Package org.apfloat.spi

Interface AdditionStrategy<T>

Type Parameters:
T - The element type of the addition strategy.
All Known Implementing Classes:
DoubleAdditionStrategy, FloatAdditionStrategy, IntAdditionStrategy, LongAdditionStrategy
public interface AdditionStrategy<T>
Generic addition strategy. Also subtraction, "short" multiplication and "short" division can be done.
Since:
1.7.0
Version:
1.7.0
Author:
Mikko Tommila

  • Method Details

    • add

      T add(DataStorage.Iterator src1, DataStorage.Iterator src2, T carry, DataStorage.Iterator dst, long size) throws ApfloatRuntimeException
      Addition in some base. Adds the data words of src1 and src2 and stores the result to dst. src2 may be null, in which case it is ignored (only the carry is propagated).

      Essentially calculates dst[i] = src1[i] + src2[i].

      Parameters:
      src1 - First source data sequence. Can be null, in which case it's ignored.
      src2 - Second source data sequence. Can be null, in which case it's ignored.
      carry - Input carry bit. This is added to the first (rightmost) word in the accessed sequence.
      dst - Destination data sequence.
      size - Number of elements to process.
      Returns:
      Overflow carry bit. Propagated carry bit from the addition of the last (leftmost) word in the accessed sequence.
      Throws:
      ApfloatRuntimeException

    • subtract

      T subtract(DataStorage.Iterator src1, DataStorage.Iterator src2, T carry, DataStorage.Iterator dst, long size) throws ApfloatRuntimeException
      Subtraction in some base. Subtracts the data words of src1 and src2 and stores the result to dst. src1 and src2 may be null, in which case they are ignored (the values are assumed to be zero and only the carry is propagated).

      Essentially calculates dst[i] = src1[i] - src2[i].

      Parameters:
      src1 - First source data sequence. Can be null, in which case the input values are assumed to be zero.
      src2 - Second source data sequence. Can be null, in which case it's ignored, or can be the same as dst.
      carry - Input carry bit. This is subtracted from the first (rightmost) word in the accessed sequence.
      dst - Destination data sequence.
      size - Number of elements to process.
      Returns:
      Overflow carry bit. Propagated carry bit from the subtraction of the last (leftmost) word in the accessed sequence. The value is 1 if the carry is set, and 0 otherwise.
      Throws:
      ApfloatRuntimeException

    • multiplyAdd

      T multiplyAdd(DataStorage.Iterator src1, DataStorage.Iterator src2, T src3, T carry, DataStorage.Iterator dst, long size) throws ApfloatRuntimeException
      Multiplication and addition in some base. Multiplies the data words of src1 by src3 and adds the result to the words in src2, and stores the result to dst. src2 may be null, in which case it is ignored (the values are assumed to be zero).

      Assumes that the result from the addition doesn't overflow the upper result word (to larger than the base). This is the case e.g. when using this method to perform an arbitrary precision multiplication.

      Essentially calculates dst[i] = src1[i] * src3 + src2[i].

      Parameters:
      src1 - First source data sequence.
      src2 - Second source data sequence. Can be null, in which case it's ignored, or can be the same as dst.
      src3 - Multiplicand. All elements of src1 are multiplied by this value.
      carry - Input carry word. This is added to the first (rightmost) word in the accessed sequence.
      dst - Destination data sequence.
      size - Number of elements to process.
      Returns:
      Overflow carry word. Propagated carry word from the multiplication and addition of the last (leftmost) word in the accessed sequence.
      Throws:
      ApfloatRuntimeException

    • divide

      T divide(DataStorage.Iterator src1, T src2, T carry, DataStorage.Iterator dst, long size) throws ApfloatRuntimeException
      Division in some base. Divides the data words of src1 by src2 and stores the result to dst. src1 may be null, in which case it is ignored (the values are assumed to be zero and only the carry division is propagated).

      Essentially calculates dst[i] = src1[i] / src2.

      Parameters:
      src1 - First source data sequence. Can be null, in which case the input values are assumed to be zero.
      src2 - Divisor. All elements of src1 are divided by this value.
      carry - Input carry word. Used as the upper word for the division of the first input element. This should be the remainder word returned from the previous block processed.
      dst - Destination data sequence.
      size - Number of elements to process.
      Returns:
      Remainder word of the propagated division of the last (rightmost) word in the accessed sequence.
      Throws:
      ApfloatRuntimeException

    • zero

      T zero()
      Returns the zero element.
      Returns:
      Zero of the correct element type.