The NSNumber class is itself a subclass of the basic NSValue class, which provides an object wrapper around a single value or data item. In addition to the basic C scalar types, NSValue can also be used to represent pointers and structures.
The NSValue class offers various factory methods to create a value with a given standard structure, which makes it easy to create an instance to represent, for example, an NSRange, like the example from earlier in the chapter:
NSString *mainString = @"This is a long string";
NSRange substringRange = [mainString rangeOfString:@"long"];
NSValue *rangeValue = [NSValue valueWithRange:substringRange];
It’s also possible to create NSValue objects to represent custom structures. If you have a particular need to use a C structure (rather than an Objective-C object) to store information, like this:
typedef struct {
int i;
float f;
} MyIntegerFloatStruct;
you can create an NSValue instance by providing a pointer to the structure as well as an encoded Objective-C type. The @encode() compiler directive is used to create the correct Objective-C type, like this:
struct MyIntegerFloatStruct aStruct;
aStruct.i = 42;
aStruct.f = 3.14;
NSValue *structValue = [NSValue value:&aStruct
withObjCType:@encode(MyIntegerFloatStruct)];
The standard C reference operator (&) is used to provide the address of aStruct for the value parameter.