macOS Exercise 8-4 of Kochan 2.0

[mdeh]

"Write a Rectangle method called translate: that takes a vector Point (xv, yv) as its argument. Have it translate the rectangle's origin by the specified vector"

Not that familiar with vectors. Is the argument simply a couple of integers. And by "translate", is the implication that one is assigning a new value to an existing point?

thanks

 

[skochan]

"Write a Rectangle method called translate: that takes a vector Point (xv, yv) as its argument. Have it translate the rectangle's origin by the specified vector"

Not that familiar with vectors. Is the argument simply a couple of integers. And by "translate", is the implication that one is assigning a new value to an existing point?

thanks

The arguments are a rectangle and an XPoint object (a set of two integers). Translation simply means to add the corresponding x, y coordinates from that argument to the rectangle's origin. Hope that clears it up!

Cheers,

Steve Kochan

 

[mdeh]

The arguments are a rectangle and an XPoint object (a set of two integers). Translation simply means to add the corresponding x, y coordinates from that argument to the rectangle's origin. Hope that clears it up!

Cheers,

Steve Kochan

So the declaration should look something like this?

-(void) translate: ( XYPoint *) myPt; ?????


Thanks Steve...

 

[savar]

The arguments are a rectangle and an XPoint object (a set of two integers). Translation simply means to add the corresponding x, y coordinates from that argument to the rectangle's origin. Hope that clears it up!

Cheers,

Steve Kochan

More specifically, add the dx and the dy to each corner of the rectangle.

 

[skochan]

More specifically, add the dx and the dy to each corner of the rectangle.

Yes, that's correct. As defined in the text, a Rectangle object has an origin and a width and a height. So in this case, you'd be adding the offset vector (the argument to the translate method) to the Rectangle's origin.

Cheers,

Steve

 

[mdeh]

Yes, that's correct. As defined in the text, a Rectangle object has an origin and a width and a height. So in this case, you'd be adding the offset vector (the argument to the translate method) to the Rectangle's origin.

Cheers,

Steve

OK...got it...thank you

 

[mdeh]

OK....here is my code for the answer to the exercise. Anyones' criticism is welcome. ( Gentle please )

So, given a rectangle which as 3 ivars; height, width, and a point which again, are 2 values x, and y.
I am not sure whether one needs to test of the "non-nil" value of origin first, as it should not be called before the value of origin is set, but if it is called, then an exception is generated. ( I am not sure whether to include a small test program or not. If others start posting their answers too, then maybe we can get some sort of consensus)

-(void) translate: (XYPpoint_f *) myPt
{
	if  (! origin)
		origin = [[XYPpoint_f alloc] init];
	
	origin.x = myPt.x ;
	origin.y = myPt.y ;
	
}

 

[skochan]

OK....here is my code for the answer to the exercise. Anyones' criticism is welcome. ( Gentle please )

So, given a rectangle which as 3 ivars; height, width, and a point which again, are 2 values x, and y.
I am not sure whether one needs to test of the "non-nil" value of origin first, as it should not be called before the value of origin is set, but if it is called, then an exception is generated. ( I am not sure whether to include a small test program or not. If others start posting their answers too, then maybe we can get some sort of consensus)

-(void) translate: (XYPpoint_f *) myPt
{
	if  (! origin)
		origin = [[XYPpoint_f alloc] init];
	
	origin.x = myPt.x ;
	origin.y = myPt.y ;
	
}

You need to add the offsets to the current origin:

	origin.x += myPt.x ;
	origin.y += myPt.y ;

Cheers,

Steve Kochan

 

[mdeh]

You need to add the offsets to the current origin:

	origin.x += myPt.x ;
	origin.y += myPt.y ;

Cheers,

Steve Kochan

OOPS..of course! ( I guess that's what "add to" means ) Thanks.

 

[TotalLuck]

Is this what I'm supposed to get?

First let me say I'm glad someone else asked this question. So here's what I got.

The implementation
I used -(XYpoint *) instead of -(void). is that not correct? :
did not however check for the (! origin) (and i am not understanding why I should do that. please explain)

-(XYpoint *) translate: (XYpoint *) mypt
{
	origin.x += mypt.x;
	origin.y += mypt.y;
	
	return origin;
	
}

the main

int main(int argc, char *argv[])
{
	NSAutoreleasePool * pool = [[NSAutoreleasePool alloc]init];
	
	Rectangle *myRect = [[Rectangle alloc]init]; // create instance of a Rectangle
	XYpoint *myPoint = [[XYpoint alloc]init]; // create origin xypoint
	XYpoint *myVector = [[XYpoint alloc] init]; // create vector
	
	[myRect setwidth: 10 andheight: 80]; 
	[myPoint setX:10 andY: 80]; 
	
	
	myRect.origin = myPoint; // sets the origin of the rectangle to the coordinates in myPoint
	
	NSLog(@"The Origin of the rectangle is at (%g,%g)", myRect.origin.x, myRect.origin.y);
	
	[myVector setX:15 andY: 40]; // set a vector point 
	NSLog(@"A Vector point  at (%g,%g)", myVector.x, myVector.y);
	
	[myRect translate: myVector];
	
	NSLog(@"The Rectangles new origin is translated to (%g,%g)", myRect.origin.x, myRect.origin.y);
	
	
	[[myRect origin] release];
	[myVector release];	
	[myRect release];
	[myPoint release];
	[pool drain];
	
	
	return 0;
}

i get this output:

The Origin of the rectangle is at (10,80)
A Vector point at (15,40)
The Rectangles new origin is translated to (25,120)

this compiles and runs fine.
what should i have gotten?

 

[skochan]

I used -(XYpoint *) instead of -(void). is that not correct?

It depends on how you (or others who will use your Rectangle class) will use the translate method in expressions. If the translated origin might be used direclty in an expression after the method call, then you would declare the method as you did: -(XYPoint *) and return the origin from the method (also as you did). This would be a more flexible implementation than not returning a value and declaring the method as void.

I did not however check for the (! origin) (and i am not understanding why I should do that. please explain)

This is defensive programming. It's possible that someone could call the translate method without having first set the Rectangle's origin (it's also possible that someone could call the translate method without even first allocating or initializing a Rectangle object).

One of the good things about Objective-C is that sending a message to a nil object has no effect and is also considered "OK" programming practice. Thus, writing a statement like

origin.x += mypt.x;

works even if origin is nil (it has no effect).

Again, the correct approach here depends on how you want your Rectangle class to work. If you write a translate method you could decide that it's a catastrophic error if it's called with a nil origin and throw an exception; or you could allocate an origin object and set it to (0, 0); or you might decide to just add the offset vector to a possibly nil origin object, knowing that it will effectively do nothing if the origin is in fact nil. The choice is a design consideration and the behavior should be documented with the method.

i get this output:

The Origin of the rectangle is at (10,80)
A Vector point at (15,40)
The Rectangles new origin is translated to (25,120)

this compiles and runs fine.
what should i have gotten?

This output is correct.

Cheers,



Steve Kochan