Immutable modifier View as Markdown

On 16 November 2015 at 09:33, Lorenzo Fontana fontanalorenz@gmail.com
wrote:

I really like the concept of immutability, but I think that it should be
applicable at instance level rather than declaration.

I would also prefer another keyword than immutable.

Final does not make the properties immutable, it makes the class not
extensible.

Any differance from the final keyword?

http://php.net/manual/en/language.oop5.final.php

On Mon, Nov 16, 2015 at 10:15 AM, Chris Riley t.carnage@gmail.com
wrote:

Hi,

There has been a lot of interest recently (eg psr-7) in immutable data.
I'm
considering putting an RFC together to add language support for
immutables:

immutable class Foo {
public $bar;
public function __construct($bar) {
$this->bar = $bar;
}
}

Immutable on a class declaration makes all (maybe only public?)
properties
of the class immutable after construct; assigning to a property would
result in a Fatal error.

class Foo {
public $bar;
immutable public $baz;
}

Immutable on a property makes the property immutable once it takes on a
none null value. Attempts to modify the property after this results in a
fatal error.

Any thoughts?
~C

What instance level syntax would you propose?

$foo = new immutable bar(); ?

or

immutable $foo = new bar();

Gives less flexibility imo and less guarantees of correctness.

Consider the current user land implementation of immutable classes:

class Foo {
private $bar;
public function __construct($bar) {
$this->bar = $bar;
}

public function getBar() {
return $this->bar;
}
}

Is already done at declaration declaration based immutability is probably
more desirable.

What keyword would you suggest other than immutable?

~C

hi,

I really like the concept of immutability, but I think that it should be
applicable at instance level rather than declaration.
I'm not sure about that, most PHP code assume mutable objects, I doubt
much existing classes would be usable as immutable

I would also prefer another keyword than immutable.

Final does not make the properties immutable, it makes the class not
extensible.

Any differance from the final keyword?

http://php.net/manual/en/language.oop5.final.php

Hi,

There has been a lot of interest recently (eg psr-7) in immutable data.
I'm
considering putting an RFC together to add language support for
immutables:
immutable class Foo {
public $bar;
public function __construct($bar) {
$this->bar = $bar;
}
}

Immutable on a class declaration makes all (maybe only public?)
properties
I don't think it's enough to make only public properties immutable. this
would lead to a lot of mutable "immutable" classes, eg:

immutable class my_class {
private $my_prop;
__construct($my_prop) {
$this->my_prop = $my_prop;
}
public function getMyProp() {
return $this->my_prop;
}
public function setMyProp($my_prop) {
return $this->my_prop = $my_prop;
}
}

how is this class immutable ?
also, what about __isset/__set ? Would it be allowed to assign non
immutable types to properties of an immutable class ?

of the class immutable after construct; assigning to a property would
result in a Fatal error.

class Foo {
public $bar;
immutable public $baz;
}

Immutable on a property makes the property immutable once it takes on a
none null value. Attempts to modify the property after this results in a
fatal error.

Any thoughts?
~C

Immutable on a property makes the property immutable once it takes on a
none null value. Attempts to modify the property after this results
in a
fatal error.

Any thoughts?
~C

As a firm believer in the value of immutable data structures, I don't
believe such a simple approach would be useful and would likely be
counter-productive.
agreed

The trick is there are 2 cases for immutability: An immutable service
object, and an immutable data object.

A service object should be immutable once it's setup. However, that
setup may, in some cases, involve setter injection. That's inferior
or constructor injection but that is not always feasible. In
practice, I'm not sure we need new syntax here at all.
I'm curious here, why is construction not always enough ? do you have
a use case in mind ?

There are various cases, like a logger, a container-aware service (such
as factories), or other cases where you may want a class to opt-in to a
specific dependency by interface. In that case your object creation
process (DI Container or otherwise) can do something like this:

$object = create_object_via_constructor_injection(...);
if ($object instanceof Loggable) {
$object->setLogger($this->getSystemLogger());
}

There are also cases in, eg, Drupal where forcing constructor injection
for a class that is extending another results in unnecessary and
annoying boilerplate. (I'm thinking of our new Plugin system.) As I
said, it's suboptimal but use cases for setter injection (really
interface injection) do exist, so a language-strict enforcement of
constructor-or-nothing are not always applicable.

In any event, I don't think service objects really need this feature
anyway as there's no data on the service to access; if it had any, it
wouldn't be a service object. :-)

For data objects, this is where it gets interesting. Data objects
do need to be modifiable for a given context... for that context.
Immutable data objects are, largely, useless and in my experience
harmful unless they have a ->giveNewVersionWithThisOneChange()
method. In PSR-7, that's the with*() methods. DateTimeImmutable has
the same methods as DateTime, but they return new instances rather
than modifying the invoked object (despite their confusing names,
which are like that for historical reasons).
How does that makes them mutable in a "given context" ? the with*()
can always create a new object with everything in the constructor, am
I missing something ?

The "everything in the constructor" is the problem. That results in,
essentially, an obscenely long function call that just happens to be
named __construct(). If I wanted something that obscure and hard to
work with I'd just use anonymous arrays. :-)

Consider PSR-7's ServerRequest object. It's immutable. However, when
passing it from one middleware to the next you want to be able to pass a
"modified" version of it, containing extra properties that indicate the
result of routing, for example. Or you have a middleware that does
cookie encryption, so it needs to "modify" the cookie header.

If that meant each middleware needed to do something like this:

function routing_middleware(ServerRequest $r, callable $next) {
$attributes = $r->getAttributes();
$attributes['controller'] = 'foo';

$new = new ServerRequest($r->getMethod(), $r->getUri(),
$r->getHeaders(), $r->getBody(), $r->getAttributes());

return $next($new);
}

Well, you can see how gross that is, and it gets worse as soon as you
need to modify more than one property. You essentially need to extract()
the object back to primitives and build a new one, which is more
cumbersome, more verbose, slower, and makes kittens cry. At this point
it's easier to just have a big anonymous array and pass that around;
it's no less self-documenting and easier to work with.

(I had to work with such an object once when doing a Symfony REST
project, because that's what the HATEOAS bundle did. It was impossible
to work with.)

Instead, with the with*() methods, PSR-7 lets you do this:

function routing_middleware(ServerRequest $r, callable $next) {
$new = $r->withAttribute('controller', 'foo');

return $next($new);
}

Or even inline that to a single line (if you really do have a hard coded
controller, which is unlikely but other cases like cookie encryption might).

function cookie_middleware(ServerRequest $r, callable $next) {
return $next($r->withHeader('Cookie',
decode_cookie($r->getHeader('Cookie')));
}

Which looks nice and functional, too.

It's much easier to read, much easier to write, still communicates
intent, and thanks to copy-on-write is no more expensive under usual usage.

Most of the objection to immutable PSR-7 objects, including my own, was
based on the assumption that we'd be stuck with the former example,
which is a major PITA. The development of the with*() methods is what
made immutable PSR-7 feasible.

In less generic form, imagine if DateTimeImmutable worked like this:

$d = new DateTimeImmutable();
$d2 = new DateTimeImmutable($d->format('c') . ' + 1 day');

Instead of like this:

$d2 = $d->modify('+1 day');

I think you'll agree the second is much nicer.

In all of these cases, though, we get the benefit of locality of
effect. Changes to an object in one function cannot impact another
function, because it's not a change but a new object.

My ideal would be to revive the property RFC and leverage
per-property access control separately for read and write to create
externally-immutable objects. That would allow for with*() methods
to be implemented as appropriate for a given object.
are you talking about this one ?
https://wiki.php.net/rfc/readonly_properties

No, I meant this earlier one:

https://wiki.php.net/rfc/propertygetsetsyntax-v1.2

I dislike the readonly keyword for basically all the same reasons I gave
earlier in this thread. :-)

--
--Larry Garfield

Hello, internals!

We discussed this topic early, see this thread:
http://www.serverphorums.com/read.php?7,1123371,1124223 for any additional
thoughts and ideas. I want to attach a link to the Java draft of partial
immutability of objects:
http://cr.openjdk.java.net/~jrose/values/values-0.html, it's pretty
interesting and can be applied for PHP too.

I very like the whole idea of having native immutability for objects and
see a many cases where it can be applied.

I'm thinking about immutability for objects as passing variables as a copy
by value, not by reference. Const keyword can be used for that:

class Foo {
public static function bar(const Baz $object) { // Require an
instance of Baz to be passed as a copy, not a reference
$object->test = '12345'; // Will throw an exception, because for
immutable object property fetching for writing is not allowed
$object = 12345; // Will throw an exception, immutable variable
can not modified

     $myObject = $object; // Create our local copy of immutable object,

this will be a full copy (clone)
$myObject->test = '456'; // OK
echo $myObject->test, ' original is ', $object->test; // Will
output: '456 original is foo'
}
}

$obj = new Baz; // Our local instance, #1
$obj->test = 'foo';
Foo::bar($obj); // Method requires immutability, so a new copy of $obj will
be created before passing value to the method

echo $obj->test; // Outputs 'foo'

Here, an original object will be preserved in any case, because a copy will
be used (probably, via cloning) and engine will check some restrictions to
keep variable immutability.

Some more code examples:

function foo() {
const $bar = new stdClass; // We use const keyword with variable name
to declare it immutable
$bar = 1; // Will throw an exception, because $bar is immutable const
variable
$bar->test = 123; // Will throw an exception, because for immutable
variable which is object property fetch for writing is not allowed

const $immutableArray = [1, 2, 3];
$immutableArray[] = 4; // Immutable array can not be modified, only

read operation is supported

$mutableArray = $immutableArray; // If we make a copy of variable

without const, it can be changed later via Copy-on-Write
$mutableArray[] = 5; // OK, local copy is used
}

Function or methods can declare a return result to be immutable by adding a
"const" keyword into the declaration:

const function foo() : Baz // declare our function to return an immutable
copy of object
{
$object = new Baz;
$object->test = 456;
return $object;
}

$value = foo(); // Will throw an exception, immutable return result from
the function can be assigned only to the immutable variable
const $immutableValue = foo(); // OK, immutable result is stored in the
immutable variable

ReflectionParameter, ReflectionFunctionAbstract can have an additional
methods: ReflectionParameter->isImmutable() or
RelectionParameter->isConst();
ReflectionFunctionAbstract->returnsImmutableValue()

2015-11-17 10:18 GMT+03:00 Chris Riley t.carnage@gmail.com:

Hi,

Chris Riley wrote:

There has been a lot of interest recently (eg psr-7) in immutable data.
I'm
considering putting an RFC together to add language support for
immutables:

I wonder if immutable classes are really the right way to go. Immutable
reference types with manual copying are somewhat alien to PHP: instead,
the language has used copy-on-write mutable value types for arrays and
strings, and prior to PHP 5, objects.

Copy-on-write value types have all the benefits immutable types do. They
can't be mutated at a distance unless you make a reference to them,
trees made of them can be compared for equality with just ===, etc.

But they also have benefits that immutable types lack. They're easier to
work with, because you can directly mutate them like everyone's used to
doing: no messing around with withFooBar() and withoutFooBar(), you just
set foobar or remove it directly. And PHP handles the duplication of
objects for you implicitly, so there's far less code to write.

Unfortunately, PHP 5 got rid of PHP 4's value type objects and replaced
them with reference type objects. But we could always introduce a way to
get value type objects again. How about a struct keyword? It would be
equivalent to class, but produce a value type rather than a reference
type.

Any thoughts?

Thanks.

Andrea Faulds
http:// http://ajf.me/ajf.me/ http://ajf.me/

--

To unsubscribe, visit: http:// http://www.php.net/unsub.phpwww.php.net
http://www.php.net/unsub.php/ http://www.php.net/unsub.phpunsub.php
http://www.php.net/unsub.php

My main motivation for this was for event objects when doing event sourcing

• they are by definition unchangeable after creation. That said,
  considering the wider use case there does seem to be a need to emulate the
  with* functions.

One option would be to hide the messy constructor call within user defined
methods, but that would add a lot of boilerplate - something I was wanting
to reduce not increase.

I can't think of a clean easy way to add with* functionality at a language
level without even more magic (parameter skipping?)

As for setter injection - my initial proposal would support that - a
property declared immutable allows itself to be set - once.

If someone can give my wiki account rfc karma I can write this up so far
too help focus discussion.

~C