Bug #52706 [Com]: __sleep inhertiance

From: Date: Sun, 29 Aug 2010 11:48:55 +0000
Subject: Bug #52706 [Com]: __sleep inhertiance
References: 1  Groups: php.bugs 
Request: Send a blank email to php-bugs+get-153894@lists.php.net to get a copy of this message
Edit report at http://bugs.php.net/bug.php?id=52706&edit=1

 ID:                 52706
 Comment by:         dev dot php at linke-t dot net
 Reported by:        dev dot php at linke-t dot net
 Summary:            __sleep inhertiance
 Status:             Bogus
 Type:               Bug
 Package:            Scripting Engine problem
 Operating System:   Win7
 PHP Version:        5.3.3
 Block user comment: N

 New Comment:

I think the primary problems i have are not tested.

I wrote a class with a private property.
class foo{
	private $v;
	function foo(){
		$this->v="bar";
	}
}
And a class wich extends it:
class foo2 extends foo{
	function foo2(){
		$this->foo();
	}
	function __sleep(){
		return array("v");
	}
}

with __sleep serialize will Output: O:10:"foo2":1:{s:1:"v";N;}
without __sleep is the output like:
O:10:"foo2":1:{s:6:"foov";s:3:"bar";}

so serialize without __sleep allow acces to private properties. I think,
if foo2 doesnt have access to a private property of class foo, than
serialize didnt had it too. 

But if i want to save an object ... so i need to save it completly ...
so it have to save the private propertys as well. In C++ for as an
example i will copy the memory ... so i have copied the private
properties as all others. 


and if this isnt a bug ... it's an bad feature ... I think we need to
call it "The dont use __sleep()" feature ;) 

I hope i wont get another "its not a bug" - correct the behaivor or
wrote in the documention how to save an object with private property of
super class, thanks.

"Yes, class design has to be well thought." 
Yes, I aggree - and i think the behavior of functions should it be too.


Previous Comments:
------------------------------------------------------------------------
[2010-08-26 15:55:50] johannes@php.net

Thank you for taking the time to write to us, but this is not
a bug. Please double-check the documentation available at
http://www.php.net/manual/ and the instructions on how to
report
a bug at http://bugs.php.net/how-to-report.php

Yes, class design has to be well thought.

------------------------------------------------------------------------
[2010-08-26 12:14:10] dev dot php at linke-t dot net

Description:
------------
i the case of class "B" extends a class "A" wich define the function
__sleep(), instances of class "B" would class A::__sleep. In this case
class serialzing would be very buggy - i tried something like:
Try: let return the __sleep of class "A" all propertys wich are
currently definied
resoulting Problem: private properety of class "B" are null or NAN
Try: define the function __sleep() in class "B" 
resoulting Problem: private properety of class "A" are null or NAN


Test script:
---------------
class A{
	public $A_A = 10;
	protected $A_B = 11;
	private $A_C = 12;
	public function A(){
		
	}
	private function __sleep(){
		echo "sleep of A()";
		if(get_class($this)	!= "A")
		{
			$d = array_keys( get_class_vars(get_class($this)));
			return ($d);
		}
		return array("A_A","A_B","A_C");
	}
	public function __get($v){
		
		return "PR";
	}
}
class B extends A{
	public 		$B_A = 20;
	protected 	$B_B = 21;
	private 	$B_C = 22;
	public function B(){
		$this->A();
	}
	public function __sleep(){ 
		return
array("A_A","A_B","A_C","B_A","B_B","B_C");
	} /* remove this __sleep would let php use A::__sleep()

}
$a = new A();
$b = new B();
echo "<br/>A";
echo "<br/>" .  htmlspecialchars(serialize($a));
echo "<br/>B";
echo "<br/>" .  htmlspecialchars(serialize($b));


Expected result:
----------------
the complettely serialized Object like with out __sleep()!

Actual result:
--------------
private propertys of parent or child are null or NAN with defined
__sleep()


------------------------------------------------------------------------



-- 
Edit this bug report at http://bugs.php.net/bug.php?id=52706&edit=1


Thread (3 messages)

« previous php.bugs (#153894) next »