RFC: about class names as values

14 years ago by Martin Scotta — view source

unread

Hi all,

PHP uses class names as values, but only in a couple of places.
String constanst after the "new" or "instanceof" operator --correct me if
I'm missing others-- are treated as class names.

if( $obj instanceof MyClass ) {
// $obj is instanceof Bar/Foo/MyClass
}

but what happen if the class name is used on other places?

function newInstance($class) {
return new $class;
}

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

so you have to end up by writting....

$obj = newInstance( '\Bar\Foo\MyClass' );

which lacks of any semantic, it's not a class name, it's just a string.

My idea is that PHP could include a constant on each class, a string with
the full qualified class name

echo MyClass::CLASS; // \Bar\Foo\MyClass
$obj = newInstance( MyClass::CLASS );

As the "CLASS" is currently a reserved word is guarantied that nobody is
using as it today.

What do you think about?

Martin Scotta

14 years ago by johannes@php.net — view source

unread

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

This describes the major change with your idea.

What happens if a constant MyClass exists?

Another question is something like this:

<?php
function factory($class) {
return new $class();
}

factory( SomeClass );
?>

To proper support this we'd have to make classes first class elements.
For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

johannes

14 years ago by John LeSueur — view source

unread

2011/1/5 Johannes Schlüter johannes@php.net

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

This describes the major change with your idea.

What happens if a constant MyClass exists?

Another question is something like this:

<?php
function factory($class) {
return new $class();
}

factory( SomeClass );
?>

To proper support this we'd have to make classes first class elements.
For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

johannes

--

I think he's actually proposing creating for each class the magic class
constant CLASS, so your example becomes:

<?php

function factory($class) {
return new $class();
}

factory( SomeClass::CLASS );

?>

This is actually doable without a magic class constant, but requires a
function or class constant to be declared in each class.

<?php
class SomeClass {
const CLASS = NAMESPACE . '' . CLASS;
static function getNameWithNSPath()
{
return NAMESPACE . '' . CLASS;
}
}

factory( SomeClass::getNameWithNSPath() );
?>

Perhaps this could be simplified with traits, if NAMESPACE and CLASS
work in traits that way. In fact, that's an interesting question, what is
NAMESPACE in a trait defined in one namespace, then used in a class in a
different namespace?

I think the point is that the factory function could exist without any
knowledge of the namespaces of the classes it would work on. Then, somewhere
else where the class has been aliased or is otherwise accessible without the
full namespace path, the developer wouldn't need to specify the full
namespace path to the factory, but could ask the class itself what it's full
namespace path was. I don't know that I agree with the idea, but I don't
think it requires making classes first class elements.

John

14 years ago by Martin Scotta — view source

unread

Yes, my intention was to only add a magic constant with the class, similar
to this

namespace Bar {
class Foo {
const KLASS = CLASS;
}
}

namespace Buzz {
use \Bar\Foo as BazFoo;

class Bar extends BazFoo {
const KLASS = CLASS;
}

$bar = new Bar;
$baz = new BazFoo;

var_dump( get_class($baz), BazFoo::KLASS);
var_dump( get_class($bar), Bar::KLASS );
}

This is 100% valid PHP 5.3.3 code, but that includes a lot of effort from
the developer. Someone miss to include the KLASS constant on a class and the
result is undefined.

If that PHP could add a magic constant --named CLASS or whatever you like--
to each class it will reduce the amount of class names hardcoded onto
strings, probably to zero.

The only issue that I found today is related to interfaces. I'm not sure if
they should include this sort of magic constant, but I would rather include
them just for consistency but, as I previously said, I'm not sure about this
one.

Martin Scotta

2011/1/5 John LeSueur john.lesueur@gmail.com

2011/1/5 Johannes Schlüter johannes@php.net

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

This describes the major change with your idea.

What happens if a constant MyClass exists?

Another question is something like this:

<?php
function factory($class) {
return new $class();
}

factory( SomeClass );
?>

To proper support this we'd have to make classes first class elements.
For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

johannes

--

I think he's actually proposing creating for each class the magic class
constant CLASS, so your example becomes:

<?php

function factory($class) {
return new $class();
}

factory( SomeClass::CLASS );

?>

This is actually doable without a magic class constant, but requires a
function or class constant to be declared in each class.

<?php
class SomeClass {
const CLASS = NAMESPACE . '' . CLASS;
static function getNameWithNSPath()
{
return NAMESPACE . '' . CLASS;
}
}

factory( SomeClass::getNameWithNSPath() );
?>

Perhaps this could be simplified with traits, if NAMESPACE and
CLASS work in traits that way. In fact, that's an interesting question,
what is NAMESPACE in a trait defined in one namespace, then used in a
class in a different namespace?

I think the point is that the factory function could exist without any
knowledge of the namespaces of the classes it would work on. Then, somewhere
else where the class has been aliased or is otherwise accessible without the
full namespace path, the developer wouldn't need to specify the full
namespace path to the factory, but could ask the class itself what it's full
namespace path was. I don't know that I agree with the idea, but I don't
think it requires making classes first class elements.

John

14 years ago by Adi Nita — view source

unread

Hi,

To proper support this we'd have to make classes first class elements.

For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

Closures are first-class citizens and it seems natural that they are so.
Perhaps we should rethink what consistency means for PHP. I don't see any
good reason why first-class objects could be evil, except that the $ could
indeed be a problem.

Regards,
Adrian

2011/1/6 Johannes Schlüter johannes@php.net

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

This describes the major change with your idea.

What happens if a constant MyClass exists?

Another question is something like this:

<?php
function factory($class) {
return new $class();
}

factory( SomeClass );
?>

To proper support this we'd have to make classes first class elements.
For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

johannes

14 years ago by Martin Scotta — view source

unread

Hi all,

I've wrote a simple snippet to show the importance of this feature.
This little piece of code, at a glance, tries to instantiate 2 same objects
by different ways, but fails because of a silently code bug.

-- start of code
<?php
namespace B\C {
class Foo {
function test() {
echo CLASS, PHP_EOL;
}
}
}

namespace A\B\C {
class Foo {
function test() {
echo CLASS, PHP_EOL;
}
}
}

namespace A {
$f1 = new B\C\Foo;
$f1->test();

$class = 'B\C\Foo';
$f2 = new $class;
$f2->test();

}
?>
--- end of code

This works this way because class names in strings are treated as fully
qualified class names.

Although this script was created to reproduce an programmer error it helps
to show the need for a safe way to pass class name as a values.

As a side note, '\A\B\C\Foo' refers to the same class that 'A\B\C\Foo', but
the string comparison will return false

Regards,
Martin Scotta

On Thu, Jan 6, 2011 at 11:47 AM, Michael Morris dmgx.michael@gmail.comwrote:

+1 to this. In a similar vein (and similar reasons) consider ::PARENT as
well. Since parent is also a reserved word here too is a case where no code
in existence will be in conflict. So

class Foo {}
class Bar extends Foo{}

echo Bar::PARENT; // "Foo"

2011/1/6 Martin Scotta martinscotta@gmail.com

Yes, my intention was to only add a magic constant with the class, similar

to this

namespace Bar {
class Foo {
const KLASS = CLASS;
}
}

namespace Buzz {
use \Bar\Foo as BazFoo;

class Bar extends BazFoo {
const KLASS = CLASS;
}

$bar = new Bar;
$baz = new BazFoo;

var_dump( get_class($baz), BazFoo::KLASS);
var_dump( get_class($bar), Bar::KLASS );
}

This is 100% valid PHP 5.3.3 code, but that includes a lot of effort from
the developer. Someone miss to include the KLASS constant on a class and
the
result is undefined.

If that PHP could add a magic constant --named CLASS or whatever you
like--
to each class it will reduce the amount of class names hardcoded onto
strings, probably to zero.

The only issue that I found today is related to interfaces. I'm not sure
if
they should include this sort of magic constant, but I would rather
include
them just for consistency but, as I previously said, I'm not sure about
this
one.

Martin Scotta

2011/1/5 John LeSueur john.lesueur@gmail.com

2011/1/5 Johannes Schlüter johannes@php.net

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

This describes the major change with your idea.

What happens if a constant MyClass exists?

Another question is something like this:

<?php
function factory($class) {
return new $class();
}

factory( SomeClass );
?>

To proper support this we'd have to make classes first class elements.
For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

johannes

--

I think he's actually proposing creating for each class the magic class
constant CLASS, so your example becomes:

<?php

function factory($class) {
return new $class();
}

factory( SomeClass::CLASS );

?>

This is actually doable without a magic class constant, but requires a
function or class constant to be declared in each class.

<?php
class SomeClass {
const CLASS = NAMESPACE . '' . CLASS;
static function getNameWithNSPath()
{
return NAMESPACE . '' . CLASS;
}
}

factory( SomeClass::getNameWithNSPath() );
?>

Perhaps this could be simplified with traits, if NAMESPACE and
CLASS work in traits that way. In fact, that's an interesting
question,
what is NAMESPACE in a trait defined in one namespace, then used in
a
class in a different namespace?

I think the point is that the factory function could exist without any
knowledge of the namespaces of the classes it would work on. Then,
somewhere
else where the class has been aliased or is otherwise accessible without
the
full namespace path, the developer wouldn't need to specify the full
namespace path to the factory, but could ask the class itself what it's
full
namespace path was. I don't know that I agree with the idea, but I don't
think it requires making classes first class elements.

John

14 years ago by Martin Vium — view source

unread

I think adding a magic constant or method for getting the class name would
be usefull in many scenarios, when referencing a specific class (e.g.
factories, configurations). It would also work well with namespaces and
refactoring tools e.g.:

$mock = $this->getMock('\My\Custom\Namespace\MyClass');

vs.

use My\Custom\Namespace\MyClass;
$mock = $this->getMock(MyClass::CLASS);

I think doing something like this is a good idea for classes and
interfaces.

Ben.

Yes, my intention was to only add a magic constant with the class, similar
to this

namespace Bar {
class Foo {
const KLASS = CLASS;
}
}

namespace Buzz {
use \Bar\Foo as BazFoo;

class Bar extends BazFoo {
const KLASS = CLASS;
}

$bar = new Bar;
$baz = new BazFoo;

var_dump( get_class($baz), BazFoo::KLASS);
var_dump( get_class($bar), Bar::KLASS );
}

This is 100% valid PHP 5.3.3 code, but that includes a lot of effort from
the developer. Someone miss to include the KLASS constant on a class and
the
result is undefined.

If that PHP could add a magic constant --named CLASS or whatever you
like--
to each class it will reduce the amount of class names hardcoded onto
strings, probably to zero.

The only issue that I found today is related to interfaces. I'm not sure
if
they should include this sort of magic constant, but I would rather
include
them just for consistency but, as I previously said, I'm not sure about
this
one.

Martin Scotta

2011/1/5 John LeSueurjohn.lesueur@gmail.com

2011/1/5 Johannes Schlüterjohannes@php.net

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

This describes the major change with your idea.

What happens if a constant MyClass exists?

Another question is something like this:

<?php
function factory($class) {
return new $class();
}

factory( SomeClass );
?>

To proper support this we'd have to make classes first class elements.
For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

johannes

--

I think he's actually proposing creating for each class the magic class
constant CLASS, so your example becomes:

<?php

function factory($class) {
return new $class();
}

factory( SomeClass::CLASS );

?>

This is actually doable without a magic class constant, but requires a
function or class constant to be declared in each class.

<?php
class SomeClass {
const CLASS = NAMESPACE . '' . CLASS;
static function getNameWithNSPath()
{
return NAMESPACE . '' . CLASS;
}
}

factory( SomeClass::getNameWithNSPath() );
?>

Perhaps this could be simplified with traits, if NAMESPACE and
CLASS work in traits that way. In fact, that's an interesting
question,
what is NAMESPACE in a trait defined in one namespace, then used in a
class in a different namespace?

I think the point is that the factory function could exist without any
knowledge of the namespaces of the classes it would work on. Then,
somewhere
else where the class has been aliased or is otherwise accessible without
the
full namespace path, the developer wouldn't need to specify the full
namespace path to the factory, but could ask the class itself what it's
full
namespace path was. I don't know that I agree with the idea, but I don't
think it requires making classes first class elements.

John

--

--
Mvh
Martin Vium
Senior System Arkitekt
Sitevision

14 years ago by Martin Scotta — view source

unread

Hi all,

I don't know how the internal development process of PHP works.

First at all: was this feature approved?

if that is a "yes"...
is this feature going to be scheduled for some release?
Is it supposed that I will submit a patch?

Thanks you all,
Martin Scotta

I think adding a magic constant or method for getting the class name would
be usefull in many scenarios, when referencing a specific class (e.g.
factories, configurations). It would also work well with namespaces and
refactoring tools e.g.:

$mock = $this->getMock('\My\Custom\Namespace\MyClass');

vs.

use My\Custom\Namespace\MyClass;
$mock = $this->getMock(MyClass::CLASS);

On 8 January 2011 11:21, Ben Schmidt mail_ben_schmidt@yahoo.com.au
wrote:

I think doing something like this is a good idea for classes and
interfaces.

Ben.

Yes, my intention was to only add a magic constant with the class,
similar
to this

namespace Bar {
class Foo {
const KLASS = CLASS;
}
}

namespace Buzz {
use \Bar\Foo as BazFoo;

class Bar extends BazFoo {
const KLASS = CLASS;
}

$bar = new Bar;
$baz = new BazFoo;

var_dump( get_class($baz), BazFoo::KLASS);
var_dump( get_class($bar), Bar::KLASS );
}

This is 100% valid PHP 5.3.3 code, but that includes a lot of effort
from
the developer. Someone miss to include the KLASS constant on a class and
the
result is undefined.

If that PHP could add a magic constant --named CLASS or whatever you
like--
to each class it will reduce the amount of class names hardcoded onto
strings, probably to zero.

The only issue that I found today is related to interfaces. I'm not sure
if
they should include this sort of magic constant, but I would rather
include
them just for consistency but, as I previously said, I'm not sure about
this
one.

Martin Scotta

2011/1/5 John LeSueurjohn.lesueur@gmail.com

2011/1/5 Johannes Schlüterjohannes@php.net

$obj = newInstance( MyClass ); // notice. undefined constant MyClass

This describes the major change with your idea.

What happens if a constant MyClass exists?

Another question is something like this:

<?php
function factory($class) {
return new $class();
}

factory( SomeClass );
?>

To proper support this we'd have to make classes first class elements.
For making this consistent it would make sense to make functions first
class elements. And best drop the $ in front of variables and create a
new language. Everything else becomes a mess.

johannes

--

I think he's actually proposing creating for each class the magic
class
constant CLASS, so your example becomes:

<?php

function factory($class) {
return new $class();
}

factory( SomeClass::CLASS );

?>

This is actually doable without a magic class constant, but requires a
function or class constant to be declared in each class.

<?php
class SomeClass {
const CLASS = NAMESPACE . '' . CLASS;
static function getNameWithNSPath()
{
return NAMESPACE . '' . CLASS;
}
}

factory( SomeClass::getNameWithNSPath() );
?>

Perhaps this could be simplified with traits, if NAMESPACE and
CLASS work in traits that way. In fact, that's an interesting
question,
what is NAMESPACE in a trait defined in one namespace, then used in
a
class in a different namespace?

I think the point is that the factory function could exist without any
knowledge of the namespaces of the classes it would work on. Then,
somewhere
else where the class has been aliased or is otherwise accessible
without
the
full namespace path, the developer wouldn't need to specify the full
namespace path to the factory, but could ask the class itself what it's
full
namespace path was. I don't know that I agree with the idea, but I
don't
think it requires making classes first class elements.

John

--

--
Mvh
Martin Vium
Senior System Arkitekt
Sitevision

14 years ago by Chad Fulton — view source

unread

Take this with a grain of salt, but:

Hi all,

I don't know how the internal development process of PHP works.

First at all: was this feature approved?

From my experience, don't consider something "approved" until it's in
SVN, and even then maybe not (e.g. type hinting...), regardless of how
many people say "I think that sounds like a good idea".

if that is a "yes"...
is this feature going to be scheduled for some release?

See below.

Is it supposed that I will submit a patch?

Mostly, the burden is on the person requesting a feature to submit a
patch, if for no other reason than that everyone is busy and that, as
the requester, you probably have the most interest in getting the
feature implemented.

Once a patch is out there, you're going to need to convince at least
one core dev to get behind your idea, since they are in the position
of getting it into SVN, and they have the best big picture view of
PHP. The core devs do often seem happy to help improve patches, if
they think they're worthwhile.

Then, perhaps it will get into SVN if a core dev likes it and no core
devs oppose it too much.
Then, perhaps it will be scheduled for a release.

Thanks you all,
Martin Scotta