方法addTarget：action：forControlEvents：如何编程？

Question

In UIKit, UIControl has the following methods:

• addTarget:action:forControlEvents:
• removeTarget:action:forControlEvents:

I am interested to know how these are programmed. Could someone point to a resource of how these methods are programmed (or provide an example how these are programmed)?

I have implemented something similar for my own project that is working. However it would be great to know the best technique to go about this, or how Apple have implemented it, as it is a useful pattern. (And, if I am correct, differs from a delegate pattern?).

Answer 1

It's pretty simple. Here's one way you could imagine it working:

@interface UIControlTargetActionPair
@property (nonatomic, assign) id target;
@property (nonatomic, assign) SEL action
@end

(with an @implementation to just synthesize those properties)

Then:

- (void)addTarget:(id)target action:(SEL)action forControlEvents:(UIControlEvents)events {
  if (events & UIControlEventTouchDown) {
    UIControlTargetActionPair *pair = [[UIControlTargetActionPair alloc] init];
    [pair setTarget:target];
    [pair setAction:action];
    [[self mutableArrayOfTargetActionPairsForControlEvent:UIControlEventTouchDown] addObject:pair];
    [pair release];
  }
  else if (....) {
    ...
  }
  // repeat for all relevant control events
}

Removal would be to simply remote the pair object from the appropriate array, and sending the events would simply be something like this:

- (void)sendActionsForControlEvents:(UIControlEvents)controlEvents {
  NSArray *pairs = [self mutableArrayOfTargetActionPairsForControlEvent:controlEvents];
  for (UIControlTargetActionPair *pair in pairs) {
    [[pair target] performSelector:[pair action] withObject:self withObject:theCurrentEvent];
  }
}

Of course, that's not the actual code. But having the code here in front of me, it is the general principle. :)