0 "ˆl–äY>”TËm q@µq¦®Å1hßÁ ƒh¹À¿0xfeffffff 0ˆl–ßi~””ÂXÔåÆEpN©‹Fÿ „u÷ÿ_‹uÐb &=LV[(ÂÁ?÷package Math::BigInt::Calc; use 5.006002; use strict; # use warnings; # dont use warnings for older Perls our $VERSION = '1.997'; # Package to store unsigned big integers in decimal and do math with them # Internally the numbers are stored in an array with at least 1 element, no # leading zero parts (except the first) and in base 1eX where X is determined # automatically at loading time to be the maximum possible value # todo: # - fully remove funky $# stuff in div() (maybe - that code scares me...) # USE_MUL: due to problems on certain os (os390, posix-bc) "* 1e-5" is used # instead of "/ 1e5" at some places, (marked with USE_MUL). Other platforms # BS2000, some Crays need USE_DIV instead. # The BEGIN block is used to determine which of the two variants gives the # correct result. # Beware of things like: # $i = $i * $y + $car; $car = int($i / $BASE); $i = $i % $BASE; # This works on x86, but fails on ARM (SA1100, iPAQ) due to whoknows what # reasons. So, use this instead (slower, but correct): # $i = $i * $y + $car; $car = int($i / $BASE); $i -= $BASE * $car; ############################################################################## # global constants, flags and accessory # announce that we are compatible with MBI v1.83 and up sub api_version () { 2; } # constants for easier life my ($BASE,$BASE_LEN,$RBASE,$MAX_VAL); my ($AND_BITS,$XOR_BITS,$OR_BITS); my ($AND_MASK,$XOR_MASK,$OR_MASK); sub _base_len { # Set/get the BASE_LEN and assorted other, connected values. # Used only by the testsuite, the set variant is used only by the BEGIN # block below: shift; my ($b, $int) = @_; if (defined $b) { # avoid redefinitions undef &_mul; undef &_div; if ($] >= 5.008 && $int && $b > 7) { $BASE_LEN = $b; *_mul = \&_mul_use_div_64; *_div = \&_div_use_div_64; $BASE = int("1e".$BASE_LEN); $MAX_VAL = $BASE-1; return $BASE_LEN unless wantarray; return ($BASE_LEN, $BASE, $AND_BITS, $XOR_BITS, $OR_BITS, $BASE_LEN, $MAX_VAL,); } # find whether we can use mul or div in mul()/div() $BASE_LEN = $b+1; my $caught = 0; while (--$BASE_LEN > 5) { $BASE = int("1e".$BASE_LEN); $RBASE = abs('1e-'.$BASE_LEN); # see USE_MUL $caught = 0; $caught += 1 if (int($BASE * $RBASE) != 1); # should be 1 $caught += 2 if (int($BASE / $BASE) != 1); # should be 1 last if $caught != 3; } $BASE = int("1e".$BASE_LEN); $RBASE = abs('1e-'.$BASE_LEN); # see USE_MUL $MAX_VAL = $BASE-1; # ($caught & 1) != 0 => cannot use MUL # ($caught & 2) != 0 => cannot use DIV if ($caught == 2) # 2 { # must USE_MUL since we cannot use DIV *_mul = \&_mul_use_mul; *_div = \&_div_use_mul; } else # 0 or 1 { # can USE_DIV instead *_mul = \&_mul_use_div; *_div = \&_div_use_div; } } return $BASE_LEN unless wantarray; return ($BASE_LEN, $BASE, $AND_BITS, $XOR_BITS, $OR_BITS, $BASE_LEN, $MAX_VAL); } sub _new { # (ref to string) return ref to num_array # Convert a number from string format (without sign) to internal base # 1ex format. Assumes normalized value as input. my $il = length($_[1])-1; # < BASE_LEN due len-1 above return [ int($_[1]) ] if $il < $BASE_LEN; # shortcut for short numbers # this leaves '00000' instead of int 0 and will be corrected after any op [ reverse(unpack("a" . ($il % $BASE_LEN+1) . ("a$BASE_LEN" x ($il / $BASE_LEN)), $_[1])) ]; } BEGIN { # from Daniel Pfeiffer: determine largest group of digits that is precisely # multipliable with itself plus carry # Test now changed to expect the proper pattern, not a result off by 1 or 2 my ($e, $num) = 3; # lowest value we will use is 3+1-1 = 3 do { $num = ('9' x ++$e) + 0; $num *= $num + 1.0;