jncraton / floating-point

Floating Point Conversion

This assignment demonstrates converting from a floating point number to an integer using only integer and bitwise operations.

You will create implementations for the following functions:

• is_negative
• get_raw_exponent
• get_exponent
• get_raw_mantissa
• get_mantissa
• float_to_int

The first five functions provide the framework to create a software-based float to int converter. These function will then be combined to create a float to int converter in the function float_to_int.

This tool may be useful for exploring floating point numbers.

Implementation Notes

The IEEE Standard for Floating-Point Arithmetic includes a number of special cases. The implementation in this assignment need only properly handle normal numbers. It could be extended to support other numbers, such as zero.

Testing

The included makefile will build the fp.c file and run a series of tests to confirm that a valid implementation has been added to the file. The tests can be run by executing:

make test

It is strongly recommend to read the tests included at the bottom of the fp.c program in order to help understand the specification for each of the functions to be implemented.