isinf is available on MSVC

[nobu]

At least Visual Studio 2015 (MSVC 14.0) provides the macro conforming to C99.
And, float.h is necessary to use _finite() like the current code.

[nobu]

I see, the problem is mingw, not mswin.

https://github.com/ruby/prism/actions/runs/12141353366/job/33853261523?pr=3274#step:4:37

src/prism.c:4145:34: error: conversion from 'double' to 'float' may change value [-Werror=float-conversion]
 4145 |     if (errno == ERANGE && isinf(value)) {
      |                                  ^~~~~

[kddnewton]

Yeah, I couldn't get the build to pass using isinf directly. I think mingw is non-conforming.

[kddnewton]

Unless there's a way around it, I think we need to keep it as-is.

[kddnewton]

Superseded by #3307