Menu
Search

Library Reference

overview

Report a bug
If you spot a problem with this page, click here to create a GitHub issue.
Improve this page
Quickly fork, edit online, and submit a pull request for this page. Requires a signed-in GitHub account. This works well for small changes. If you'd like to make larger changes you may want to consider using a local clone.

mir.bignum.internal.phobos_kernel

Fundamental operations for arbitrary-precision arithmetic
These functions are for internal use only.
package(mir.bignum) pure nothrow @nogc @safe void mulInternal(BigDigit[] result, const(BigDigit)[] x, const(BigDigit)[] y, BigDigit[] scratchbuff);
General unsigned multiply routine for bigints. Sets result = x * y.
The length of y must not be larger than the length of x. Different algorithms are used, depending on the lengths of x and y.

TODO "Modern Computer Arithmetic" suggests the OddEvenKaratsuba algorithm for the unbalanced case. (But I doubt it would be faster in practice).

package(mir.bignum) pure nothrow @nogc @safe void squareInternal(BigDigit[] result, const BigDigit[] x, BigDigit[] scratchbuff);
General unsigned squaring routine for BigInts. Sets result = x*x.

NOTE If the highest half-digit of x is zero, the highest digit of result will also be zero.

package(mir.bignum) pure nothrow @nogc @safe void divModInternal(BigDigit[] quotient, BigDigit[] remainder, const BigDigit[] u, const BigDigit[] v, BigDigit[] buffer);
if remainder is null, only calculate quotient.
package(mir.bignum) pure nothrow @nogc @safe size_t biguintToOctal(char[] buff, const(BigDigit)[] data);
Convert a big uint into an octal string.
Parameters:
char[] buff The destination buffer for the octal string. Must be large enough to store the result, including leading zeroes, which is ceil(data.length * BigDigitBits / 3) characters. The buffer is filled from back to front, starting from buff[$-1].
const(BigDigit)[] data The biguint to be converted.
Returns:
The index of the leading non-zero digit in buff. Will be buff.length - 1 if the entire big uint is zero.
package(mir.bignum) pure nothrow @nogc @safe size_t biguintToDecimal(char[] buff, BigDigit[] data);
Convert a big uint into a decimal string.
Parameters:
char[] buff The destination buffer for the decimal string. Must be large enough to store the result, including leading zeros. Will be filled backwards, starting from buff[$-1].
BigDigit[] data The biguint to be converted. Will be destroyed.
buff.length must be >= (data.length*32)/log2(10) = 9.63296 * data.length.
Returns:
the lowest index of buff which was used.
int biguintFromDecimal(Range)(BigDigit[] data, Range s)
if (isInputRange!Range && isSomeChar!(ElementType!Range) && !isInfinite!Range);
Convert a decimal string into a big uint.
Parameters:
BigDigit[] data The biguint to be receive the result. Must be large enough to store the result.
Range s The decimal string. May contain or 0 .. 9
The required length for the destination buffer is slightly less than 1 + s.length/log2(10) = 1 + s.length/3.3219.
Returns:
the highest index of data which was used. 0 if case of failure.