用于将浮点格式化为十进制数字的算法

Question

Recently, I've studied the IEEE 754-1985 standard. After reading some materials on the web, I think I've grasped the basic ideas underlie floating point numbers. While, I have a question that can not be solved by searching the web. That is, how the computer system convert the floating point number(in binary form) to its decimal equivalent and print it on the screen as a string for our reading?

To be clear, for example, In C/C++ language, we just write:

...
float x,y,z;
x=3.14;
y=2.71;
z=x+y;
printf("%f",z);

we will get 5.85 as a result on the screen. But how computer(or just the C language and its compiler) convert x,y to IEEE binary floating point form and after the calculation by CPU, convert the IEEE form back to the decimal string form?

Is there some algorithms and for this conversion? By the way, is it hardware or software that does this trick?

Thank you! :-)

Answer 1

you should have a look at the glibc , especially the printf function. this is where the authoritative answer for at least one popular implementation lies. i doubt that it is too readable, though.

it is definitely done in software.

http://www.gnu.org/software/libc/download.html