtod worked as planned. */
#ifdef PERLING
	return newSVnv (d);
#elif defined (TOKENING)
	return json_token_new (parser, (unsigned char *) start,
			       parser->end,
			       json_token_number);
#else
	return;
#endif
    }
    else {
	/* Failure, strtod rejected the number. */
	goto string_number_end;
    }

 int_number_end:

    parser->end--;
    if (parser->end - (unsigned char *) start < INT_MAX_DIGITS + minus) {
	if (minus) {
	    guess = -guess;
	}
	/*
	printf ("number debug: '%.*s': %d\n",
		parser->end - (unsigned char *) start, start, guess);
	*/
#ifdef PERLING
	return newSViv (guess);
#elif defined (TOKENING)
	return json_token_new (parser, (unsigned char *) start,
			       parser->end - 1, json_token_number);
#else
	return;
#endif
    }
    else {
	goto string_number_end;
    }

string_number_end:

    /* We could not convert this number using a number conversion
       routine, so we are going to convert it to a string.  This might
       happen with ridiculously long numbers or something. The JSON
       standard doesn't explicitly disallow integers with a million
       digits. */

#ifdef PERLING
    return newSVpv (start, (STRLEN) ((char *) parser->end - start));
#elif defined (TOKENING)
    return json_token_new (parser, (unsigned char *) start,
			   parser->end - 1, json_token_number);
#else
    return;
#endif
}

#ifdef PERLING

/* This copies our on-stack buffer "buffer" of size "size" into the
   end of a Perl SV called "string". */

#define COPYBUFFER {					\
	if (! string) {					\
	    string = newSVpvn ((char *) buffer, size);	\
	}						\
	else {						\
	    char * svbuf;				\
	    STRLEN cur = SvCUR (string);		\
	    if (SvLEN (string) <= cur + size) {		\
		SvGROW (string, cur + size);		\
	    }						\
	    svbuf = SvPVX (string);			\
	    memcpy (svbuf + cur, buffer, size);		\
	    SvCUR_set (string, cur + size);		\
	}						\
    }

/* The size of the on-stack buffer. */

#define BUFSIZE 0x1000

/* We need a safety margin when dealing with the buffer, for example
   if we hit a Unicode \uabcd escape which needs to be decoded, we
   need to have enough bytes to write into the buffer. */

#define MARGIN 0x10

/* Speedup hack, a special "get_string" for Perl parsing which doesn't
   use parser->buffer but its own buffer on the stack. */

static INLINE SV *
perl_get_string (json_parse_t * parser, STRLEN prefixlen)
{
    unsigned char * b;
    unsigned char c;
    unsigned char * start;
    unsigned char buffer[BUFSIZE];
    STRLEN size;
    SV * string;
    string = 0;
    start = parser->end;
    b = buffer;

    if (prefixlen > 0) {

	/* The string from parser->end to parser->end + prefixlen has
	   already been checked and found not to contain the end of
	   the string or any escapes, so we just copy the memory
	   straight into the buffer. This was supposed to speed things
	   up, but it didn't seem to. However this presumably cannot
	   hurt either. */

	if (prefixlen > BUFSIZE - MARGIN) {
	    /* This is to account for the very unlikely case that the
	       key of the JSON object is more than BUFSIZE - MARGIN
	       bytes long and has an escape after more than BUFSIZE -
	       MARGIN bytes. */
	    prefixlen = BUFSIZE - MARGIN;
	}

	memcpy (buffer, parser->end, prefixlen);
	start += prefixlen;
    }

 string_start:

    size = b - buffer;
    if (size >= BUFSIZE - MARGIN) {
	/* Spot-check for an overflow. */
	if (STRINGEND) {
	    STRINGFAIL (unexpected_end_of_input);
	}
	/* "string_start" is a label for a goto which is applied until
	   we get to the end of the string, so size keeps getting
	   larger and larger.  Now the string being parsed has proved
	   to be too big for our puny BUFSIZE buffer, so we copy the
	   contents of the buffer into the nice Perl scalar. */
	COPYBUFFER;
	/* Set the point of copying bytes back to the beginning of
	   buffer. We don't reset the memory in buffer. */
	b = buffer;
	size = b - buffer;
    }
    NEXTBYTE;

    /* "if" statements seem to compile to something marginally faster
       than "switch" statements, for some reason. */

    if (c < 0x20) {
	ILLEGALBYTE;
    }
    else if (c >= 0x20 && c <= 0x80) {
	/* For some reason or another, putting the following "if"
	   statements after the above one results in about 4% faster
	   code than putting them before it. */
	if (c == '"') {
	    goto string_end;
	}
	if (c == '\\') {
	    HANDLE_ESCAPES (parser->end, start - 1);
	    goto string_start;
	}
	* b++ = c;
	goto string_start;
    }
    else {

	/* Resort to switch statements for the UTF-8 stuff. This
	   actually also contains statements to handle ASCII but they
	   will never be executed. */

	switch (c) {
#define ADDBYTE * b = c; b++
#define startofutf8string start
#include "utf8-byte-one.c"

	default:

	    /* We have to give up, this byte is too mysterious for our
	       weak minds. */

	    ILLEGALBYTE;
	}
    }

 string_end:

    if (STRINGEND) {
	STRINGFAIL (unexpected_end_of_input);
    }

    COPYBUFFER;
    return string;

/* The rest of the UTF-8 stuff goes in here. */

#include "utf8-next-byte.c"
#undef ADDBYTE

    goto string_end;
}

#endif /* PERLING */

static SVPTR
PREFIX (string) (json_parse_t * parser)
{
    unsigned char c;
#ifdef PERLING
    SV * string;
    STRLEN len;
    STRLEN prefixlen;
#elif defined (TOKENING)
    json_token_t * string;
    int len;
#else
    int len;
#endif

    unsigned char * start;

    start = parser->end;
    len = 0;

    /* First of all, we examine the string to work out how long it is
       and to look for escapes. If we find them, we go to "contains_escapes"
       and go back and do all the hard work of converting the escapes
       into the right things. If we don't find any escapes, we just
       use "start" and "len" and copy the string from inside
       "input". This is a trick to increase the speed of
       processing. */

 string_start:
    switch (NEXTBYTE) {
    case '"':
	goto string_end;
    case '\\':
	goto contains_escapes;

#define ADDBYTE len++
#include "utf8-byte-one.c"

	/* Not a fall through. */
    case BADBYTES:
	ILLEGALBYTE;
    }
    /* Parsing of the string ended due to a \0 byte flipping the
       "while" switch and we dropped into this section before
       reaching the string's end. */
    ILLEGALBYTE;

#include "utf8-next-byte.c"
#undef ADDBYTE

 string_end:

#ifdef PERLING

    /* Our string didn't contain any escape sequences, so we can just
       make a new SV * by copying the string from "start", the old
       position within the thing we're parsing to start + len. */

    string = newSVpvn ((char *) start, len);

#elif defined (TOKENING)

    string = json_token_new (parser, start - 1,
			     start + len,
			     json_token_string);

#endif

    goto string_done;

 contains_escapes:

#ifdef PERLING

    /* Use "perl_get_string" which keeps the buffer on the
       stack. Results in a minor speed increase. */
    parser->end = start;
    prefixlen = (STRLEN) (parser->end - start);
    string = perl_get_string (parser, prefixlen);

#elif defined (TOKENING)
    /* Don't use "len" here since it subtracts the escapes. */
    parser->end = start;
    len = get_string (parser);
    string = json_token_new (parser,
			     /* Location of first quote. */
			     start - 1,
			     /* Location of last quote. */
			     parser->end - 1,
			     json_token_string);
#else
    parser->end = start;
    len = get_string (parser);
#endif

 string_done:

#ifdef PERLING
    if (parser->unicode || parser->force_unicode || parser->upgrade_utf8) {
	SvUTF8_on (string);
	parser->force_unicode = 0;
    }
#endif

#if defined (PERLING) || defined (TOKENING)
    return string;
#else
    return;
#endif
}

#define FAILLITERAL(c)					\
    parser->expected = XIN_LITERAL;			\
    parser->literal_char = c;				\
    parser->bad_beginning = start;			\
    parser->error = json_error_unexpected_character;	\
    parser->bad_type = json_literal;			\
    parser->bad_byte = parser->end - 1;			\
    failbadinput (parser)

static SVPTR
PREFIX (literal_true) (json_parse_t * parser)
{
    unsigned char * start;
    start = parser->end - 1;
    if (* parser->end++ == 'r') {
	if (* parser->end++ == 'u') {
	    if (* parser->end++ == 'e') {
#ifdef PERLING
		if (parser->user_true) {
		    return newSVsv (parser->user_true);
		}
		else if (parser->copy_literals) {
		    return newSVsv (&PL_sv_yes);
		}
		else {
		    return &PL_sv_yes;
		}
#elif defined (TOKENING)
		return json_token_new (parser, start, parser->end - 1,
				       json_token_literal);
#else
		return;
#endif
	    }
	    FAILLITERAL ('e');
	}
	FAILLITERAL ('u');
    }
    FAILLITERAL ('r');
}

static SVPTR
PREFIX (literal_false) (json_parse_t * parser)
{
    unsigned char * start;
    start = parser->end - 1;
    if (* parser->end++ == 'a') {
	if (* parser->end++ == 'l') {
	    if (* parser->end++ == 's') {
		if (* parser->end++ == 'e') {
#ifdef PERLING
		if (parser->user_false) {
		    return newSVsv (parser->user_false);
		}
		else if (parser->copy_literals) {
		    return newSVsv (&PL_sv_no);
		}
		else {
		    return &PL_sv_no;
		}
#elif defined (TOKENING)
		return json_token_new (parser, start, parser->end - 1,
				       json_token_literal);
#else
		return;
#endif
		}
		FAILLITERAL ('e');
	    }
	    FAILLITERAL ('s');
	}
	FAILLITERAL ('l');
    }
    FAILLITERAL ('a');
}

static SVPTR
PREFIX (literal_null) (json_parse_t * parser)
{
    unsigned char * start;
    start = parser->end - 1;
    if (* parser->end++ == 'u') {
	if (* parser->end++ == 'l') {
	    if (* parser->end++ == 'l') {
#ifdef PERLING
		if (parser->user_null) {
		    return newSVsv (parser->user_null);
		}
		else if (parser->copy_literals) {
		    return newSVsv (&PL_sv_undef);
		}
		else {
		    SvREFCNT_inc (json_null);
		    return json_null;
		}
#elif defined (TOKENING)
		return json_token_new (parser, start, parser-> end - 1,
				       json_token_literal);
#else
		return;
#endif
	    }
	    FAILLITERAL ('l');
	}
	FAILLITERAL ('l');
    }
    FAILLITERAL ('u');
}

static SVPTR PREFIX (object) (json_parse_t * parser);

/* Given one character, decide what to do next. This goes in the
   switch statement in both "object ()" and "array ()". */

#define PARSE(start,expected)			\
						\
 case WHITESPACE:				\
 goto start;					\
						\
 case '"':					\
 SETVALUE PREFIX (string) (parser);		\
 break;						\
						\
 case '-':					\
 case DIGIT:					\
 parser->end_expected = expected;	        \
 SETVALUE PREFIX (number) (parser);		\
 break;						\
						\
 case '{':					\
 INCDEPTH;					\
 SETVALUE PREFIX (object) (parser);		\
 break;						\
						\
 case '[':					\
 INCDEPTH;					\
 SETVALUE PREFIX (array) (parser);		\
 break;						\
						\
 case 'f':					\
 SETVALUE PREFIX (literal_false) (parser);	\
 break;			                        \
						\
 case 'n':					\
 SETVALUE PREFIX (literal_null) (parser);	\
 break;			                        \
						\
 case 't':					\
 SETVALUE PREFIX (literal_true) (parser);	\
 break

#define FAILARRAY(err)				\
    parser->bad_byte = parser->end - 1;		\
    parser->bad_type = json_array;		\
    parser->bad_beginning = start;		\
    parser->error = json_error_ ## err;		\
    failbadinput (parser)

/* We have seen "[", so now deal with the contents of an array. At the
   end of this routine, "parser->end" is pointing one beyond the final
   "]" of the array. */

static SVPTR
PREFIX (array) (json_parse_t * parser)
{
    unsigned char c;
    unsigned char * start;
#ifdef PERLING
    AV * av;
    SV * value = & PL_sv_undef;
#elif defined (TOKENING)
    json_token_t * av;
    json_token_t * prev;
    json_token_t * value;
#endif

    start = parser->end - 1;
#ifdef PERLING
    av = newAV ();
#elif defined (TOKENING)
    av = json_token_new (parser, start, 0, json_token_array);
    prev = 0;
#endif

 array_start:

    switch (NEXTBYTE) {

	PARSE (array_start, XARRAY_END);

    case ']':
	goto array_end;

    default:
	parser->expected = VALUE_START | XWHITESPACE | XARRAY_END;
	FAILARRAY (unexpected_character);
    }

#ifdef PERLING
    av_push (av, value);
#elif defined (TOKENING)
    prev = json_token_set_child (parser, av, value);
#endif

    /* Accept either a comma or whitespace or the end of the array. */

 array_middle:

    switch (NEXTBYTE) {

    case WHITESPACE:
	goto array_middle;

    case ',':
#ifdef TOKENING
	value = json_token_new (parser, parser->end - 1,
				parser->end - 1,
				json_token_comma);
	prev = json_token_set_next (prev, value);
#endif
	goto array_next;

    case ']':
	/* Array with at least one element. */
	goto array_end;

    default:

	parser->expected = XWHITESPACE | XCOMMA | XARRAY_END;
	FAILARRAY (unexpected_character);
    }

 array_next:

    switch (NEXTBYTE) {

	PARSE (array_next, XARRAY_END);

    default:
	parser->expected = VALUE_START | XWHITESPACE;
	FAILARRAY (unexpected_character);
    }

#ifdef PERLING
    av_push (av, value);
#elif defined (TOKENING)
    prev = json_token_set_next (prev, value);
#endif

    goto array_middle;

 array_end:
    DECDEPTH;

#ifdef PERLING
    return newRV_noinc ((SV *) av);
#elif defined (TOKENING)
    /* We didn't know where the end was until now. */
    json_token_set_end (parser, av, parser->end - 1);
    return av;
#else
    return;
#endif
}

#define FAILOBJECT(err)				\
    parser->bad_byte = parser->end - 1;		\
    parser->bad_type = json_object;		\
    parser->bad_beginning = start;		\
    parser->error = json_error_ ## err;		\
    failbadinput (parser)

/* We have seen "{", so now deal with the contents of an object. At
   the end of this routine, "parser->end" is pointing one beyond the
   final "}" of the object. */

static SVPTR
PREFIX (object) (json_parse_t * parser)
{
    char c;
#ifdef PERLING
    HV * hv;
    SV * value;
    /* This is set to -1 if we want a Unicode key. See "perldoc
       perlapi" under "hv_store". */
    int uniflag;
#elif defined (TOKENING)
    json_token_t * hv;
    json_token_t * value;
    json_token_t * prev;
#endif
    string_t key;
    /* Start of parsing. */
    unsigned char * start;

    start = parser->end - 1;

#ifdef PERLING
    if (parser->unicode || parser->upgrade_utf8) {
	/* Keys are unicode. */
	uniflag = -1;
    }
    else {
	/* Keys are not unicode. */
	uniflag = 1;
    }
    hv = newHV ();
#elif defined (TOKENING)
    hv = json_token_new (parser, start, 0, json_token_object);
    prev = 0;
#endif

 hash_start:

    switch (NEXTBYTE) {
    case WHITESPACE:
	goto hash_start;
    case '}':
	goto hash_end;
    case '"':
#ifdef TOKENING
	value = json_token_new (parser, parser->end - 1, 0,
				json_token_string);
	/* We only come past the label "hash_start" once, so we don't
	   need to check that there is not already a child. */
	json_token_set_child (parser, hv, value);
	prev = value;
#endif
	get_key_string (parser, & key);
#ifdef TOKENING
	/* We didn't know where the end of the string was until now so
	   we wait until after "get_key_string" to set the end. */
	json_token_set_end (parser, value, parser->end - 1);
#endif
	goto hash_next;
    default:
	parser->expected = XWHITESPACE | XSTRING_START | XOBJECT_END;
	FAILOBJECT (unexpected_character);
    }

 hash_middle:

    /* We are in the middle of a hash. We have seen a key:value pair,
       and now we're looking for either a comma and then another
       key-value pair, or a closing curly brace and the end of the
       hash. */

    switch (NEXTBYTE) {
    case WHITESPACE:
	goto hash_middle;
    case '}':
	goto hash_end;
    case ',':
#ifdef TOKENING
	value = json_token_new (parser, parser->end - 1,
				parser->end - 1,
				json_token_comma);
	prev = json_token_set_next (prev, value);
#endif
	goto hash_key;
    default:
	parser->expected = XWHITESPACE | XCOMMA | XOBJECT_END;
	FAILOBJECT (unexpected_character);
    }

 hash_key:

    /* We're looking for a key in the hash, which is a string starting
       with a double quotation mark. */

    switch (NEXTBYTE) {
    case WHITESPACE:
	goto hash_key;
    case '"':
#ifdef TOKENING
	value = json_token_new (parser, parser->end - 1, 0,
				json_token_string);
	prev = json_token_set_next (prev, value);
#endif
	get_key_string (parser, & key);
#ifdef TOKENING
	/* We didn't know where the end of the string was until now so
	   we wait until after "get_key_string" to set the end. */
	json_token_set_end (parser, value, parser->end - 1);
#endif
	goto hash_next;
    default:
	parser->expected = XWHITESPACE | XSTRING_START;
	FAILOBJECT (unexpected_character);
    }

 hash_next:

    /* We've seen a key, now we're looking for a colon. */

    switch (NEXTBYTE) {
    case WHITESPACE:
	goto hash_next;
    case ':':
#ifdef TOKENING
	value = json_token_new (parser, parser->end - 1,
				parser->end - 1,
				json_token_colon);
	prev = json_token_set_next (prev, value);
#endif
	goto hash_value;
    default:
	parser->expected = XWHITESPACE | XVALUE_SEPARATOR;
	FAILOBJECT (unexpected_character);
    }

 hash_value:

    /* We've seen a colon, now we're looking for a value, which can be
       anything at all, including another hash. Most of the cases are
       dealt with in the PARSE macro. */

    switch (NEXTBYTE) {
	PARSE (hash_value, XOBJECT_END);
    default:
	parser->expected = XWHITESPACE | VALUE_START;
	FAILOBJECT (unexpected_character);
    }

    if (key.contains_escapes) {

	/* The key had something like "\n" in it, so we can't just
	   copy the value but have to process it to remove the
	   escapes. */

#ifdef PERLING
	int klen;
	klen = resolve_string (parser, & key);
	key.start = parser->buffer;
	key.length = klen;
#else
	resolve_string (parser, & key);
#endif
    }
#ifdef PERLING
    if (parser->detect_collisions) {
	/* Look in hv for an existing key with our values. */
	SV ** sv_ptr;
	sv_ptr = hv_fetch (hv, (char *) key.start, key.length * uniflag, 0);
	if (sv_ptr) {
	    parser->bad_byte = key.start;
	    parser->bad_length = key.length;
	    parser->bad_type = json_object;
	    parser->bad_beginning = start;
	    parser->error = json_error_name_is_not_unique;
	    failbadinput (parser);
	}
    }
    (void) hv_store (hv, (char *) key.start, key.length * uniflag, value, 0);
#endif

#if defined(TOKENING)
    prev = json_token_set_next (prev, value);
#endif
    goto hash_middle;

 hash_end:
    DECDEPTH;

#ifdef PERLING
    return newRV_noinc ((SV *) hv);
#elif defined (TOKENING)
    json_token_set_end (parser, hv, parser->end - 1);
    return hv;
#else
    return;
#endif
}

#undef PREFIX
#undef SVPTR
#undef SETVALUE

#ifdef PERLING

/* Set and delete user-defined literals. */

static void
json_parse_delete_true (json_parse_t * parser)
{
    if (parser->user_true) {
	SvREFCNT_dec (parser->user_true);
	parser->user_true = 0;
    }
}

static void
json_parse_set_true (json_parse_t * parser, SV * user_true)
{
    json_parse_delete_true (parser);
    if (! SvTRUE (user_true) && ! parser->no_warn_literals) {
	warn ("User-defined value for JSON true evaluates as false");
    }
    if (parser->copy_literals && ! parser->no_warn_literals) {
	warn ("User-defined value overrules copy_literals");
    }
    parser->user_true = user_true;
    SvREFCNT_inc (user_true);
}

static void
json_parse_delete_false (json_parse_t * parser)
{
    if (parser->user_false) {
	SvREFCNT_dec (parser->user_false);
	parser->user_false = 0;
    }
}

static void
json_parse_set_false (json_parse_t * parser, SV * user_false)
{
    json_parse_delete_false (parser);
    if (SvTRUE (user_false) && ! parser->no_warn_literals) {
	warn ("User-defined value for JSON false evaluates as true");
    }
    if (parser->copy_literals && ! parser->no_warn_literals) {
	warn ("User-defined value overrules copy_literals");
    }
    parser->user_false = user_false;
    SvREFCNT_inc (user_false);
}

static void
json_parse_delete_null (json_parse_t * parser)
{
    if (parser->user_null) {
	SvREFCNT_dec (parser->user_null);
	parser->user_null = 0;
    }
}

static void
json_parse_set_null (json_parse_t * parser, SV * user_null)
{
    if (parser->copy_literals && ! parser->no_warn_literals) {
	warn ("User-defined value overrules copy_literals");
    }
    json_parse_delete_null (parser);
    parser->user_null = user_null;
    SvREFCNT_inc (user_null);
}

static void
json_parse_free (json_parse_t * parser)
{
    /* We can get here with depth > 0 if the parser fails and then the
       error is caught. */
    if (parser->depth < 0) {
	warn ("Parser depth underflow %d", parser->depth);
    }
    json_parse_delete_true (parser);
    json_parse_delete_false (parser);
    json_parse_delete_null (parser);
    Safefree (parser);
}

static void
json_parse_copy_literals (json_parse_t * parser, SV * onoff)
{
    if (! parser->no_warn_literals &&
	(parser->user_true || parser->user_false || parser->user_null)) {
	warn ("User-defined value overrules copy_literals");
    }
    parser->copy_literals = SvTRUE (onoff) ? 1 : 0;
}

#endif /* def PERLING */