Newer
Older
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
/* -*- mode: c -*- */
/*
number.c -- Numeric constants.
*/
/*
Copyright (c) 1984, Taiichi Yuasa and Masami Hagiya.
Copyright (c) 1990, Giuseppe Attardi.
Copyright (c) 2001, Juan Jose Garcia Ripoll.
Copyright (c) 2011, Jean-Claude Beaudoin.
MKCL is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 3 of the License, or (at your option) any later version.
See file '../../Copyright' for full details.
*/
#include <mkcl/mkcl.h>
#include <mkcl/mkcl-math.h>
#include <float.h>
#include <mkcl/internal.h>
#include <mkcl/mkcl-fenv.h>
#if defined(MKCL_IEEE_FP) && defined(HAVE_FEENABLEEXCEPT)
/*
* We are using IEEE arithmetics and can rely on FPE exceptions
* to be raised when invalid operations are performed.
*/
# define DO_DETECT_FPE(e, f) mkcl_detect_fpe()
#else
/*
* Either we can not rely on C signals or we do not want IEEE NaNs and
* infinities. The first case typically happens for instance under OS
* X, where the status of the FPE control word is changed by
* printf. We have two alternatives.
*/
# ifdef MKCL_IEEE_FP
# if defined(HAVE_FENV_H) && !defined(MKCL_AVOID_FENV_H)
# define DO_DETECT_FPE(e, f) \
if (isnan(f) || !isfinite(f)) \
mkcl_deliver_fpe(e);
# else
# define DO_DETECT_FPE(e, f) \
if (isnan(f)) { \
if ((e)->fpe_control_bits & FE_INVALID) \
mk_cl_error(e,1, @'floating-point-invalid-operation'); \
} else if (!isfinite(f)) { \
if ((e)->fpe_control_bits & FE_DIVBYZERO) \
mk_cl_error(e,1, @'division-by-zero'); \
}
# endif
# else
# define DO_DETECT_FPE(e, f) \
if (isnan(f)) { \
mk_cl_error(e, 1, @'floating-point-invalid-operation'); \
} else if (!isfinite(f)) { \
mk_cl_error(e, 1, @'division-by-zero'); \
}
# endif
#endif
mkcl_word
mkcl_integer_to_word(MKCL, mkcl_object x)
{
if (MKCL_FIXNUMP(x))
return mkcl_fixnum_to_word(x);
if (mkcl_type_of(x) == mkcl_t_bignum) {
#if MKCL_LONG_BITS >= MKCL_WORD_BITS
if (mpz_fits_slong_p(x->big.big_num)) {
return mpz_get_si(x->big.big_num);
}
#elif MKCL_WORD_BITS == 64
return mkcl_to_int64_t(env, x);
#else
#error "Don't know how to convert from lisp bignum to native C integer."
#endif
}
#if MKCL_WORD_BITS == 64
mkcl_FEwrong_type_argument(env, @'mkcl::integer64', x);
#elif MKCL_WORD_BITS == 32
mkcl_FEwrong_type_argument(env, @'mkcl::integer32', x);
#else
mkcl_FEwrong_type_argument(env, @'integer', x);
#endif
}
mkcl_index
mkcl_integer_to_index(MKCL, mkcl_object x)
{
if (MKCL_FIXNUMP(x)) {
mkcl_word i = mkcl_fixnum_to_word(x);
if (i >= 0)
return i;
} else if (mkcl_type_of(x) == mkcl_t_bignum) {
#if MKCL_LONG_BITS >= MKCL_WORD_BITS
if (mpz_fits_ulong_p(x->big.big_num)) {
return mpz_get_ui(x->big.big_num);
}
#elif MKCL_WORD_BITS == 64
return mkcl_to_uint64_t(env, x);
#else
#error "Don't know how to convert from lisp bignum to native C unsigned integer."
#endif
}
mk_cl_error(env, 9, @'simple-type-error', @':format-control',
mkcl_make_simple_base_string(env, "Not a non-negative fixnum ~S"),
@':format-arguments', mk_cl_list(env, 1,x),
@':expected-type',
#if MKCL_WORD_BITS == 64
@'mkcl::natural64',
#elif MKCL_WORD_BITS == 32
@'mkcl::natural32',
#else
mk_cl_list(env, 3,
@'integer',
MKCL_MAKE_FIXNUM(0),
mkcl_one_minus(env, mkcl_ash(env,
MKCL_MAKE_FIXNUM(1),
MKCL_WORD_BITS))
),
#endif
@':datum', x);
}
mkcl_word
mkcl_fixnum_in_range(MKCL, mkcl_object fun, const char *what, mkcl_object value,
mkcl_word min, mkcl_word max)
{
do {
if (MKCL_FIXNUMP(value)) {
mkcl_word output = mkcl_fixnum_to_word(value);
if ((min <= output) && (output <= max)) {
return output;
}
}
value = mkcl_type_error(env, fun, what, value,
mk_cl_list(env, 3,
@'integer',
MKCL_MAKE_FIXNUM(min),
MKCL_MAKE_FIXNUM(max)));
} while(1);
}
mkcl_object mkcl_make_big_integer(MKCL, mkcl_word l)
{
#if MKCL_WORD_BITS > MKCL_LONG_BITS
return mkcl_make_int64_t(env, l);
#else
mkcl_object z = _mkcl_big_register0();
_mkcl_big_set_si(z, l);
return _mkcl_big_register_copy(env, z);
#endif
}
mkcl_object mkcl_make_big_unsigned_integer(MKCL, mkcl_index l)
{
#if MKCL_WORD_BITS > MKCL_LONG_BITS
return mkcl_make_uint64_t(env, l);
#else
mkcl_object z = _mkcl_big_register0();
_mkcl_big_set_ui(z, l);
return _mkcl_big_register_copy(env, z);
#endif
}
mkcl_uint8_t
mkcl_to_uint8_t(MKCL, mkcl_object x)
{
do {
if (MKCL_FIXNUMP(x)) {
mkcl_word y = mkcl_fixnum_to_word(x);
if (y >= 0 && y <= UINT8_MAX) {
return (uint8_t)y;
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3, @'integer', MKCL_MAKE_FIXNUM(0), MKCL_MAKE_FIXNUM(UINT8_MAX)));
} while(1);
}
mkcl_int8_t
mkcl_to_int8_t(MKCL, mkcl_object x)
{
do {
if (MKCL_FIXNUMP(x)) {
mkcl_word y = mkcl_fixnum_to_word(x);
if (y >= INT8_MIN && y <= INT8_MAX) {
return (int8_t)y;
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3, @'integer', MKCL_MAKE_FIXNUM(INT8_MIN), MKCL_MAKE_FIXNUM(INT8_MAX)));
} while(1);
}
#if MKCL_WORD_BITS < 32
# error "Unsupported platform with mkcl_word < mkcl_uint32_t"
#endif
mkcl_uint16_t
mkcl_to_uint16_t(MKCL, mkcl_object x)
{
do {
if (MKCL_FIXNUMP(x)) {
mkcl_word y = mkcl_fixnum_to_word(x);
if (y >= 0 && y <= UINT16_MAX) {
return (mkcl_uint16_t)y;
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3, @'integer', MKCL_MAKE_FIXNUM(0), MKCL_MAKE_FIXNUM(UINT16_MAX)));
} while(1);
}
mkcl_int16_t
mkcl_to_int16_t(MKCL, mkcl_object x)
{
do {
if (MKCL_FIXNUMP(x)) {
mkcl_word y = mkcl_fixnum_to_word(x);
if (y >= INT16_MIN && y <= INT16_MAX) {
return (mkcl_int16_t)y;
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3, @'integer', MKCL_MAKE_FIXNUM(INT16_MIN), MKCL_MAKE_FIXNUM(INT16_MAX)));
} while(1);
}
#if (MKCL_WORD_BITS > 32)
mkcl_uint32_t
mkcl_to_uint32_t(MKCL, mkcl_object x)
{
do {
if (MKCL_FIXNUMP(x)) {
mkcl_word y = mkcl_fixnum_to_word(x);
if (y >= 0 && y <= UINT32_MAX) {
return (mkcl_uint32_t)y;
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3, @'integer', MKCL_MAKE_FIXNUM(0), mkcl_make_unsigned_integer(env, UINT32_MAX)));
} while(1);
}
mkcl_int32_t
mkcl_to_int32_t(MKCL, mkcl_object x)
{
do {
if (MKCL_FIXNUMP(x)) {
mkcl_word y = mkcl_fixnum_to_word(x);
if (y >= INT32_MIN && y <= INT32_MAX) {
return (mkcl_int32_t)y;
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3,
@'integer',
mkcl_make_integer(env, INT32_MIN),
mkcl_make_integer(env, INT32_MAX)));
} while(1);
}
#endif /* (MKCL_WORD_BITS > 32) */
#if (MKCL_WORD_BITS < 64) || (MKCL_LONG_BITS < 64)
mkcl_uint64_t
mkcl_to_uint64_t(MKCL, mkcl_object x)
{
do {
if (!mkcl_minusp(env, x)) {
if (MKCL_FIXNUMP(x)) {
return mkcl_fixnum_to_word(x);
} else if (mkcl_type_of(x) != mkcl_t_bignum) {
(void)0;
} else if (mpz_fits_ulong_p(x->big.big_num)) {
return mpz_get_ui(x->big.big_num);
} else {
mkcl_object copy = _mkcl_big_register0();
mpz_fdiv_q_2exp(copy->big.big_num, x->big.big_num, 32);
if (mpz_fits_ulong_p(copy->big.big_num)) {
mkcl_uint64_t output = mpz_get_ui(copy->big.big_num);
output = (output << 32) + mpz_get_ui(x->big.big_num);
return output;
}
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3,@'integer',MKCL_MAKE_FIXNUM(0),
mkcl_one_minus(env, mkcl_ash(env, MKCL_MAKE_FIXNUM(1), 64))));
} while(1);
}
mkcl_int64_t
mkcl_to_int64_t(MKCL, mkcl_object x)
{
do {
if (MKCL_FIXNUMP(x)) {
return mkcl_fixnum_to_word(x);
} else if (mkcl_type_of(x) != mkcl_t_bignum) {
(void)0;
} else if (mpz_fits_slong_p(x->big.big_num)) {
return mpz_get_si(x->big.big_num);
} else {
mkcl_object copy = _mkcl_big_register0();
mpz_fdiv_q_2exp(copy->big.big_num, x->big.big_num, 32);
if (mpz_fits_slong_p(copy->big.big_num)) {
mkcl_int64_t output = mpz_get_si(copy->big.big_num);
mpz_fdiv_r_2exp(copy->big.big_num, x->big.big_num, 32);
return (output << 32) + mpz_get_ui(copy->big.big_num);
}
}
x = mkcl_type_error(env, @'coerce', "variable", x,
mk_cl_list(env, 3,@'integer',
mkcl_negate(env, mkcl_ash(env, MKCL_MAKE_FIXNUM(1), 63)),
mkcl_one_minus(env, mkcl_ash(env, MKCL_MAKE_FIXNUM(1), 63))));
} while(1);
}
mkcl_object
mkcl_make_uint64_t(MKCL, mkcl_uint64_t i)
{
if (i <= MKCL_MOST_POSITIVE_FIXNUM) {
return MKCL_MAKE_FIXNUM(i);
} else if (i <= ~((mkcl_uint32_t)0)) {
return mkcl_make_uint32_t(env, i);
} else {
mkcl_object aux = mkcl_make_uint32_t(env, i >> 32);
return mk_cl_logior(env, 2, mkcl_ash(env, aux, 32), mkcl_make_uint32_t(env, (mkcl_uint32_t)i));
}
}
mkcl_object
mkcl_make_int64_t(MKCL, mkcl_int64_t i)
{
if (i >= MKCL_MOST_NEGATIVE_FIXNUM && i <= MKCL_MOST_POSITIVE_FIXNUM) {
return MKCL_MAKE_FIXNUM(i);
} else {
mkcl_object aux = mkcl_make_int32_t(env, i >> 32);
mkcl_object aux0 = mkcl_ash(env, aux, 32);
mkcl_object aux1 = mkcl_make_uint32_t(env, (mkcl_uint32_t)i);
mkcl_object aux2 = mk_cl_logior(env, 2, aux0, aux1);
return aux2;
}
}
#endif /* (MKCL_WORD_BITS < 64) */
mkcl_ulong_long_t
mkcl_to_ulong_long(MKCL, mkcl_object x) {
return (mkcl_ulong_long_t)mkcl_to_uint64_t(env, x);
}
mkcl_long_long_t
mkcl_to_long_long(MKCL, mkcl_object x) {
return (mkcl_long_long_t)mkcl_to_int64_t(env, x);
}
mkcl_object
mkcl_make_ulong_long(MKCL, mkcl_ulong_long_t i) {
return mkcl_make_uint64_t(env, i);
}
mkcl_object
mkcl_make_long_long(MKCL, mkcl_long_long_t i) {
return mkcl_make_int64_t(env, i);
}
mkcl_object
mkcl_make_ratio(MKCL, mkcl_object num, mkcl_object den)
{
mkcl_object g, r;
/* INV: the arguments NUM & DEN are integers */
if (den == MKCL_MAKE_FIXNUM(0))
mkcl_FEdivision_by_zero(env, num, den);
if (num == MKCL_MAKE_FIXNUM(0) || den == MKCL_MAKE_FIXNUM(1))
return(num);
if (mkcl_minusp(env, den)) {
num = mkcl_negate(env, num);
den = mkcl_negate(env, den);
}
g = mkcl_gcd(env, num, den);
if (g != MKCL_MAKE_FIXNUM(1)) {
num = mkcl_integer_divide(env, num, g);
den = mkcl_integer_divide(env, den, g);
}
if (den == MKCL_MAKE_FIXNUM(1))
return num;
if (den == MKCL_MAKE_FIXNUM(-1))
return mkcl_negate(env, num);
r = mkcl_alloc_raw_ratio(env);
r->ratio.num = num;
r->ratio.den = den;
return(r);
}
#if defined(HAVE_FENV_H) && !defined(MKCL_AVOID_FENV_H)
void
mkcl_deliver_fpe(MKCL)
{
int bits = env->fpe_control_bits;
if (fetestexcept(env->fpe_control_bits)) {
mkcl_object condition;
if (fetestexcept(bits & FE_DIVBYZERO))
condition = @'division-by-zero';
else if (fetestexcept(bits & FE_INVALID))
condition = @'floating-point-invalid-operation';
else if (fetestexcept(bits & FE_OVERFLOW))
condition = @'floating-point-overflow';
else if (fetestexcept(bits & FE_UNDERFLOW))
condition = @'floating-point-underflow';
else if (fetestexcept(bits & FE_INEXACT))
condition = @'floating-point-inexact';
else
condition = @'arithmetic-error';
feclearexcept(FE_ALL_EXCEPT);
mk_cl_error(env, 1, condition);
}
feclearexcept(FE_ALL_EXCEPT);
}
#endif
mkcl_object
mkcl_make_singlefloat(MKCL, float f)
{
mkcl_object x;
DO_DETECT_FPE(env, f); /* What the F!!! for? JCB */
if (f == 0.0F) {
#if defined(MKCL_SIGNED_ZERO)
if (signbit(f))
return mkcl_core.singlefloat_minus_zero;
#endif
return mkcl_core.singlefloat_zero;
}
x = mkcl_alloc_raw_singlefloat(env);
mkcl_single_float(x) = f;
return(x);
}
mkcl_object
mkcl_make_doublefloat(MKCL, double f)
{
mkcl_object x;
DO_DETECT_FPE(env, f); /* What the F!!! for? JCB */
if (f == 0.0) {
#if defined(MKCL_SIGNED_ZERO)
if (signbit(f))
return mkcl_core.doublefloat_minus_zero;
#endif
return mkcl_core.doublefloat_zero;
}
x = mkcl_alloc_raw_doublefloat(env);
mkcl_double_float(x) = f;
return(x);
}
#ifdef MKCL_LONG_FLOAT
mkcl_object
mkcl_make_longfloat(MKCL, long double f)
{
mkcl_object x;
DO_DETECT_FPE(env, f); /* What the F!!! for? JCB */
if (f == 0.0L) {
#if defined(MKCL_SIGNED_ZERO)
if (signbit(f))
return mkcl_core.longfloat_minus_zero;
#endif
return mkcl_core.longfloat_zero;
}
x = mkcl_alloc_raw_longfloat(env);
x->longfloat.value = f;
return x;
}
#else
mkcl_object
mkcl_make_longfloat(MKCL, long double f)
{
return mkcl_make_doublefloat(env, f); /* downcast. */
}
#endif
mkcl_object
mkcl_make_complex(MKCL, mkcl_object r, mkcl_object i)
{
mkcl_object c;
mkcl_type ti;
AGAIN:
ti = mkcl_type_of(i);
/* Both R and I are promoted to a common type */
switch (mkcl_type_of(r)) {
case mkcl_t_fixnum:
case mkcl_t_bignum:
case mkcl_t_ratio:
switch (ti) {
case mkcl_t_fixnum:
if (i == MKCL_MAKE_FIXNUM(0))
return(r);
case mkcl_t_bignum:
case mkcl_t_ratio:
break;
case mkcl_t_singlefloat:
r = mkcl_make_singlefloat(env, (float)mkcl_to_double(env, r));
break;
case mkcl_t_doublefloat:
r = mkcl_make_doublefloat(env, mkcl_to_double(env, r));
break;
#ifdef MKCL_LONG_FLOAT
case mkcl_t_longfloat:
r = mkcl_make_longfloat(env, mkcl_to_double(env, r));
break;
#endif
default:
i = mkcl_type_error(env, @'complex',"imaginary part", i, @'real');
goto AGAIN;
}
break;
case mkcl_t_singlefloat:
switch (ti) {
case mkcl_t_fixnum:
case mkcl_t_bignum:
case mkcl_t_ratio:
i = mkcl_make_singlefloat(env, (float)mkcl_to_double(env, i));
break;
case mkcl_t_singlefloat:
break;
case mkcl_t_doublefloat:
r = mkcl_make_doublefloat(env, (double)(mkcl_single_float(r)));
break;
#ifdef MKCL_LONG_FLOAT
case mkcl_t_longfloat:
r = mkcl_make_longfloat(env, (long double)mkcl_single_float(r));
break;
#endif
default:
i = mkcl_type_error(env, @'complex',"imaginary part", i, @'real');
goto AGAIN;
}
break;
case mkcl_t_doublefloat:
switch (ti) {
case mkcl_t_fixnum:
case mkcl_t_bignum:
case mkcl_t_ratio:
case mkcl_t_singlefloat:
i = mkcl_make_doublefloat(env, mkcl_to_double(env, i));
case mkcl_t_doublefloat:
break;
#ifdef MKCL_LONG_FLOAT
case mkcl_t_longfloat:
r = mkcl_make_longfloat(env, (long double)mkcl_double_float(r));
break;
#endif
default:
i = mkcl_type_error(env, @'complex',"imaginary part", i, @'real');
goto AGAIN;
}
break;
#ifdef MKCL_LONG_FLOAT
case mkcl_t_longfloat:
if (ti != mkcl_t_longfloat)
i = mkcl_make_longfloat(env, (long double)mkcl_to_double(env, i));
break;
#endif
default:
r = mkcl_type_error(env, @'complex',"real part", r, @'real');
goto AGAIN;
}
c = mkcl_alloc_raw_complex(env);
c->_complex.real = r;
c->_complex.imag = i;
return(c);
}
static mkcl_object
into_bignum(MKCL, mkcl_object bignum, mkcl_object integer)
{
if (MKCL_FIXNUMP(integer)) {
#if MKCL_LONG_BITS < MKCL_WORD_BITS
_mkcl_big_set_fixnum(bignum, mkcl_fixnum_to_word(integer));
#else
mpz_set_si(bignum->big.big_num, mkcl_fixnum_to_word(integer));
#endif
} else {
mpz_set(bignum->big.big_num, integer->big.big_num);
}
return bignum;
}
static mkcl_word
remove_zeros(MKCL, mkcl_object *integer)
{
mkcl_object buffer = into_bignum(env, _mkcl_big_register0(), *integer);
unsigned long den_twos = mpz_scan1(buffer->big.big_num, 0);
if (den_twos < ULONG_MAX) {
mpz_div_2exp(buffer->big.big_num, buffer->big.big_num, den_twos);
*integer = _mkcl_big_register_normalize(env, buffer);
return -den_twos;
} else {
_mkcl_big_register_free(env, buffer);
return 0;
}
}
static mkcl_object
prepare_ratio_to_float(MKCL, mkcl_object num, mkcl_object den,
int digits, mkcl_word *scaleout)
{
/* We have to cook our own routine because GMP does not round.
* The recipe is simple: we multiply the numberator by a large
* enough number so that the division by the denominator fits
* the floating point number. The result is scaled back by the
* appropriate exponent.
*/
/* Scale down the denominator, eliminating the zeros
* so that we have smaller operands.
*/
mkcl_word scale = remove_zeros(env, &den);
mkcl_word num_size, delta;
num_size = mkcl_integer_length(env, num);
delta = mkcl_integer_length(env, den) - num_size;
scale -= delta;
{
mkcl_word adjust = digits + delta + 1;
if (adjust > 0)
num = mkcl_ash(env, num, adjust);
else if (adjust < 0)
den = mkcl_ash(env, den, -adjust);
}
do {
mkcl_object fraction = mkcl_truncate2(env, num, den);
mkcl_object rem = MKCL_VALUES(1);
mkcl_word len = mkcl_integer_length(env, fraction);
if ((len - digits) == 1) {
if (mkcl_oddp(env, fraction)) {
mkcl_object one = mkcl_minusp(env, num) ? MKCL_MAKE_FIXNUM(-1) : MKCL_MAKE_FIXNUM(1);
if (rem == MKCL_MAKE_FIXNUM(0)) {
if (!mkcl_Null(mk_cl_logbitp(env, MKCL_MAKE_FIXNUM(1), fraction)))
fraction = mkcl_plus(env, fraction, one);
} else {
fraction = mkcl_plus(env, fraction, one);
}
}
*scaleout = scale - (digits + 1);
return fraction;
}
num = mkcl_ash(env, num, -1);
scale++;
--delta;
} while (1);
}
static float
ratio_to_float(MKCL, mkcl_object num, mkcl_object den)
{
mkcl_word scale;
mkcl_object bits = prepare_ratio_to_float(env, num, den, FLT_MANT_DIG, &scale);
#if (MKCL_WORD_BITS - MKCL_TAG_BITS) >= FLT_MANT_DIG
/* In this case the output of prepare_ratio_to_float() always fits within a fixnum. */
float output = mkcl_fixnum_to_word(bits);
#else
float output = MKCL_FIXNUMP(bits) ? mkcl_fixnum_to_word(bits) : _mkcl_big_to_truncated_double(bits);
#endif
return ldexpf(output, scale);
}
static double
ratio_to_double(MKCL, mkcl_object num, mkcl_object den)
{
mkcl_word scale;
mkcl_object bits = prepare_ratio_to_float(env, num, den, DBL_MANT_DIG, &scale);
#if (MKCL_WORD_BITS - MKCL_TAG_BITS) >= DBL_MANT_DIG
/* In this case the output of prepare_ratio_to_float() always fits within a fixnum. */
double output = mkcl_fixnum_to_word(bits);
#else
double output = MKCL_FIXNUMP(bits) ? mkcl_fixnum_to_word(bits) : _mkcl_big_to_truncated_double(bits);
#endif
return ldexp(output, scale);
}
float
mkcl_to_float(MKCL, mkcl_object x)
{
switch(mkcl_type_of(x)) {
case mkcl_t_fixnum:
return((float)(mkcl_fixnum_to_word(x)));
case mkcl_t_bignum:
return ratio_to_double(env, x, MKCL_MAKE_FIXNUM(1));
case mkcl_t_ratio:
return ratio_to_float(env, x->ratio.num, x->ratio.den);
case mkcl_t_singlefloat:
return mkcl_single_float(x);
case mkcl_t_doublefloat:
return((float) mkcl_double_float(x));
#ifdef MKCL_LONG_FLOAT
case mkcl_t_longfloat:
return (float) mkcl_long_float(x);
#endif
default:
mkcl_FEtype_error_real(env, x);
}
}
double
mkcl_to_double(MKCL, mkcl_object x)
{
switch(mkcl_type_of(x)) {
case mkcl_t_fixnum:
return((double)(mkcl_fixnum_to_word(x)));
case mkcl_t_bignum:
return ratio_to_double(env, x, MKCL_MAKE_FIXNUM(1));
case mkcl_t_ratio:
return ratio_to_double(env, x->ratio.num, x->ratio.den);
case mkcl_t_singlefloat:
return (double)mkcl_single_float(x);
case mkcl_t_doublefloat:
return(mkcl_double_float(x));
#ifdef MKCL_LONG_FLOAT
case mkcl_t_longfloat:
return (double)mkcl_long_float(x);
#endif
default:
mkcl_FEtype_error_real(env, x);
}
}
#ifdef MKCL_LONG_FLOAT
static long double
ratio_to_long_double(MKCL, mkcl_object num, mkcl_object den)
{
mkcl_word scale;
mkcl_object bits = prepare_ratio_to_float(env, num, den, LDBL_MANT_DIG, &scale);
long double output = mkcl_to_long_double(env, bits);
return ldexpl(output, scale);
}
#endif
#ifdef MKCL_LONG_FLOAT
long double
mkcl_to_long_double(MKCL, mkcl_object x)
{
switch(mkcl_type_of(x)) {
case mkcl_t_fixnum:
return (long double)mkcl_fixnum_to_word(x);
case mkcl_t_bignum: {
long double output = 0;
int i, l = mpz_size(x->big.big_num);
for (i = 0; i < l; i++) {
output += mpz_getlimbn(x->big.big_num, i);
output = ldexpl(output, -GMP_LIMB_BITS);
}
output = ldexpl(output, l * GMP_LIMB_BITS);
return (mpz_sgn(x->big.big_num) < 0) ? -output : output;
}
case mkcl_t_ratio:
return ratio_to_long_double(env, x->ratio.num, x->ratio.den);
case mkcl_t_singlefloat:
return (long double)mkcl_single_float(x);
case mkcl_t_doublefloat:
return (long double)mkcl_double_float(x);
case mkcl_t_longfloat:
return mkcl_long_float(x);
default:
mkcl_FEtype_error_real(env, x);
}
}
#endif
mkcl_object
mk_cl_rational(MKCL, mkcl_object x)
{
double d;
mkcl_call_stack_check(env);
AGAIN:
switch (mkcl_type_of(x)) {
case mkcl_t_fixnum:
case mkcl_t_bignum:
case mkcl_t_ratio:
break;
case mkcl_t_singlefloat:
d = mkcl_single_float(x);
goto GO_ON;
case mkcl_t_doublefloat:
d = mkcl_double_float(x);
GO_ON:
if (d == 0) {
x = MKCL_MAKE_FIXNUM(0);
} else {
int e;
d = frexp(d, &e);
e -= DBL_MANT_DIG;
x = mkcl_double_to_integer(env, ldexp(d, DBL_MANT_DIG));
if (e != 0) {
x = mkcl_times(env,
mk_cl_expt(env, MKCL_MAKE_FIXNUM(FLT_RADIX), MKCL_MAKE_FIXNUM(e)),
x);
}
}
break;
#ifdef MKCL_LONG_FLOAT
case mkcl_t_longfloat: {
long double d = mkcl_long_float(x);
if (d == 0) {
x = MKCL_MAKE_FIXNUM(0);
} else {
int e;
d = frexpl(d, &e);
e -= LDBL_MANT_DIG;
d = ldexpl(d, LDBL_MANT_DIG);
x = mkcl_long_double_to_integer(env, d);
if (e != 0) {
x = mkcl_times(env,
mk_cl_expt(env, MKCL_MAKE_FIXNUM(FLT_RADIX), MKCL_MAKE_FIXNUM(e)),
x);
}
}
break;
}
#endif
default:
x = mkcl_type_error(env, @'rational',"argument",x,@'number');
goto AGAIN;
}
@(return x);
}
mkcl_object
mkcl_long_double_to_integer(MKCL, long double d0)
{
const int fb = MKCL_WORD_BITS - 3;
int e;
long double d = frexpl(d0, &e);
if (e <= fb) {
return MKCL_MAKE_FIXNUM((mkcl_word)d0);
} else if (e > LDBL_MANT_DIG) {
return mkcl_ash(env, mkcl_long_double_to_integer(env, ldexp(d, LDBL_MANT_DIG)),
e - LDBL_MANT_DIG);
} else {
long double d1 = floorl(d = ldexpl(d, fb));
int newe = e - fb;
mkcl_object o = mkcl_ash(env, mkcl_long_double_to_integer(env, d1), newe);
long double d2 = ldexpl(d - d1, newe);
if (d2) o = mkcl_plus(env, o, mkcl_long_double_to_integer(env, d2));
return o;
}
}
mkcl_object
mkcl_double_to_integer(MKCL, double d)
{
if (d <= MKCL_MOST_POSITIVE_FIXNUM && d >= MKCL_MOST_NEGATIVE_FIXNUM)
return MKCL_MAKE_FIXNUM((mkcl_word)d);
else {
mkcl_object z = _mkcl_big_register0();
_mkcl_big_set_d(z, d);
return _mkcl_big_register_copy(env, z);
}
}
mkcl_object
mkcl_float_to_integer(MKCL, float d)
{
if (d <= MKCL_MOST_POSITIVE_FIXNUM && d >= MKCL_MOST_NEGATIVE_FIXNUM)
return MKCL_MAKE_FIXNUM((mkcl_word)d);
else {
mkcl_object z = _mkcl_big_register0();
_mkcl_big_set_d(z, d);
return _mkcl_big_register_copy(env, z);
}
}
void
mkcl_init_number(MKCL)
{
mkcl_object num;
num = mkcl_make_singlefloat(env, FLT_MAX);
MKCL_SET(@'MOST-POSITIVE-SHORT-FLOAT', num);
MKCL_SET(@'MOST-POSITIVE-SINGLE-FLOAT', num);
num = mkcl_make_singlefloat(env, -FLT_MAX);
MKCL_SET(@'MOST-NEGATIVE-SHORT-FLOAT', num);
MKCL_SET(@'MOST-NEGATIVE-SINGLE-FLOAT', num);
num = mkcl_make_singlefloat(env, FLT_MIN);
MKCL_SET(@'LEAST-POSITIVE-SHORT-FLOAT', num);
MKCL_SET(@'LEAST-POSITIVE-SINGLE-FLOAT', num);
MKCL_SET(@'LEAST-POSITIVE-NORMALIZED-SHORT-FLOAT', num);
MKCL_SET(@'LEAST-POSITIVE-NORMALIZED-SINGLE-FLOAT', num);
num = mkcl_make_singlefloat(env, -FLT_MIN);
MKCL_SET(@'LEAST-NEGATIVE-SHORT-FLOAT', num);
MKCL_SET(@'LEAST-NEGATIVE-SINGLE-FLOAT', num);
MKCL_SET(@'LEAST-NEGATIVE-NORMALIZED-SHORT-FLOAT', num);
MKCL_SET(@'LEAST-NEGATIVE-NORMALIZED-SINGLE-FLOAT', num);
num = mkcl_make_doublefloat(env, DBL_MAX);
MKCL_SET(@'MOST-POSITIVE-DOUBLE-FLOAT', num);
#ifdef MKCL_LONG_FLOAT
num = mkcl_make_longfloat(env, LDBL_MAX);
#endif
MKCL_SET(@'MOST-POSITIVE-LONG-FLOAT', num);
num = mkcl_make_doublefloat(env, -DBL_MAX);
MKCL_SET(@'MOST-NEGATIVE-DOUBLE-FLOAT', num);
#ifdef MKCL_LONG_FLOAT
num = mkcl_make_longfloat(env, -LDBL_MAX);
#endif
MKCL_SET(@'MOST-NEGATIVE-LONG-FLOAT', num);
num = mkcl_make_doublefloat(env, DBL_MIN);
MKCL_SET(@'LEAST-POSITIVE-DOUBLE-FLOAT', num);
MKCL_SET(@'LEAST-POSITIVE-NORMALIZED-DOUBLE-FLOAT', num);
#ifdef MKCL_LONG_FLOAT
num = mkcl_make_longfloat(env, LDBL_MIN);
#endif
MKCL_SET(@'LEAST-POSITIVE-LONG-FLOAT', num);
MKCL_SET(@'LEAST-POSITIVE-NORMALIZED-LONG-FLOAT', num);
num = mkcl_make_doublefloat(env, -DBL_MIN);
MKCL_SET(@'LEAST-NEGATIVE-DOUBLE-FLOAT', num);
MKCL_SET(@'LEAST-NEGATIVE-NORMALIZED-DOUBLE-FLOAT', num);
#ifdef MKCL_LONG_FLOAT
num = mkcl_make_longfloat(env, -LDBL_MIN);
#endif
MKCL_SET(@'LEAST-NEGATIVE-LONG-FLOAT', num);
MKCL_SET(@'LEAST-NEGATIVE-NORMALIZED-LONG-FLOAT', num);
mkcl_core.singlefloat_zero = mkcl_alloc_raw_singlefloat(env);
mkcl_single_float(mkcl_core.singlefloat_zero) = 0.0F;
mkcl_core.doublefloat_zero = mkcl_alloc_raw_doublefloat(env);
mkcl_double_float(mkcl_core.doublefloat_zero) = 0.0;
#ifdef MKCL_LONG_FLOAT
mkcl_core.longfloat_zero = mkcl_alloc_raw_longfloat(env);
mkcl_core.longfloat_zero->longfloat.value = 0.0L;
#else
mkcl_core.longfloat_zero = mkcl_core.doublefloat_zero;
#endif
#ifdef MKCL_SIGNED_ZERO
mkcl_core.singlefloat_minus_zero = mkcl_alloc_raw_singlefloat(env);
mkcl_single_float(mkcl_core.singlefloat_minus_zero) = -0.0F;
mkcl_core.doublefloat_minus_zero = mkcl_alloc_raw_doublefloat(env);
mkcl_double_float(mkcl_core.doublefloat_minus_zero) = -0.0;
# ifdef MKCL_LONG_FLOAT
mkcl_core.longfloat_minus_zero = mkcl_alloc_raw_longfloat(env);
mkcl_core.longfloat_minus_zero->longfloat.value = -0.0L;
# else
mkcl_core.longfloat_minus_zero = mkcl_core.doublefloat_minus_zero;
# endif
#else
mkcl_core.singlefloat_minus_zero = mkcl_core.singlefloat_zero;
mkcl_core.doublefloat_minus_zero = mkcl_core.doublefloat_zero;
mkcl_core.longfloat_minus_zero = mkcl_core.longfloat_zero;
#endif
mkcl_core.plus_half = mkcl_make_ratio(env, MKCL_MAKE_FIXNUM(1), MKCL_MAKE_FIXNUM(2));
mkcl_core.minus_half = mkcl_make_ratio(env, MKCL_MAKE_FIXNUM(-1), MKCL_MAKE_FIXNUM(2));
mkcl_core.imag_unit =
mkcl_make_complex(env, mkcl_make_singlefloat(env, 0.0),
mkcl_make_singlefloat(env, 1.0));
mkcl_core.minus_imag_unit =
mkcl_make_complex(env, mkcl_make_singlefloat(env, 0.0),
mkcl_make_singlefloat(env, -1.0));
mkcl_core.imag_two =
mkcl_make_complex(env, mkcl_make_singlefloat(env, 0.0),
mkcl_make_singlefloat(env, 2.0));
#ifdef MKCL_LONG_FLOAT
MKCL_SET(@'pi', mkcl_make_longfloat(env, MKCL_PI_L));
#else
MKCL_SET(@'pi', mkcl_make_doublefloat(env, MKCL_PI_D));
#endif
MKCL_SET(@'*random-state*', mkcl_make_random_state(env, mk_cl_Ct));
}