Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.java.programmer > #17357 > unrolled thread

why does this work?

Back to article view | Back to comp.lang.java.programmer

Contents

  why does this work? dkoleary <dkoleary@olearycomputers.com> - 2012-08-08 10:30 -0700
    Re: why does this work? Daniel Pitts <newsgroup.nospam@virtualinfinity.net> - 2012-08-08 10:38 -0700
      Re: why does this work? dkoleary <dkoleary@olearycomputers.com> - 2012-08-08 11:39 -0700
        Re: why does this work? Lew <lewbloch@gmail.com> - 2012-08-08 17:33 -0700
    Re: why does this work? Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-08-08 14:03 -0400
    Re: why does this work? Roedy Green <see_website@mindprod.com.invalid> - 2012-08-09 02:41 -0700
      Re: why does this work? glen herrmannsfeldt <gah@ugcs.caltech.edu> - 2012-08-09 20:17 +0000

#17357 — why does this work?

Hi;

New java programmer.  So new, in fact, that I'm still working my way through the O'Reilly Head First Java book.  One of the end of chapter questions involves identifying if a sample class will compile and what to do to make it compile if it won't.  

The sample class from chapter 4 is:

class XCopy
{  public static void main(String[] args)
   {  int orig = 42;
      XCopy x = new XCopy();
      int y = x.go(orig);
      System.out.println(orig + " " + y);
   }

   int go(int arg)
   {  return arg * 2;   }
}

The book says that it'll compile and run, displaying "42 84" and, sure enough, it does:

$ javac XCopy.java
$ java XCopy
42 84

How come that isn't recursive?  XCopy.main() instantiates a new XCopy.  Shouldn't that new XCopy instance also instantiate a new XCopy?  

I was figuring this would run until the XCopy.go function tried returning a number that wouldn't fit in int anymore... That's obviously not the case, but I don't know why.

Can someone provide the missing concept?

Thanks.

Doug O'Leary

[toc] | [next] | [standalone]

#17359

On 8/8/12 10:30 AM, dkoleary wrote:
> Hi;
>
> New java programmer.  So new, in fact, that I'm still working my way through the O'Reilly Head First Java book.  One of the end of chapter questions involves identifying if a sample class will compile and what to do to make it compile if it won't.
>
> The sample class from chapter 4 is:
>
> class XCopy
> {  public static void main(String[] args)
>     {  int orig = 42;
>        XCopy x = new XCopy();
>        int y = x.go(orig);
>        System.out.println(orig + " " + y);
>     }
>
>     int go(int arg)
>     {  return arg * 2;   }
> }
>
> The book says that it'll compile and run, displaying "42 84" and, sure enough, it does:
>
> $ javac XCopy.java
> $ java XCopy
> 42 84
>
> How come that isn't recursive?  XCopy.main() instantiates a new XCopy.  Shouldn't that new XCopy instance also instantiate a new XCopy?
new XCopy() creates a new instance of the XCopy class, which executes 
Constructors.  main isn't executed again because of this.  main is only 
automatically executed by the JVM on start-up, once.
>
> I was figuring this would run until the XCopy.go function tried returning a number that wouldn't fit in int anymore... That's obviously not the case, but I don't know why.
int can fit any number in the range [-2^31, 2^31).  42 and 84 are both 
within that range.  It is possible that if you pass in a number with a 
large enough magnitude, you will end up with an overflow.  In Java (and 
many 2s-compliment integer systems), overflow will simply throw-away the 
upper bits, and you will have what is called "wrap-around".  This 
actually makes many of the basic operations easier, because signed 
numbers and unsigned numbers behave the same way.


Hopefully this helps.

[toc] | [prev] | [next] | [standalone]

#17365

On Wednesday, August 8, 2012 12:38:27 PM UTC-5, Daniel Pitts wrote:
> new XCopy() creates a new instance of the XCopy class, which executes 
> Constructors.  main isn't executed again because of this.  main is only 
> automatically executed by the JVM on start-up, once.
> 

Got it... I do remember reading that somewhere else.  main only gets executed once.  

Thanks for the concept correction.  Hopefully, it'll stick now :)

Doug O'Leary

[toc] | [prev] | [next] | [standalone]

#17441

dkoleary wrote:
> Got it... I do remember reading that somewhere else.  main only gets executed once.  

That isn't strictly true. The 'main()' method is only called for one class one time by 
the JVM when it starts, but nothing prevents code in the program from explicitly 
calling some class's 'main()' at some later time.

> Thanks for the concept correction.  Hopefully, it'll stick now :)

Read the Java tutorials. They explain constructors vs. methods and such 
concepts.

-- 
Lew

[toc] | [prev] | [next] | [standalone]

#17362

On 8/8/2012 1:30 PM, dkoleary wrote:
> [...]
> How come that isn't recursive?  XCopy.main() instantiates a new XCopy.  Shouldn't that new XCopy instance also instantiate a new XCopy?

     See Daniel Pitts' explanation.  Another thing you could try to
help you see what's going on is to sprinkle some more println()
calls through the code to help trace through the execution.  In
this case you're interested in where the constructor fits with
relation to everything else, but there's no explicit constructor
written in the code.  As I'm sure you've learned, this means the
compiler will write a simple constructor for you -- but there's
no way to get the compiler to stick println() calls in what it
writes, so your recourse is to write your own explicit constructor.
With this in mind, the code might look like:

	class XCopy {
	    public static void main(String[] args) {
	        System.out.println("entering main()");
	        int orig = 42;
	        System.out.println("main() creates an XCopy");
	        XCopy x = new XCopy();
	        System.out.println("main() created an XCopy");
	        int y = x.go(orig);
	        System.out.println(orig + " " + y);
	        System.out.println("main() is finished");
	    }

	    int go(int arg) {
	        System.out.println("executing go(), arg = " + arg);
	        return arg * 2;
	    }

	    // Explicit constructor, just for the println
	    XCopy() {
	        System.out.println("constructing an XCopy");
	    }
	}

Run this version, study the output, and see if the sequence of
events becomes clearer.

     This technique is sometimes called "printf debugging" (the name
comes from a different programming language).  Despite its simplicity,
it can be astonishingly effective, and the overall approach can be
used in most environments and most languages.  Indeed, Java's various
logging frameworks (you may learn about them later) are basically
just fancied-up versions of printf debugging: A piece of the program
blurts "Look! I'm *here*, and these are a few interesting values."

-- 
Eric Sosman
esosman@ieee-dot-org.invalid

[toc] | [prev] | [next] | [standalone]

#17450

On Wed, 8 Aug 2012 10:30:02 -0700 (PDT), dkoleary
<dkoleary@olearycomputers.com> wrote, quoted or indirectly quoted
someone who said :

>How come that isn't recursive?  

new allocates some space and XCopy() runs the initialitzer code in the
constructor.

There is no new in the constructor, so why would it be recursive?

There is no more reason for the constructor to be recursive than any
other method.  The constructor does not reinvoke the constructor.
-- 
Roedy Green Canadian Mind Products http://mindprod.com
A new scientific truth does not triumph by convincing its opponents and making them see the light,
but rather because its opponents eventually die, and a new generation grows up that is familiar with it.
~ Max Planck 1858-04-23 1947-10-04 

[toc] | [prev] | [next] | [standalone]

#17566

Roedy Green <see_website@mindprod.com.invalid> wrote:

(snip)

> There is no new in the constructor, so why would it be recursive?

> There is no more reason for the constructor to be recursive than any
> other method.  The constructor does not reinvoke the constructor.

But if there IS a new in the constructor...


-- glen

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.java.programmer

csiph-web