mtlb_int8

Description

Matlab and Scilab int8 behave differently :

• Beyong the main interval bounds, Scilab's int8() wraps, while Matlab's mtlb_int8() saturates.
• For non integer values, Scilab's int8() truncates the fractional part, while Matlab's mtlb_int8() rounds to the nearest integer.

x	int8(x)	mtlb_int8(x)
%nan	0	0
-%inf	-128	-128
-130	126	-128
-129	127	-128
-128	-128	-128
-127	-127	-127
-80.7	-80	-81
-80.5	-80	-81
-80.4	-80	-80
10.2	10	10
10.5	10	11
10.51	10	11
127	127	127
128	-128	127
129	-127	127
%inf	127	127

mtlb_int8 has not to be used for hand coded functions. The function mtlb_int8(A) is used by mfile2sci to replace int8(A) when it is not possible to know what is the input while porting Matlab code to Scilab. This function then determines the correct semantic at run time.