A bug that caused negative numbers and exponential notation to be formatted improperly

[kiritannplum]

I am sorry if there are any unnatural sentences as I am using machine translation.

As the title says, there is a bug that causes negative numbers and exponential notation to be formatted improperly.

Here is an example.
Before:

float x = 1.0e-3;
float y = -1.0;

After:

float x = 1.0e - 3;
float y = - 1.0;