In C# or Java (I don’t care which) write a program which correctly calculates add, subtract, multiply and divide using our ‘minifloat’ binary format using an algorithm you code yourself (see below)
Your program only needs to work on two ‘numbers’ at a time, read those in from a text file. For each ‘number’ store the sign, exponent and mantissa separately. You can hard code your examples.
E.g.
Struct minifloat{
Int sign; // sign, actually 1 bit
Int exponent; // this is the exponent and is 4 bits long (0-15)
Int mantissa; //this is the fraction part and is 3bits long (0-7), but I would keep the 1. Stored explicitly, so it becomes 0-15 and 4 digits long.
}
Show testing for all 4 (add, sub, mul, div) with at least 3 examples (and compare to results you do by hand).
Make sure you are error checking for NAN and +/- 0.
Note: This is an obviously contrived example, the intent is to force you to think about the algorithm, not to make software that anyone would ever want to use.