C++ library: (sw) fixed point (through overload)

[legacy]

git clone https://[email protected]/p/fixedpoint/

source browse

Code: [Select]

./test2 1.1 1.2
sizeof fixed_t = 12..12 bytes
a    = 1.1
b    = 1.2
a+b  = 2.3
a-b  = -0.1
a*b  = 1.32
-a   = -1.1
b++  = 1.3
a[4] = 1.1000 ... operator precision
a/b = 0.9
a.whole() = 1
is a == b? no
is a != b? yes
is a <  b? yes
is a <= b? yes
is a >  b? no
is a >= b? no

to summarize, this library make possible to use a fixed point type as natural as the floating point type
to make that, it defines a class and then it uses the C++ overload feature in order to redefine operators {+,-,*,/, ...}

enjoy =)

[legacy]

p.s.
it is not a professional code, also a lot of improvements are needed, being warned about that 

[nctnico]

I didn't look at the source but I'm wondering what kind of code the compiler creates to turn the floating point values into integers. That usually is pretty ugly and slow without FPU. But there are some neat tricks here. See:
http://stackoverflow.com/questions/429632/how-to-speed-up-floating-point-to-integer-number-conversion/429812#429812