公开(公告)号：FR2396360A1

公开(公告)日：1979-01-26

申请号：FR7817690

申请日：1978-06-05

Applicant: IBM

Inventor： LANGDON GLEN G JR

IPC: G06F7/38 ,  G06F7/00 ,  G06F7/483 ,  G06F7/493 ,  G06F7/494 ,  G06F7/50 ,  G06F7/506 ,  G06F7/385

Abstract: Method and means are described for the tracking of digit significance upon operands arithmetically combined in a series of binary operations such as addition, subtraction, or shifting in a decimal computer. The digits are decimally encoded in a format having enough excess capacity such that nonsignificant digits are unique. As part of the arithmetic combining of the operand, pairs of digits of like order but possibly mismatched as to significance and by observing a predetermined rounding rule may also cause a carry value to be propagated to a digit position of higher order. In subtraction by complement addition, an additional carry is propagated to a higher order position conditioned upon there being either a local overflow, a nonsignificant subtrahend, or a nonsignificant minuend and a subtrahend less than an amount specified by a rounding rule. Between the two operands, this results in the rounding of the more precise operand to the least significant digit position of the less precise operand. The method and means are applicable to floating point, sign plus magnitude, radix and diminished radix complement number representation forms.

公开(公告)号：FR2382718A1

公开(公告)日：1978-09-29

申请号：FR7803456

申请日：1978-02-01

Applicant: IBM

Inventor： LANGDON GLEN G JR ,  RISSANEN JORMA J

IPC: H03M7/40 ,  G06F5/00 ,  H03K13/24

Abstract: There is disclosed a method and means for compacting (encoding) and decompacting (decoding) binary bit strings which avoids the blocking of string elements required by Huffman coding and the ever increasing memory as is the case in simple enumerative coding. The method and means arithmetically encodes successive terms or symbols in a symbol string s=ai aj . . . , in which each new term ak in a source alphabet of N symbols gives rise to a new encoded string C(sak) and a new length indicator L(sak). The method and means comprises the steps of forming L(sak) from the recursion L(sak)=L(s)+l(ak), where l(ak) is a rational approximation of log2 1/p(ak), p(ak) being the a'priori probability of occurrence of ak, and l(ak) being so constrained that the Kraft inequality is satisfied: AND FORMING C(sak) from the recursion C(s)+[Pk-12L(sak)],where: AND WHERE Pk-1 is the cumulative probability of occurrence of an arbitrary ordering of all symbols.