The floating point number X.1 in C/C ++

I saw this question in a topic just now.
Cannot be correctly divided into N.1 floating point numbers

After debugging, we found that
When reading a floating point number, 2.1 or 3.1,
The actual values are 2.0999999 and 3.099999.

The reason is that because 0.1 is represented as a circular decimal number in binary format, it cannot be correctly expressed. 0. 0001100110011 ....
Therefore, it is 0.09999999

======================================

So when we split up floating-point numbers,
You can first read the data in a string and then split it.
Of course, this is just my idea. Maybe there is a better way.