It's easy to forget that CGFloat may actually be a double despite it's name. On 64-bit systems, where CGFLOAT_IS_DOUBLE is set to 1 it will actually be a double.

This may become cumbersome when you are developing an application where you want a single code base which builds for both 32-bit and 64-bit such as when developing against the 32-bit simulator and deploying arm64.

CGFloatType is a simple library which provides methods on NSNumber to help deal with this. Along with providing various math and rounding functions which accepts CGFloat.

This helps you change:

#if CGFLOAT_IS_DOUBLE
  CGFloat points = ceil(pointsPerMinute * durationMinutes);
#else
  CGFloat points = ceilf(pointsPerMinute * durationMinutes);
#endif

Into something like:

CGFloat points = ceilCGFloat(pointsPerMinute * durationMinutes);

Usage

#import <CGFloatType/CGFloatType.h>

@implementation Example

- (void)example {
    NSNumber *number = [NSNumber numberWithCGFloat:CGRectGetHeight(self.view.frame)];
    NSLog(@"NSNumber with value: %@", number);
    CGFloat value = [number CGFloatValue];

    CGFloat floorValue = floorCGFloat(value);
    CGFloat ceilValue = ceilCGFloat(value);

    NSLog(@"%@ (floor = %@, ceil = %@)", @(value), @(floorValue), @(ceilValue));
}

@end

Installation

Podfile

pod 'CGFloatType'

License

CGFloatType is released under the BSD license. See LICENSE.