Commit | Line | Data |
0716bf9b |
1 | package Math::BigInt::Calc; |
2 | |
3 | use 5.005; |
4 | use strict; |
574bacfe |
5 | # use warnings; # dont use warnings for older Perls |
0716bf9b |
6 | |
7 | require Exporter; |
bd05a461 |
8 | use vars qw/@ISA $VERSION/; |
0716bf9b |
9 | @ISA = qw(Exporter); |
10 | |
dccbb853 |
11 | $VERSION = '0.13'; |
0716bf9b |
12 | |
13 | # Package to store unsigned big integers in decimal and do math with them |
14 | |
15 | # Internally the numbers are stored in an array with at least 1 element, no |
16 | # leading zero parts (except the first) and in base 100000 |
17 | |
18 | # todo: |
19 | # - fully remove funky $# stuff (maybe) |
0716bf9b |
20 | |
21 | # USE_MUL: due to problems on certain os (os390, posix-bc) "* 1e-5" is used |
ee15d750 |
22 | # instead of "/ 1e5" at some places, (marked with USE_MUL). Other platforms |
23 | # BS2000, some Crays need USE_DIV instead. |
bd05a461 |
24 | # The BEGIN block is used to determine which of the two variants gives the |
25 | # correct result. |
0716bf9b |
26 | |
27 | ############################################################################## |
28 | # global constants, flags and accessory |
29 | |
30 | # constants for easier life |
31 | my $nan = 'NaN'; |
ee15d750 |
32 | my ($BASE,$RBASE,$BASE_LEN,$MAX_VAL); |
33 | |
34 | sub _base_len |
35 | { |
dccbb853 |
36 | # set/get the BASE_LEN and assorted other, connected values |
37 | # used only be the testsuite, set is used only by the BEGIN block below |
ee15d750 |
38 | my $b = shift; |
39 | if (defined $b) |
40 | { |
dccbb853 |
41 | $b = 8 if $b > 8; # cap, for VMS, OS/390 and other 64 bit |
ee15d750 |
42 | $BASE_LEN = $b; |
43 | $BASE = int("1e".$BASE_LEN); |
44 | $RBASE = abs('1e-'.$BASE_LEN); # see USE_MUL |
45 | $MAX_VAL = $BASE-1; |
46 | # print "BASE_LEN: $BASE_LEN MAX_VAL: $MAX_VAL\n"; |
47 | # print "int: ",int($BASE * $RBASE),"\n"; |
48 | if (int($BASE * $RBASE) == 0) # should be 1 |
49 | { |
50 | # must USE_MUL |
ee15d750 |
51 | *{_mul} = \&_mul_use_mul; |
52 | *{_div} = \&_div_use_mul; |
53 | } |
54 | else |
55 | { |
ee15d750 |
56 | # can USE_DIV instead |
57 | *{_mul} = \&_mul_use_div; |
58 | *{_div} = \&_div_use_div; |
59 | } |
60 | } |
dccbb853 |
61 | $BASE_LEN; |
ee15d750 |
62 | } |
574bacfe |
63 | |
64 | BEGIN |
65 | { |
bd05a461 |
66 | # from Daniel Pfeiffer: determine largest group of digits that is precisely |
574bacfe |
67 | # multipliable with itself plus carry |
dccbb853 |
68 | # Test now changed to expect the proper pattern, not a result off by 1 or 2 |
69 | my ($e, $num) = 3; # lowest value we will use is 3+1-1 = 3 |
bd05a461 |
70 | do |
71 | { |
72 | $num = ('9' x ++$e) + 0; |
574bacfe |
73 | $num *= $num + 1; |
dccbb853 |
74 | # print "$num $e\n"; |
75 | } while ("$num" =~ /9{$e}0{$e}/); # must be a certain pattern |
76 | # last test failed, so retract one step: |
ee15d750 |
77 | _base_len($e-1); |
574bacfe |
78 | } |
79 | |
0716bf9b |
80 | ############################################################################## |
81 | # create objects from various representations |
82 | |
83 | sub _new |
84 | { |
85 | # (string) return ref to num_array |
86 | # Convert a number from string format to internal base 100000 format. |
87 | # Assumes normalized value as input. |
574bacfe |
88 | my $d = $_[1]; |
0716bf9b |
89 | # print "_new $d $$d\n"; |
90 | my $il = CORE::length($$d)-1; |
91 | # these leaves '00000' instead of int 0 and will be corrected after any op |
574bacfe |
92 | return [ reverse(unpack("a" . ($il % $BASE_LEN+1) |
93 | . ("a$BASE_LEN" x ($il / $BASE_LEN)), $$d)) ]; |
0716bf9b |
94 | } |
95 | |
96 | sub _zero |
97 | { |
98 | # create a zero |
99 | return [ 0 ]; |
100 | } |
101 | |
102 | sub _one |
103 | { |
104 | # create a one |
105 | return [ 1 ]; |
106 | } |
107 | |
108 | sub _copy |
109 | { |
574bacfe |
110 | return [ @{$_[1]} ]; |
0716bf9b |
111 | } |
112 | |
bd05a461 |
113 | # catch and throw away |
114 | sub import { } |
115 | |
0716bf9b |
116 | ############################################################################## |
117 | # convert back to string and number |
118 | |
119 | sub _str |
120 | { |
121 | # (ref to BINT) return num_str |
122 | # Convert number from internal base 100000 format to string format. |
123 | # internal format is always normalized (no leading zeros, "-0" => "+0") |
574bacfe |
124 | my $ar = $_[1]; |
0716bf9b |
125 | my $ret = ""; |
126 | my $l = scalar @$ar; # number of parts |
127 | return $nan if $l < 1; # should not happen |
128 | # handle first one different to strip leading zeros from it (there are no |
129 | # leading zero parts in internal representation) |
130 | $l --; $ret .= $ar->[$l]; $l--; |
131 | # Interestingly, the pre-padd method uses more time |
574bacfe |
132 | # the old grep variant takes longer (14 to 10 sec) |
133 | my $z = '0' x ($BASE_LEN-1); |
0716bf9b |
134 | while ($l >= 0) |
135 | { |
574bacfe |
136 | $ret .= substr($z.$ar->[$l],-$BASE_LEN); # fastest way I could think of |
0716bf9b |
137 | $l--; |
138 | } |
139 | return \$ret; |
140 | } |
141 | |
142 | sub _num |
143 | { |
144 | # Make a number (scalar int/float) from a BigInt object |
574bacfe |
145 | my $x = $_[1]; |
0716bf9b |
146 | return $x->[0] if scalar @$x == 1; # below $BASE |
147 | my $fac = 1; |
148 | my $num = 0; |
149 | foreach (@$x) |
150 | { |
151 | $num += $fac*$_; $fac *= $BASE; |
152 | } |
153 | return $num; |
154 | } |
155 | |
156 | ############################################################################## |
157 | # actual math code |
158 | |
159 | sub _add |
160 | { |
161 | # (ref to int_num_array, ref to int_num_array) |
574bacfe |
162 | # routine to add two base 1eX numbers |
0716bf9b |
163 | # stolen from Knuth Vol 2 Algorithm A pg 231 |
b22b3e31 |
164 | # there are separate routines to add and sub as per Knuth pg 233 |
0716bf9b |
165 | # This routine clobbers up array x, but not y. |
166 | |
574bacfe |
167 | my ($c,$x,$y) = @_; |
0716bf9b |
168 | |
169 | # for each in Y, add Y to X and carry. If after that, something is left in |
170 | # X, foreach in X add carry to X and then return X, carry |
171 | # Trades one "$j++" for having to shift arrays, $j could be made integer |
b22b3e31 |
172 | # but this would impose a limit to number-length of 2**32. |
0716bf9b |
173 | my $i; my $car = 0; my $j = 0; |
174 | for $i (@$y) |
175 | { |
176 | $x->[$j] -= $BASE |
177 | if $car = (($x->[$j] += $i + $car) >= $BASE) ? 1 : 0; |
178 | $j++; |
179 | } |
180 | while ($car != 0) |
181 | { |
182 | $x->[$j] -= $BASE if $car = (($x->[$j] += $car) >= $BASE) ? 1 : 0; $j++; |
183 | } |
184 | return $x; |
185 | } |
186 | |
187 | sub _sub |
188 | { |
189 | # (ref to int_num_array, ref to int_num_array) |
574bacfe |
190 | # subtract base 1eX numbers -- stolen from Knuth Vol 2 pg 232, $x > $y |
b22b3e31 |
191 | # subtract Y from X (X is always greater/equal!) by modifying x in place |
574bacfe |
192 | my ($c,$sx,$sy,$s) = @_; |
0716bf9b |
193 | |
194 | my $car = 0; my $i; my $j = 0; |
195 | if (!$s) |
196 | { |
197 | #print "case 2\n"; |
198 | for $i (@$sx) |
199 | { |
200 | last unless defined $sy->[$j] || $car; |
201 | #print "x: $i y: $sy->[$j] c: $car\n"; |
202 | $i += $BASE if $car = (($i -= ($sy->[$j] || 0) + $car) < 0); $j++; |
203 | #print "x: $i y: $sy->[$j-1] c: $car\n"; |
204 | } |
205 | # might leave leading zeros, so fix that |
206 | __strip_zeros($sx); |
207 | return $sx; |
208 | } |
209 | else |
210 | { |
211 | #print "case 1 (swap)\n"; |
212 | for $i (@$sx) |
213 | { |
214 | last unless defined $sy->[$j] || $car; |
215 | #print "$sy->[$j] $i $car => $sx->[$j]\n"; |
216 | $sy->[$j] += $BASE |
217 | if $car = (($sy->[$j] = $i-($sy->[$j]||0) - $car) < 0); |
218 | #print "$sy->[$j] $i $car => $sy->[$j]\n"; |
219 | $j++; |
220 | } |
221 | # might leave leading zeros, so fix that |
222 | __strip_zeros($sy); |
223 | return $sy; |
224 | } |
225 | } |
226 | |
ee15d750 |
227 | sub _mul_use_mul |
0716bf9b |
228 | { |
229 | # (BINT, BINT) return nothing |
230 | # multiply two numbers in internal representation |
b22b3e31 |
231 | # modifies first arg, second need not be different from first |
574bacfe |
232 | my ($c,$xv,$yv) = @_; |
dccbb853 |
233 | |
0716bf9b |
234 | my @prod = (); my ($prod,$car,$cty,$xi,$yi); |
235 | # since multiplying $x with $x fails, make copy in this case |
574bacfe |
236 | $yv = [@$xv] if "$xv" eq "$yv"; # same references? |
0716bf9b |
237 | for $xi (@$xv) |
238 | { |
239 | $car = 0; $cty = 0; |
574bacfe |
240 | |
241 | # slow variant |
242 | # for $yi (@$yv) |
243 | # { |
244 | # $prod = $xi * $yi + ($prod[$cty] || 0) + $car; |
245 | # $prod[$cty++] = |
246 | # $prod - ($car = int($prod * RBASE)) * $BASE; # see USE_MUL |
247 | # } |
248 | # $prod[$cty] += $car if $car; # need really to check for 0? |
249 | # $xi = shift @prod; |
250 | |
251 | # faster variant |
252 | # looping through this if $xi == 0 is silly - so optimize it away! |
253 | $xi = (shift @prod || 0), next if $xi == 0; |
0716bf9b |
254 | for $yi (@$yv) |
255 | { |
256 | $prod = $xi * $yi + ($prod[$cty] || 0) + $car; |
574bacfe |
257 | ## this is actually a tad slower |
258 | ## $prod = $prod[$cty]; $prod += ($car + $xi * $yi); # no ||0 here |
0716bf9b |
259 | $prod[$cty++] = |
574bacfe |
260 | $prod - ($car = int($prod * $RBASE)) * $BASE; # see USE_MUL |
0716bf9b |
261 | } |
262 | $prod[$cty] += $car if $car; # need really to check for 0? |
263 | $xi = shift @prod; |
264 | } |
0716bf9b |
265 | push @$xv, @prod; |
266 | __strip_zeros($xv); |
267 | # normalize (handled last to save check for $y->is_zero() |
268 | return $xv; |
269 | } |
270 | |
ee15d750 |
271 | sub _mul_use_div |
272 | { |
273 | # (BINT, BINT) return nothing |
274 | # multiply two numbers in internal representation |
275 | # modifies first arg, second need not be different from first |
276 | my ($c,$xv,$yv) = @_; |
277 | |
278 | my @prod = (); my ($prod,$car,$cty,$xi,$yi); |
279 | # since multiplying $x with $x fails, make copy in this case |
280 | $yv = [@$xv] if "$xv" eq "$yv"; # same references? |
281 | for $xi (@$xv) |
282 | { |
283 | $car = 0; $cty = 0; |
284 | # looping through this if $xi == 0 is silly - so optimize it away! |
285 | $xi = (shift @prod || 0), next if $xi == 0; |
286 | for $yi (@$yv) |
287 | { |
288 | $prod = $xi * $yi + ($prod[$cty] || 0) + $car; |
289 | $prod[$cty++] = |
290 | $prod - ($car = int($prod / $BASE)) * $BASE; |
291 | } |
292 | $prod[$cty] += $car if $car; # need really to check for 0? |
293 | $xi = shift @prod; |
294 | } |
295 | push @$xv, @prod; |
296 | __strip_zeros($xv); |
297 | # normalize (handled last to save check for $y->is_zero() |
298 | return $xv; |
299 | } |
300 | |
301 | sub _div_use_mul |
0716bf9b |
302 | { |
b22b3e31 |
303 | # ref to array, ref to array, modify first array and return remainder if |
0716bf9b |
304 | # in list context |
b22b3e31 |
305 | # no longer handles sign |
574bacfe |
306 | my ($c,$x,$yorg) = @_; |
0716bf9b |
307 | my ($car,$bar,$prd,$dd,$xi,$yi,@q,$v2,$v1); |
308 | |
309 | my (@d,$tmp,$q,$u2,$u1,$u0); |
310 | |
311 | $car = $bar = $prd = 0; |
312 | |
313 | my $y = [ @$yorg ]; |
314 | if (($dd = int($BASE/($y->[-1]+1))) != 1) |
315 | { |
316 | for $xi (@$x) |
317 | { |
318 | $xi = $xi * $dd + $car; |
319 | $xi -= ($car = int($xi * $RBASE)) * $BASE; # see USE_MUL |
320 | } |
321 | push(@$x, $car); $car = 0; |
322 | for $yi (@$y) |
323 | { |
324 | $yi = $yi * $dd + $car; |
325 | $yi -= ($car = int($yi * $RBASE)) * $BASE; # see USE_MUL |
326 | } |
327 | } |
328 | else |
329 | { |
330 | push(@$x, 0); |
331 | } |
332 | @q = (); ($v2,$v1) = @$y[-2,-1]; |
333 | $v2 = 0 unless $v2; |
334 | while ($#$x > $#$y) |
335 | { |
336 | ($u2,$u1,$u0) = @$x[-3..-1]; |
337 | $u2 = 0 unless $u2; |
338 | #warn "oups v1 is 0, u0: $u0 $y->[-2] $y->[-1] l ",scalar @$y,"\n" |
339 | # if $v1 == 0; |
ee15d750 |
340 | # $q = (($u0 == $v1) ? 99999 : int(($u0*$BASE+$u1)/$v1)); |
341 | $q = (($u0 == $v1) ? $MAX_VAL : int(($u0*$BASE+$u1)/$v1)); |
574bacfe |
342 | --$q while ($v2*$q > ($u0*$BASE+$u1-$q*$v1)*$BASE+$u2); |
0716bf9b |
343 | if ($q) |
344 | { |
345 | ($car, $bar) = (0,0); |
346 | for ($yi = 0, $xi = $#$x-$#$y-1; $yi <= $#$y; ++$yi,++$xi) |
347 | { |
348 | $prd = $q * $y->[$yi] + $car; |
349 | $prd -= ($car = int($prd * $RBASE)) * $BASE; # see USE_MUL |
574bacfe |
350 | $x->[$xi] += $BASE if ($bar = (($x->[$xi] -= $prd + $bar) < 0)); |
0716bf9b |
351 | } |
352 | if ($x->[-1] < $car + $bar) |
353 | { |
354 | $car = 0; --$q; |
355 | for ($yi = 0, $xi = $#$x-$#$y-1; $yi <= $#$y; ++$yi,++$xi) |
356 | { |
574bacfe |
357 | $x->[$xi] -= $BASE |
0716bf9b |
358 | if ($car = (($x->[$xi] += $y->[$yi] + $car) > $BASE)); |
359 | } |
360 | } |
361 | } |
362 | pop(@$x); unshift(@q, $q); |
363 | } |
364 | if (wantarray) |
365 | { |
366 | @d = (); |
367 | if ($dd != 1) |
368 | { |
369 | $car = 0; |
370 | for $xi (reverse @$x) |
371 | { |
372 | $prd = $car * $BASE + $xi; |
373 | $car = $prd - ($tmp = int($prd / $dd)) * $dd; # see USE_MUL |
374 | unshift(@d, $tmp); |
375 | } |
376 | } |
377 | else |
378 | { |
379 | @d = @$x; |
380 | } |
381 | @$x = @q; |
382 | __strip_zeros($x); |
383 | __strip_zeros(\@d); |
384 | return ($x,\@d); |
385 | } |
386 | @$x = @q; |
387 | __strip_zeros($x); |
388 | return $x; |
389 | } |
390 | |
ee15d750 |
391 | sub _div_use_div |
392 | { |
393 | # ref to array, ref to array, modify first array and return remainder if |
394 | # in list context |
395 | # no longer handles sign |
396 | my ($c,$x,$yorg) = @_; |
397 | my ($car,$bar,$prd,$dd,$xi,$yi,@q,$v2,$v1); |
398 | |
399 | my (@d,$tmp,$q,$u2,$u1,$u0); |
400 | |
401 | $car = $bar = $prd = 0; |
402 | |
403 | my $y = [ @$yorg ]; |
404 | if (($dd = int($BASE/($y->[-1]+1))) != 1) |
405 | { |
406 | for $xi (@$x) |
407 | { |
408 | $xi = $xi * $dd + $car; |
409 | $xi -= ($car = int($xi / $BASE)) * $BASE; |
410 | } |
411 | push(@$x, $car); $car = 0; |
412 | for $yi (@$y) |
413 | { |
414 | $yi = $yi * $dd + $car; |
415 | $yi -= ($car = int($yi / $BASE)) * $BASE; |
416 | } |
417 | } |
418 | else |
419 | { |
420 | push(@$x, 0); |
421 | } |
422 | @q = (); ($v2,$v1) = @$y[-2,-1]; |
423 | $v2 = 0 unless $v2; |
424 | while ($#$x > $#$y) |
425 | { |
426 | ($u2,$u1,$u0) = @$x[-3..-1]; |
427 | $u2 = 0 unless $u2; |
428 | #warn "oups v1 is 0, u0: $u0 $y->[-2] $y->[-1] l ",scalar @$y,"\n" |
429 | # if $v1 == 0; |
430 | # $q = (($u0 == $v1) ? 99999 : int(($u0*$BASE+$u1)/$v1)); |
431 | $q = (($u0 == $v1) ? $MAX_VAL : int(($u0*$BASE+$u1)/$v1)); |
432 | --$q while ($v2*$q > ($u0*$BASE+$u1-$q*$v1)*$BASE+$u2); |
433 | if ($q) |
434 | { |
435 | ($car, $bar) = (0,0); |
436 | for ($yi = 0, $xi = $#$x-$#$y-1; $yi <= $#$y; ++$yi,++$xi) |
437 | { |
438 | $prd = $q * $y->[$yi] + $car; |
439 | $prd -= ($car = int($prd / $BASE)) * $BASE; |
440 | $x->[$xi] += $BASE if ($bar = (($x->[$xi] -= $prd + $bar) < 0)); |
441 | } |
442 | if ($x->[-1] < $car + $bar) |
443 | { |
444 | $car = 0; --$q; |
445 | for ($yi = 0, $xi = $#$x-$#$y-1; $yi <= $#$y; ++$yi,++$xi) |
446 | { |
447 | $x->[$xi] -= $BASE |
448 | if ($car = (($x->[$xi] += $y->[$yi] + $car) > $BASE)); |
449 | } |
450 | } |
451 | } |
452 | pop(@$x); unshift(@q, $q); |
453 | } |
454 | if (wantarray) |
455 | { |
456 | @d = (); |
457 | if ($dd != 1) |
458 | { |
459 | $car = 0; |
460 | for $xi (reverse @$x) |
461 | { |
462 | $prd = $car * $BASE + $xi; |
463 | $car = $prd - ($tmp = int($prd / $dd)) * $dd; |
464 | unshift(@d, $tmp); |
465 | } |
466 | } |
467 | else |
468 | { |
469 | @d = @$x; |
470 | } |
471 | @$x = @q; |
472 | __strip_zeros($x); |
473 | __strip_zeros(\@d); |
474 | return ($x,\@d); |
475 | } |
476 | @$x = @q; |
477 | __strip_zeros($x); |
478 | return $x; |
479 | } |
480 | |
dccbb853 |
481 | sub _mod |
482 | { |
483 | # if possible, use mod shortcut |
484 | my ($c,$x,$yo) = @_; |
485 | |
486 | # slow way since $y to big |
487 | if (scalar @$yo > 1) |
488 | { |
489 | my ($xo,$rem) = _div($c,$x,$yo); |
490 | return $rem; |
491 | } |
492 | my $y = $yo->[0]; |
493 | # both are single element |
494 | if (scalar @$x == 1) |
495 | { |
496 | $x->[0] %= $y; |
497 | return $x; |
498 | } |
499 | |
500 | my $b = $BASE % $y; |
501 | if ($b == 0) |
502 | { |
503 | # when BASE % Y == 0 then (B * BASE) % Y == 0 |
504 | # (B * BASE) % $y + A % Y => A % Y |
505 | # so need to consider only last element: O(1) |
506 | $x->[0] %= $y; |
507 | } |
508 | else |
509 | { |
510 | # else need to go trough all elemens: O(N) |
511 | # XXX not ready yet |
512 | my ($xo,$rem) = _div($c,$x,$yo); |
513 | return $rem; |
514 | |
515 | # my $i = 0; my $r = 1; |
516 | # print "Multi: "; |
517 | # foreach (@$x) |
518 | # { |
519 | # print "$_ $r $b $y\n"; |
520 | # print "\$_ % \$y = ",$_ % $y,"\n"; |
521 | # print "\$_ % \$y * \$b = ",($_ % $y) * $b,"\n"; |
522 | # $r += ($_ % $y) * $b; |
523 | # print "$r $b $y =>"; |
524 | # $r %= $y if $r > $y; |
525 | # print " $r\n"; |
526 | # } |
527 | # $x->[0] = $r; |
528 | } |
529 | splice (@$x,1); |
530 | return $x; |
531 | } |
532 | |
0716bf9b |
533 | ############################################################################## |
574bacfe |
534 | # shifts |
535 | |
536 | sub _rsft |
537 | { |
538 | my ($c,$x,$y,$n) = @_; |
539 | |
540 | if ($n != 10) |
541 | { |
542 | return; # we cant do this here, due to now _pow, so signal failure |
543 | } |
544 | else |
545 | { |
546 | # shortcut (faster) for shifting by 10) |
547 | # multiples of $BASE_LEN |
548 | my $dst = 0; # destination |
549 | my $src = _num($c,$y); # as normal int |
dccbb853 |
550 | my $rem = $src % $BASE_LEN; # remainder to shift |
574bacfe |
551 | $src = int($src / $BASE_LEN); # source |
552 | if ($rem == 0) |
553 | { |
554 | splice (@$x,0,$src); # even faster, 38.4 => 39.3 |
555 | } |
556 | else |
557 | { |
558 | my $len = scalar @$x - $src; # elems to go |
559 | my $vd; my $z = '0'x $BASE_LEN; |
560 | $x->[scalar @$x] = 0; # avoid || 0 test inside loop |
561 | while ($dst < $len) |
562 | { |
563 | $vd = $z.$x->[$src]; |
564 | #print "$dst $src '$vd' "; |
565 | $vd = substr($vd,-$BASE_LEN,$BASE_LEN-$rem); |
566 | #print "'$vd' "; |
567 | $src++; |
568 | $vd = substr($z.$x->[$src],-$rem,$rem) . $vd; |
569 | #print "'$vd1' "; |
570 | #print "'$vd'\n"; |
571 | $vd = substr($vd,-$BASE_LEN,$BASE_LEN) if length($vd) > $BASE_LEN; |
572 | $x->[$dst] = int($vd); |
573 | $dst++; |
574 | } |
575 | splice (@$x,$dst) if $dst > 0; # kill left-over array elems |
576 | pop @$x if $x->[-1] == 0; # kill last element if 0 |
577 | } # else rem == 0 |
578 | } |
579 | $x; |
580 | } |
581 | |
582 | sub _lsft |
583 | { |
584 | my ($c,$x,$y,$n) = @_; |
585 | |
586 | if ($n != 10) |
587 | { |
588 | return; # we cant do this here, due to now _pow, so signal failure |
589 | } |
590 | else |
591 | { |
592 | # shortcut (faster) for shifting by 10) since we are in base 10eX |
593 | # multiples of $BASE_LEN: |
594 | my $src = scalar @$x; # source |
595 | my $len = _num($c,$y); # shift-len as normal int |
dccbb853 |
596 | my $rem = $len % $BASE_LEN; # remainder to shift |
574bacfe |
597 | my $dst = $src + int($len/$BASE_LEN); # destination |
598 | my $vd; # further speedup |
599 | #print "src $src:",$x->[$src]||0," dst $dst:",$v->[$dst]||0," rem $rem\n"; |
600 | $x->[$src] = 0; # avoid first ||0 for speed |
601 | my $z = '0' x $BASE_LEN; |
602 | while ($src >= 0) |
603 | { |
604 | $vd = $x->[$src]; $vd = $z.$vd; |
605 | #print "s $src d $dst '$vd' "; |
606 | $vd = substr($vd,-$BASE_LEN+$rem,$BASE_LEN-$rem); |
607 | #print "'$vd' "; |
608 | $vd .= $src > 0 ? substr($z.$x->[$src-1],-$BASE_LEN,$rem) : '0' x $rem; |
609 | #print "'$vd' "; |
610 | $vd = substr($vd,-$BASE_LEN,$BASE_LEN) if length($vd) > $BASE_LEN; |
611 | #print "'$vd'\n"; |
612 | $x->[$dst] = int($vd); |
613 | $dst--; $src--; |
614 | } |
615 | # set lowest parts to 0 |
616 | while ($dst >= 0) { $x->[$dst--] = 0; } |
617 | # fix spurios last zero element |
618 | splice @$x,-1 if $x->[-1] == 0; |
619 | #print "elems: "; my $i = 0; |
620 | #foreach (reverse @$v) { print "$i $_ "; $i++; } print "\n"; |
621 | } |
622 | $x; |
623 | } |
624 | |
625 | ############################################################################## |
0716bf9b |
626 | # testing |
627 | |
628 | sub _acmp |
629 | { |
630 | # internal absolute post-normalized compare (ignore signs) |
631 | # ref to array, ref to array, return <0, 0, >0 |
b22b3e31 |
632 | # arrays must have at least one entry; this is not checked for |
0716bf9b |
633 | |
574bacfe |
634 | my ($c,$cx, $cy) = @_; |
0716bf9b |
635 | |
636 | #print "$cx $cy\n"; |
637 | my ($i,$a,$x,$y,$k); |
638 | # calculate length based on digits, not parts |
574bacfe |
639 | $x = _len('',$cx); $y = _len('',$cy); |
0716bf9b |
640 | # print "length: ",($x-$y),"\n"; |
574bacfe |
641 | my $lxy = $x - $y; # if different in length |
642 | return -1 if $lxy < 0; |
643 | return 1 if $lxy > 0; |
0716bf9b |
644 | #print "full compare\n"; |
645 | $i = 0; $a = 0; |
646 | # first way takes 5.49 sec instead of 4.87, but has the early out advantage |
b22b3e31 |
647 | # so grep is slightly faster, but more inflexible. hm. $_ instead of $k |
0716bf9b |
648 | # yields 5.6 instead of 5.5 sec huh? |
649 | # manual way (abort if unequal, good for early ne) |
650 | my $j = scalar @$cx - 1; |
651 | while ($j >= 0) |
652 | { |
653 | # print "$cx->[$j] $cy->[$j] $a",$cx->[$j]-$cy->[$j],"\n"; |
654 | last if ($a = $cx->[$j] - $cy->[$j]); $j--; |
655 | } |
574bacfe |
656 | return 1 if $a > 0; |
657 | return -1 if $a < 0; |
658 | return 0; # equal |
0716bf9b |
659 | # while it early aborts, it is even slower than the manual variant |
660 | #grep { return $a if ($a = $_ - $cy->[$i++]); } @$cx; |
661 | # grep way, go trough all (bad for early ne) |
662 | #grep { $a = $_ - $cy->[$i++]; } @$cx; |
663 | #return $a; |
664 | } |
665 | |
666 | sub _len |
667 | { |
dccbb853 |
668 | # compute number of digits in bigint, minus the sign |
b22b3e31 |
669 | # int() because add/sub sometimes leaves strings (like '00005') instead of |
dccbb853 |
670 | # int ('5') in this place, thus causing length() to report wrong length |
574bacfe |
671 | my $cx = $_[1]; |
0716bf9b |
672 | |
574bacfe |
673 | return (@$cx-1)*$BASE_LEN+length(int($cx->[-1])); |
0716bf9b |
674 | } |
675 | |
676 | sub _digit |
677 | { |
678 | # return the nth digit, negative values count backward |
679 | # zero is rightmost, so _digit(123,0) will give 3 |
574bacfe |
680 | my ($c,$x,$n) = @_; |
0716bf9b |
681 | |
574bacfe |
682 | my $len = _len('',$x); |
0716bf9b |
683 | |
684 | $n = $len+$n if $n < 0; # -1 last, -2 second-to-last |
685 | $n = abs($n); # if negative was too big |
686 | $len--; $n = $len if $n > $len; # n to big? |
687 | |
574bacfe |
688 | my $elem = int($n / $BASE_LEN); # which array element |
689 | my $digit = $n % $BASE_LEN; # which digit in this element |
0716bf9b |
690 | $elem = '0000'.@$x[$elem]; # get element padded with 0's |
691 | return substr($elem,-$digit-1,1); |
692 | } |
693 | |
694 | sub _zeros |
695 | { |
696 | # return amount of trailing zeros in decimal |
697 | # check each array elem in _m for having 0 at end as long as elem == 0 |
698 | # Upon finding a elem != 0, stop |
574bacfe |
699 | my $x = $_[1]; |
0716bf9b |
700 | my $zeros = 0; my $elem; |
701 | foreach my $e (@$x) |
702 | { |
703 | if ($e != 0) |
704 | { |
574bacfe |
705 | $elem = "$e"; # preserve x |
706 | $elem =~ s/.*?(0*$)/$1/; # strip anything not zero |
707 | $zeros *= $BASE_LEN; # elems * 5 |
708 | $zeros += CORE::length($elem); # count trailing zeros |
709 | last; # early out |
0716bf9b |
710 | } |
574bacfe |
711 | $zeros ++; # real else branch: 50% slower! |
0716bf9b |
712 | } |
713 | return $zeros; |
714 | } |
715 | |
716 | ############################################################################## |
717 | # _is_* routines |
718 | |
719 | sub _is_zero |
720 | { |
721 | # return true if arg (BINT or num_str) is zero (array '+', '0') |
574bacfe |
722 | my $x = $_[1]; |
0716bf9b |
723 | return (((scalar @$x == 1) && ($x->[0] == 0))) <=> 0; |
724 | } |
725 | |
726 | sub _is_even |
727 | { |
728 | # return true if arg (BINT or num_str) is even |
574bacfe |
729 | my $x = $_[1]; |
0716bf9b |
730 | return (!($x->[0] & 1)) <=> 0; |
731 | } |
732 | |
733 | sub _is_odd |
734 | { |
735 | # return true if arg (BINT or num_str) is even |
574bacfe |
736 | my $x = $_[1]; |
0716bf9b |
737 | return (($x->[0] & 1)) <=> 0; |
738 | } |
739 | |
740 | sub _is_one |
741 | { |
742 | # return true if arg (BINT or num_str) is one (array '+', '1') |
574bacfe |
743 | my $x = $_[1]; |
0716bf9b |
744 | return (scalar @$x == 1) && ($x->[0] == 1) <=> 0; |
745 | } |
746 | |
747 | sub __strip_zeros |
748 | { |
749 | # internal normalization function that strips leading zeros from the array |
750 | # args: ref to array |
0716bf9b |
751 | my $s = shift; |
752 | |
753 | my $cnt = scalar @$s; # get count of parts |
754 | my $i = $cnt-1; |
755 | #print "strip: cnt $cnt i $i\n"; |
756 | # '0', '3', '4', '0', '0', |
757 | # 0 1 2 3 4 |
758 | # cnt = 5, i = 4 |
759 | # i = 4 |
760 | # i = 3 |
761 | # => fcnt = cnt - i (5-2 => 3, cnt => 5-1 = 4, throw away from 4th pos) |
762 | # >= 1: skip first part (this can be zero) |
763 | while ($i > 0) { last if $s->[$i] != 0; $i--; } |
764 | $i++; splice @$s,$i if ($i < $cnt); # $i cant be 0 |
765 | return $s; |
766 | } |
767 | |
768 | ############################################################################### |
769 | # check routine to test internal state of corruptions |
770 | |
771 | sub _check |
772 | { |
bd05a461 |
773 | # used by the test suite |
574bacfe |
774 | my $x = $_[1]; |
0716bf9b |
775 | |
0716bf9b |
776 | return "$x is not a reference" if !ref($x); |
777 | |
778 | # are all parts are valid? |
779 | my $i = 0; my $j = scalar @$x; my ($e,$try); |
780 | while ($i < $j) |
781 | { |
782 | $e = $x->[$i]; $e = 'undef' unless defined $e; |
783 | $try = '=~ /^[\+]?[0-9]+\$/; '."($x, $e)"; |
784 | last if $e !~ /^[+]?[0-9]+$/; |
dccbb853 |
785 | $try = '=~ /^[\+]?[0-9]+\$/; '."($x, $e) (stringify)"; |
786 | last if "$e" !~ /^[+]?[0-9]+$/; |
787 | $try = '=~ /^[\+]?[0-9]+\$/; '."($x, $e) (cat-stringify)"; |
788 | last if '' . "$e" !~ /^[+]?[0-9]+$/; |
0716bf9b |
789 | $try = ' < 0 || >= $BASE; '."($x, $e)"; |
790 | last if $e <0 || $e >= $BASE; |
791 | # this test is disabled, since new/bnorm and certain ops (like early out |
792 | # in add/sub) are allowed/expected to leave '00000' in some elements |
793 | #$try = '=~ /^00+/; '."($x, $e)"; |
794 | #last if $e =~ /^00+/; |
795 | $i++; |
796 | } |
797 | return "Illegal part '$e' at pos $i (tested: $try)" if $i < $j; |
798 | return 0; |
799 | } |
800 | |
801 | 1; |
802 | __END__ |
803 | |
804 | =head1 NAME |
805 | |
806 | Math::BigInt::Calc - Pure Perl module to support Math::BigInt |
807 | |
808 | =head1 SYNOPSIS |
809 | |
ee15d750 |
810 | Provides support for big integer calculations. Not intended to be used by other |
811 | modules (except Math::BigInt::Cached). Other modules which sport the same |
812 | functions can also be used to support Math::Bigint, like Math::BigInt::Pari. |
0716bf9b |
813 | |
814 | =head1 DESCRIPTION |
815 | |
816 | In order to allow for multiple big integer libraries, Math::BigInt |
817 | was rewritten to use library modules for core math routines. Any |
818 | module which follows the same API as this can be used instead by |
819 | using the following call: |
820 | |
ee15d750 |
821 | use Math::BigInt lib => 'libname'; |
0716bf9b |
822 | |
823 | =head1 EXPORT |
824 | |
bd05a461 |
825 | The following functions MUST be defined in order to support |
0716bf9b |
826 | the use by Math::BigInt: |
827 | |
828 | _new(string) return ref to new object from ref to decimal string |
829 | _zero() return a new object with value 0 |
830 | _one() return a new object with value 1 |
831 | |
832 | _str(obj) return ref to a string representing the object |
833 | _num(obj) returns a Perl integer/floating point number |
834 | NOTE: because of Perl numeric notation defaults, |
835 | the _num'ified obj may lose accuracy due to |
836 | machine-dependend floating point size limitations |
837 | |
838 | _add(obj,obj) Simple addition of two objects |
839 | _mul(obj,obj) Multiplication of two objects |
840 | _div(obj,obj) Division of the 1st object by the 2nd |
b22b3e31 |
841 | In list context, returns (result,remainder). |
842 | NOTE: this is integer math, so no |
843 | fractional part will be returned. |
844 | _sub(obj,obj) Simple subtraction of 1 object from another |
0716bf9b |
845 | a third, optional parameter indicates that the params |
846 | are swapped. In this case, the first param needs to |
847 | be preserved, while you can destroy the second. |
848 | sub (x,y,1) => return x - y and keep x intact! |
849 | |
850 | _acmp(obj,obj) <=> operator for objects (return -1, 0 or 1) |
851 | |
852 | _len(obj) returns count of the decimal digits of the object |
853 | _digit(obj,n) returns the n'th decimal digit of object |
854 | |
855 | _is_one(obj) return true if argument is +1 |
856 | _is_zero(obj) return true if argument is 0 |
857 | _is_even(obj) return true if argument is even (0,2,4,6..) |
858 | _is_odd(obj) return true if argument is odd (1,3,5,7..) |
859 | |
860 | _copy return a ref to a true copy of the object |
861 | |
862 | _check(obj) check whether internal representation is still intact |
863 | return 0 for ok, otherwise error message as string |
864 | |
bd05a461 |
865 | The following functions are optional, and can be defined if the underlying lib |
ee15d750 |
866 | has a fast way to do them. If undefined, Math::BigInt will use a pure, but |
bd05a461 |
867 | slow, Perl way as fallback to emulate these: |
0716bf9b |
868 | |
869 | _from_hex(str) return ref to new object from ref to hexadecimal string |
870 | _from_bin(str) return ref to new object from ref to binary string |
871 | |
ee15d750 |
872 | _as_hex(str) return ref to scalar string containing the value as |
873 | unsigned hex string, with the '0x' prepended. |
874 | Leading zeros must be stripped. |
875 | _as_bin(str) Like as_hex, only as binary string containing only |
876 | zeros and ones. Leading zeros must be stripped and a |
877 | '0b' must be prepended. |
878 | |
0716bf9b |
879 | _rsft(obj,N,B) shift object in base B by N 'digits' right |
dccbb853 |
880 | For unsupported bases B, return undef to signal failure |
0716bf9b |
881 | _lsft(obj,N,B) shift object in base B by N 'digits' left |
dccbb853 |
882 | For unsupported bases B, return undef to signal failure |
0716bf9b |
883 | |
884 | _xor(obj1,obj2) XOR (bit-wise) object 1 with object 2 |
dccbb853 |
885 | Note: XOR, AND and OR pad with zeros if size mismatches |
0716bf9b |
886 | _and(obj1,obj2) AND (bit-wise) object 1 with object 2 |
887 | _or(obj1,obj2) OR (bit-wise) object 1 with object 2 |
888 | |
dccbb853 |
889 | _mod(obj,obj) Return remainder of div of the 1st by the 2nd object |
0716bf9b |
890 | _sqrt(obj) return the square root of object |
891 | _pow(obj,obj) return object 1 to the power of object 2 |
892 | _gcd(obj,obj) return Greatest Common Divisor of two objects |
893 | |
b22b3e31 |
894 | _zeros(obj) return number of trailing decimal zeros |
0716bf9b |
895 | |
896 | _dec(obj) decrement object by one (input is >= 1) |
897 | _inc(obj) increment object by one |
898 | |
b22b3e31 |
899 | Input strings come in as unsigned but with prefix (i.e. as '123', '0xabc' |
0716bf9b |
900 | or '0b1101'). |
901 | |
b22b3e31 |
902 | Testing of input parameter validity is done by the caller, so you need not |
574bacfe |
903 | worry about underflow (f.i. in C<_sub()>, C<_dec()>) nor about division by |
904 | zero or similar cases. |
905 | |
906 | The first parameter can be modified, that includes the possibility that you |
907 | return a reference to a completely different object instead. Although keeping |
dccbb853 |
908 | the reference and just changing it's contents is prefered over creating and |
909 | returning a different reference. |
574bacfe |
910 | |
911 | Return values are always references to objects or strings. Exceptions are |
912 | C<_lsft()> and C<_rsft()>, which return undef if they can not shift the |
913 | argument. This is used to delegate shifting of bases different than 10 back |
dccbb853 |
914 | to Math::BigInt, which will use some generic code to calculate the result. |
574bacfe |
915 | |
916 | =head1 WRAP YOUR OWN |
917 | |
918 | If you want to port your own favourite c-lib for big numbers to the |
919 | Math::BigInt interface, you can take any of the already existing modules as |
920 | a rough guideline. You should really wrap up the latest BigInt and BigFloat |
bd05a461 |
921 | testsuites with your module, and replace in them any of the following: |
574bacfe |
922 | |
923 | use Math::BigInt; |
924 | |
bd05a461 |
925 | by this: |
574bacfe |
926 | |
927 | use Math::BigInt lib => 'yourlib'; |
928 | |
929 | This way you ensure that your library really works 100% within Math::BigInt. |
0716bf9b |
930 | |
931 | =head1 LICENSE |
932 | |
933 | This program is free software; you may redistribute it and/or modify it under |
934 | the same terms as Perl itself. |
935 | |
936 | =head1 AUTHORS |
937 | |
938 | Original math code by Mark Biggar, rewritten by Tels L<http://bloodgate.com/> |
939 | in late 2000, 2001. |
940 | Seperated from BigInt and shaped API with the help of John Peacock. |
941 | |
942 | =head1 SEE ALSO |
943 | |
ee15d750 |
944 | L<Math::BigInt>, L<Math::BigFloat>, L<Math::BigInt::BitVect>, |
945 | L<Math::BigInt::GMP>, L<Math::BigInt::Cached> and L<Math::BigInt::Pari>. |
0716bf9b |
946 | |
947 | =cut |