I don’t like checking if the caller passed me a nil NSError** in my own methods, especially if the implementation is somewhat lengthy, so I’ve put together a very basic C macro that declares an autoreleasing NSError* to use as temporary storage in case of a null NSError** parameter from the caller:
|
1 2 3 |
#define ERROR(x,y) __autoreleasing NSError * y;\ x = x == NULL ? &y : x; \ *x = nil; |
Imagine that I have a method named foo that takes a single NSString* and an optional NSError**:
|
1 2 3 4 |
-(BOOL)foo:(NSString*)someString error:(NSError**)error { ERROR(error,tempError); // do the work } |
Now I don’t need to check if I’ve been passed a valid NSError** pointer, because it will be assigned to the local tempError variable if it was nil.