Bug #74488 [Com]: Just flat wrong: "Warning: Declaration of * should be compatible with"
| From: | spam2 at rhsoft dot net | Date: | Sat, 22 Apr 2017 12:23:59 +0000 |
| Subject: | Bug #74488 [Com]: Just flat wrong: "Warning: Declaration of * should be compatible with" | ||
| References: | 1 | Groups: | php.bugs |
| Request: | Send a blank email to php-bugs+get-208713@lists.php.net to get a copy of this message | ||
Edit report at https://bugs.php.net/bug.php?id=74488&edit=1
ID: 74488
Comment by: spam2 at rhsoft dot net
Reported by: appacc2040 at gmail dot com
Summary: Just flat wrong: "Warning: Declaration of * should
be compatible with"
Status: Not a bug
Type: Bug
Package: Compile Warning
PHP Version: 7.1.4
Block user comment: N
Private report: N
New Comment:
At least give us "somekeyword_stating_dontmatter class extends parent_class"
Previous Comments:
------------------------------------------------------------------------
[2017-04-22 12:15:24] spam2 at rhsoft dot net
No they are not because when someone wants Java then he can use Java
And that in case of return types the warning becomes a fatal error is a typical php inconsistency
like soap extension, require and call to disabled functions not throwing throwable exceptions
------------------------------------------------------------------------
[2017-04-22 11:58:39] nikic@php.net
You might find these helpful:
https://en.wikipedia.org/wiki/Liskov_substitution_principle
https://en.wikipedia.org/wiki/Composition_over_inheritance
------------------------------------------------------------------------
[2017-04-22 11:52:33] appacc2040 at gmail dot com
Description:
------------
"Warning: Declaration of * should be compatible with"
In the real world abstracted features work this way all the time. This warning is an arbitrary and
artificial idea that has no basis in the real world.
PHP Warning: Declaration of combinationLock::unlock($combination) should be compatible with
abstractLock::unlock() in test.php on line 23
PHP Warning: Declaration of paddleLock::unlock($key) should be compatible with
abstractLock::unlock() in test.php on line 32
Test script:
---------------
<?php
class abstractLock
{
private $locked = true;
public function unlock()
{
$this->locked=false;
}
public function lockStatus()
{
if($this->locked) return "Locked\n";
else return "Unlocked\n";
}
}
class combinationLock extends abstractLock
{
private $combination = '32-10-21'; // combination
public function unlock($combination) // unlock WITH combination
{
if($this->combination == $combination) parent::unlock();
}
}
class paddleLock extends abstractLock
{
private $key = '100,70,80,30,50,90,60,40,100'; // ridge heights
public function unlock($key) // unlock WITH key
{
if($this->key == $key) parent::unlock();
}
}
$lock1 = new paddleLock();
echo "paddleLock is " . $lock1->lockStatus();
$lock1->unlock('100,70,80,30,50,90,60,40,100');
echo "paddleLock is " . $lock1->lockStatus();
$lock2 = new combinationLock();
echo "combinationLock is " . $lock2->lockStatus();
$lock2->unlock('32-10-21');
echo "combinationLock is " . $lock2->lockStatus();
?>
Expected result:
----------------
No warning about things that happen all the time in the real world.
------------------------------------------------------------------------
--
Edit this bug report at https://bugs.php.net/bug.php?id=74488&edit=1