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
|
------------------------------------------------------------------------------
-- --
-- GNAT RUN-TIME COMPONENTS --
-- --
-- S Y S T E M . I M G _ R E A L --
-- --
-- B o d y --
-- --
-- Copyright (C) 1992-2009, Free Software Foundation, Inc. --
-- --
-- GNAT is free software; you can redistribute it and/or modify it under --
-- terms of the GNU General Public License as published by the Free Soft- --
-- ware Foundation; either version 3, or (at your option) any later ver- --
-- sion. GNAT is distributed in the hope that it will be useful, but WITH- --
-- OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY --
-- or FITNESS FOR A PARTICULAR PURPOSE. --
-- --
-- As a special exception under Section 7 of GPL version 3, you are granted --
-- additional permissions described in the GCC Runtime Library Exception, --
-- version 3.1, as published by the Free Software Foundation. --
-- --
-- You should have received a copy of the GNU General Public License and --
-- a copy of the GCC Runtime Library Exception along with this program; --
-- see the files COPYING3 and COPYING.RUNTIME respectively. If not, see --
-- <http://www.gnu.org/licenses/>. --
-- --
-- GNAT was originally developed by the GNAT team at New York University. --
-- Extensive contributions were provided by Ada Core Technologies Inc. --
-- --
------------------------------------------------------------------------------
with System.Img_LLU; use System.Img_LLU;
with System.Img_Uns; use System.Img_Uns;
with System.Powten_Table; use System.Powten_Table;
with System.Unsigned_Types; use System.Unsigned_Types;
package body System.Img_Real is
-- The following defines the maximum number of digits that we can convert
-- accurately. This is limited by the precision of Long_Long_Float, and
-- also by the number of digits we can hold in Long_Long_Unsigned, which
-- is the integer type we use as an intermediate for the result.
-- We assume that in practice, the limitation will come from the digits
-- value, rather than the integer value. This is true for typical IEEE
-- implementations, and at worst, the only loss is for some precision
-- in very high precision floating-point output.
-- Note that in the following, the "-2" accounts for the sign and one
-- extra digits, since we need the maximum number of 9's that can be
-- supported, e.g. for the normal 64 bit case, Long_Long_Integer'Width
-- is 21, since the maximum value (approx 1.6 * 10**19) has 20 digits,
-- but the maximum number of 9's that can be supported is 19.
Maxdigs : constant :=
Natural'Min
(Long_Long_Unsigned'Width - 2, Long_Long_Float'Digits);
Unsdigs : constant := Unsigned'Width - 2;
-- Number of digits that can be converted using type Unsigned
-- See above for the explanation of the -2.
Maxscaling : constant := 5000;
-- Max decimal scaling required during conversion of floating-point
-- numbers to decimal. This is used to defend against infinite
-- looping in the conversion, as can be caused by erroneous executions.
-- The largest exponent used on any current system is 2**16383, which
-- is approximately 10**4932, and the highest number of decimal digits
-- is about 35 for 128-bit floating-point formats, so 5000 leaves
-- enough room for scaling such values
function Is_Negative (V : Long_Long_Float) return Boolean;
pragma Import (Intrinsic, Is_Negative);
--------------------------
-- Image_Floating_Point --
--------------------------
procedure Image_Floating_Point
(V : Long_Long_Float;
S : in out String;
P : out Natural;
Digs : Natural)
is
pragma Assert (S'First = 1);
begin
-- Decide whether a blank should be prepended before the call to
-- Set_Image_Real. We generate a blank for positive values, and
-- also for positive zeroes. For negative zeroes, we generate a
-- space only if Signed_Zeroes is True (the RM only permits the
-- output of -0.0 on targets where this is the case). We can of
-- course still see a -0.0 on a target where Signed_Zeroes is
-- False (since this attribute refers to the proper handling of
-- negative zeroes, not to their existence).
if not Is_Negative (V)
or else (not Long_Long_Float'Signed_Zeros and then V = -0.0)
then
S (1) := ' ';
P := 1;
else
P := 0;
end if;
Set_Image_Real (V, S, P, 1, Digs - 1, 3);
end Image_Floating_Point;
--------------------------------
-- Image_Ordinary_Fixed_Point --
--------------------------------
procedure Image_Ordinary_Fixed_Point
(V : Long_Long_Float;
S : in out String;
P : out Natural;
Aft : Natural)
is
pragma Assert (S'First = 1);
begin
-- Output space at start if non-negative
if V >= 0.0 then
S (1) := ' ';
P := 1;
else
P := 0;
end if;
Set_Image_Real (V, S, P, 1, Aft, 0);
end Image_Ordinary_Fixed_Point;
--------------------
-- Set_Image_Real --
--------------------
procedure Set_Image_Real
(V : Long_Long_Float;
S : out String;
P : in out Natural;
Fore : Natural;
Aft : Natural;
Exp : Natural)
is
procedure Reset;
pragma Import (C, Reset, "__gnat_init_float");
-- We import the floating-point processor reset routine so that we can
-- be sure the floating-point processor is properly set for conversion
-- calls (see description of Reset in GNAT.Float_Control (g-flocon.ads).
-- This is notably need on Windows, where calls to the operating system
-- randomly reset the processor into 64-bit mode.
NFrac : constant Natural := Natural'Max (Aft, 1);
Sign : Character;
X : aliased Long_Long_Float;
-- This is declared aliased because the expansion of X'Valid passes
-- X by access and JGNAT requires all access parameters to be aliased.
-- The Valid attribute probably needs to be handled via a different
-- expansion for JGNAT, and this use of aliased should be removed
-- once Valid is handled properly. ???
Scale : Integer;
Expon : Integer;
Field_Max : constant := 255;
-- This should be the same value as Ada.[Wide_]Text_IO.Field'Last.
-- It is not worth dragging in Ada.Text_IO to pick up this value,
-- since it really should never be necessary to change it!
Digs : String (1 .. 2 * Field_Max + 16);
-- Array used to hold digits of converted integer value. This is a
-- large enough buffer to accommodate ludicrous values of Fore and Aft.
Ndigs : Natural;
-- Number of digits stored in Digs (and also subscript of last digit)
procedure Adjust_Scale (S : Natural);
-- Adjusts the value in X by multiplying or dividing by a power of
-- ten so that it is in the range 10**(S-1) <= X < 10**S. Includes
-- adding 0.5 to round the result, readjusting if the rounding causes
-- the result to wander out of the range. Scale is adjusted to reflect
-- the power of ten used to divide the result (i.e. one is added to
-- the scale value for each division by 10.0, or one is subtracted
-- for each multiplication by 10.0).
procedure Convert_Integer;
-- Takes the value in X, outputs integer digits into Digs. On return,
-- Ndigs is set to the number of digits stored. The digits are stored
-- in Digs (1 .. Ndigs),
procedure Set (C : Character);
-- Sets character C in output buffer
procedure Set_Blanks_And_Sign (N : Integer);
-- Sets leading blanks and minus sign if needed. N is the number of
-- positions to be filled (a minus sign is output even if N is zero
-- or negative, but for a positive value, if N is non-positive, then
-- the call has no effect).
procedure Set_Digs (S, E : Natural);
-- Set digits S through E from Digs buffer. No effect if S > E
procedure Set_Special_Fill (N : Natural);
-- After outputting +Inf, -Inf or NaN, this routine fills out the
-- rest of the field with * characters. The argument is the number
-- of characters output so far (either 3 or 4)
procedure Set_Zeros (N : Integer);
-- Set N zeros, no effect if N is negative
pragma Inline (Set);
pragma Inline (Set_Digs);
pragma Inline (Set_Zeros);
------------------
-- Adjust_Scale --
------------------
procedure Adjust_Scale (S : Natural) is
Lo : Natural;
Hi : Natural;
Mid : Natural;
XP : Long_Long_Float;
begin
-- Cases where scaling up is required
if X < Powten (S - 1) then
-- What we are looking for is a power of ten to multiply X by
-- so that the result lies within the required range.
loop
XP := X * Powten (Maxpow);
exit when XP >= Powten (S - 1) or else Scale < -Maxscaling;
X := XP;
Scale := Scale - Maxpow;
end loop;
-- The following exception is only raised in case of erroneous
-- execution, where a number was considered valid but still
-- fails to scale up. One situation where this can happen is
-- when a system which is supposed to be IEEE-compliant, but
-- has been reconfigured to flush denormals to zero.
if Scale < -Maxscaling then
raise Constraint_Error;
end if;
-- Here we know that we must multiply by at least 10**1 and that
-- 10**Maxpow takes us too far: binary search to find right one.
-- Because of roundoff errors, it is possible for the value
-- of XP to be just outside of the interval when Lo >= Hi. In
-- that case we adjust explicitly by a factor of 10. This
-- can only happen with a value that is very close to an
-- exact power of 10.
Lo := 1;
Hi := Maxpow;
loop
Mid := (Lo + Hi) / 2;
XP := X * Powten (Mid);
if XP < Powten (S - 1) then
if Lo >= Hi then
Mid := Mid + 1;
XP := XP * 10.0;
exit;
else
Lo := Mid + 1;
end if;
elsif XP >= Powten (S) then
if Lo >= Hi then
Mid := Mid - 1;
XP := XP / 10.0;
exit;
else
Hi := Mid - 1;
end if;
else
exit;
end if;
end loop;
X := XP;
Scale := Scale - Mid;
-- Cases where scaling down is required
elsif X >= Powten (S) then
-- What we are looking for is a power of ten to divide X by
-- so that the result lies within the required range.
loop
XP := X / Powten (Maxpow);
exit when XP < Powten (S) or else Scale > Maxscaling;
X := XP;
Scale := Scale + Maxpow;
end loop;
-- The following exception is only raised in case of erroneous
-- execution, where a number was considered valid but still
-- fails to scale up. One situation where this can happen is
-- when a system which is supposed to be IEEE-compliant, but
-- has been reconfigured to flush denormals to zero.
if Scale > Maxscaling then
raise Constraint_Error;
end if;
-- Here we know that we must divide by at least 10**1 and that
-- 10**Maxpow takes us too far, binary search to find right one.
Lo := 1;
Hi := Maxpow;
loop
Mid := (Lo + Hi) / 2;
XP := X / Powten (Mid);
if XP < Powten (S - 1) then
if Lo >= Hi then
XP := XP * 10.0;
Mid := Mid - 1;
exit;
else
Hi := Mid - 1;
end if;
elsif XP >= Powten (S) then
if Lo >= Hi then
XP := XP / 10.0;
Mid := Mid + 1;
exit;
else
Lo := Mid + 1;
end if;
else
exit;
end if;
end loop;
X := XP;
Scale := Scale + Mid;
-- Here we are already scaled right
else
null;
end if;
-- Round, readjusting scale if needed. Note that if a readjustment
-- occurs, then it is never necessary to round again, because there
-- is no possibility of such a second rounding causing a change.
X := X + 0.5;
if X >= Powten (S) then
X := X / 10.0;
Scale := Scale + 1;
end if;
end Adjust_Scale;
---------------------
-- Convert_Integer --
---------------------
procedure Convert_Integer is
begin
-- Use Unsigned routine if possible, since on many machines it will
-- be significantly more efficient than the Long_Long_Unsigned one.
if X < Powten (Unsdigs) then
Ndigs := 0;
Set_Image_Unsigned
(Unsigned (Long_Long_Float'Truncation (X)),
Digs, Ndigs);
-- But if we want more digits than fit in Unsigned, we have to use
-- the Long_Long_Unsigned routine after all.
else
Ndigs := 0;
Set_Image_Long_Long_Unsigned
(Long_Long_Unsigned (Long_Long_Float'Truncation (X)),
Digs, Ndigs);
end if;
end Convert_Integer;
---------
-- Set --
---------
procedure Set (C : Character) is
begin
P := P + 1;
S (P) := C;
end Set;
-------------------------
-- Set_Blanks_And_Sign --
-------------------------
procedure Set_Blanks_And_Sign (N : Integer) is
begin
if Sign = '-' then
for J in 1 .. N - 1 loop
Set (' ');
end loop;
Set ('-');
else
for J in 1 .. N loop
Set (' ');
end loop;
end if;
end Set_Blanks_And_Sign;
--------------
-- Set_Digs --
--------------
procedure Set_Digs (S, E : Natural) is
begin
for J in S .. E loop
Set (Digs (J));
end loop;
end Set_Digs;
----------------------
-- Set_Special_Fill --
----------------------
procedure Set_Special_Fill (N : Natural) is
F : Natural;
begin
F := Fore + 1 + Aft - N;
if Exp /= 0 then
F := F + Exp + 1;
end if;
for J in 1 .. F loop
Set ('*');
end loop;
end Set_Special_Fill;
---------------
-- Set_Zeros --
---------------
procedure Set_Zeros (N : Integer) is
begin
for J in 1 .. N loop
Set ('0');
end loop;
end Set_Zeros;
-- Start of processing for Set_Image_Real
begin
Reset;
Scale := 0;
-- Deal with invalid values first,
if not V'Valid then
-- Note that we're taking our chances here, as V might be
-- an invalid bit pattern resulting from erroneous execution
-- (caused by using uninitialized variables for example).
-- No matter what, we'll at least get reasonable behaviour,
-- converting to infinity or some other value, or causing an
-- exception to be raised is fine.
-- If the following test succeeds, then we definitely have
-- an infinite value, so we print Inf.
if V > Long_Long_Float'Last then
Set ('+');
Set ('I');
Set ('n');
Set ('f');
Set_Special_Fill (4);
-- In all other cases we print NaN
elsif V < Long_Long_Float'First then
Set ('-');
Set ('I');
Set ('n');
Set ('f');
Set_Special_Fill (4);
else
Set ('N');
Set ('a');
Set ('N');
Set_Special_Fill (3);
end if;
return;
end if;
-- Positive values
if V > 0.0 then
X := V;
Sign := '+';
-- Negative values
elsif V < 0.0 then
X := -V;
Sign := '-';
-- Zero values
elsif V = 0.0 then
if Long_Long_Float'Signed_Zeros and then Is_Negative (V) then
Sign := '-';
else
Sign := '+';
end if;
Set_Blanks_And_Sign (Fore - 1);
Set ('0');
Set ('.');
Set_Zeros (NFrac);
if Exp /= 0 then
Set ('E');
Set ('+');
Set_Zeros (Natural'Max (1, Exp - 1));
end if;
return;
else
-- It should not be possible for a NaN to end up here.
-- Either the 'Valid test has failed, or we have some form
-- of erroneous execution. Raise Constraint_Error instead of
-- attempting to go ahead printing the value.
raise Constraint_Error;
end if;
-- X and Sign are set here, and X is known to be a valid,
-- non-zero floating-point number.
-- Case of non-zero value with Exp = 0
if Exp = 0 then
-- First step is to multiply by 10 ** Nfrac to get an integer
-- value to be output, an then add 0.5 to round the result.
declare
NF : Natural := NFrac;
begin
loop
-- If we are larger than Powten (Maxdigs) now, then
-- we have too many significant digits, and we have
-- not even finished multiplying by NFrac (NF shows
-- the number of unaccounted-for digits).
if X >= Powten (Maxdigs) then
-- In this situation, we only to generate a reasonable
-- number of significant digits, and then zeroes after.
-- So first we rescale to get:
-- 10 ** (Maxdigs - 1) <= X < 10 ** Maxdigs
-- and then convert the resulting integer
Adjust_Scale (Maxdigs);
Convert_Integer;
-- If that caused rescaling, then add zeros to the end
-- of the number to account for this scaling. Also add
-- zeroes to account for the undone multiplications
for J in 1 .. Scale + NF loop
Ndigs := Ndigs + 1;
Digs (Ndigs) := '0';
end loop;
exit;
-- If multiplication is complete, then convert the resulting
-- integer after rounding (note that X is non-negative)
elsif NF = 0 then
X := X + 0.5;
Convert_Integer;
exit;
-- Otherwise we can go ahead with the multiplication. If it
-- can be done in one step, then do it in one step.
elsif NF < Maxpow then
X := X * Powten (NF);
NF := 0;
-- If it cannot be done in one step, then do partial scaling
else
X := X * Powten (Maxpow);
NF := NF - Maxpow;
end if;
end loop;
end;
-- If number of available digits is less or equal to NFrac,
-- then we need an extra zero before the decimal point.
if Ndigs <= NFrac then
Set_Blanks_And_Sign (Fore - 1);
Set ('0');
Set ('.');
Set_Zeros (NFrac - Ndigs);
Set_Digs (1, Ndigs);
-- Normal case with some digits before the decimal point
else
Set_Blanks_And_Sign (Fore - (Ndigs - NFrac));
Set_Digs (1, Ndigs - NFrac);
Set ('.');
Set_Digs (Ndigs - NFrac + 1, Ndigs);
end if;
-- Case of non-zero value with non-zero Exp value
else
-- If NFrac is less than Maxdigs, then all the fraction digits are
-- significant, so we can scale the resulting integer accordingly.
if NFrac < Maxdigs then
Adjust_Scale (NFrac + 1);
Convert_Integer;
-- Otherwise, we get the maximum number of digits available
else
Adjust_Scale (Maxdigs);
Convert_Integer;
for J in 1 .. NFrac - Maxdigs + 1 loop
Ndigs := Ndigs + 1;
Digs (Ndigs) := '0';
Scale := Scale - 1;
end loop;
end if;
Set_Blanks_And_Sign (Fore - 1);
Set (Digs (1));
Set ('.');
Set_Digs (2, Ndigs);
-- The exponent is the scaling factor adjusted for the digits
-- that we output after the decimal point, since these were
-- included in the scaled digits that we output.
Expon := Scale + NFrac;
Set ('E');
Ndigs := 0;
if Expon >= 0 then
Set ('+');
Set_Image_Unsigned (Unsigned (Expon), Digs, Ndigs);
else
Set ('-');
Set_Image_Unsigned (Unsigned (-Expon), Digs, Ndigs);
end if;
Set_Zeros (Exp - Ndigs - 1);
Set_Digs (1, Ndigs);
end if;
end Set_Image_Real;
end System.Img_Real;
|