Rational Developer for System z
PL/I for Windows, Version 8.0, Programming Guide

PRECTYPE

The PRECTYPE option determines how the compiler derives the attributes for the MULTIPLY, DIVIDE, ADD and SUBTRACT built-in functions when the operands are FIXED and at least one is FIXED BIN.

Read syntax diagramSkip visual syntax diagram                .-ANS-------.
>>-PRECTYPE--(--+-DECDIGIT--+--)-------------------------------><
                '-DECRESULT-'
 
ANS
Under PRECTYPE(ANS), the value p in BIF(x,y,p) and in BIF(x, y, p, 0) is interpreted as specifying a binary number of digits, the operation is performed as a binary operation and the result has the attributes FIXED BIN(p,0).

However, for BIF(x, y, p, q) if q is not-zero, then the operation will be performed as a decimal operation and the result will have the attributes FIXED DEC(t,u) where t and u are the decimal equivalents of p and q, namely t = 1 + ceil(p / 3.32) and u = ceil(q / 3.32). In this case, x, y, p and q are effectively all converted to decimal (in contrast to the next suboption which converts only x and y to decimal and does so even if q is zero). The compiler will issue the Informational message 1BM1053 in this situation.

DECDIGIT
Under PRECTYPE(DECDIGIT), the value p in BIF(x,y,p) and in BIF(x, y, p, 0) is interpreted as specifying a decimal number of digits, the operation is performed as a binary operation and the result has the attributes FIXED BIN(s) where s is the corresponding binary equivalent to p (namely s = ceil(3.32*p)). For an instance of BIF(x, y, p, q) where q is not-zero, the results under PRECTYPE(DECDIGIT) are the same as described below under PRECTYPE(DECRESULT).
DECRESULT
Under PRECTYPE(DECRESULT), the value p in BIF(x,y,p) and the values p and q in BIF(x,y,p,q) are interpreted as specifying a decimal number of digits, the operation is performed as a decimal operation and the result has the attributes FIXED DEC(p,0) or FIXED DEC(p,q) respectively. The result is the same as would be produced if the DECIMAL built-in were applied to x and y.

PRECTYPE(ANS) is the default.


Terms of use | Feedback

This information center is powered by Eclipse technology. (http://www.eclipse.org)