MoeBrowne

This RFC needs a re-work. While I support the idea, I don't think this is the right approach. The fact that the RFC is so long is a massive red flag. For example there are 4 different ways to define hooks:

class User
{
    public string $username {
        set(string $value) {
            $this->username = strtolower($value);
        }
    }
}

class User
{
    public string $username {
        set(string $value) {
            $field = strtolower($value);
        }
    }
}

class User
{
    public string $username {
        set => strtolower($value);
    }
}

class User
{
    public function __construct(
        public string $username { set => strtolower($value); }
    ) {}
}

Why?!

This is not what PHP needs.

It feels like it adds a huge amount of complexity just to avoid getter/setter methods. Also, I'm not sure it's a good thing to lose the explicit differentiation between a property access and method call.

Note: setter methods are something I almost never use in the first place, since it's typically better to have immutable value objects.

This kind of code looks very appealing when performing very simple (and anemic) CRUD operations on a model, but it has a very short trajectory when the code gets a little more complex.

Triggering events is shown as an example of an advantage of this feature, but it's clearly a bad idea: https://3v4l.org/ZCVpG#v8.2.11

Also, if you need a centralized validation around an attribute, a value object is a way better option: https://3v4l.org/QWm8c#v8.2.11

I think that there are a ton of better requests with a lot more of value to achieve this in userland code.

Also, this idea is very overlooked, and would require a lot of internals work after to cover all the edge cases. How should child classes behave? Are they overridable? How do you know if there is a setter behavior defined already? How should traits behave? Is there a way to call the parent class setter/getter? Would that be overriden or just called by default?

Also, in response to some comments:

It could be used by ORMs like Laravel Eloquent Model that has cast and other hooks to transform data on get/set.

Good ORMs should provide mechanisms for this!

However, I actually often find the need for more fine-grained control over input and output, but adding methods feels so heavy-weight.

This is the perfect case for value objects to kick in. You want behavior (input validation, output transformation maybe) at an atomic member, that could be exactly this but atomically standalone, and therefore reusable in other cases, and very easily testable.

While I'm not against the concept in general, this implementation is not well-done. The $value variables comes from nowhere and makes it confusing - it looks like an undefined variable and a quick glance at it, was the first thing I thought - where is $value coming from? Why not just use $name within that block? Same with $field? Why not just return the value that is going to be set? This is one of the more confusing RFCs I've seen and does not follow the PHP-code style, makes things confusing for both new and experienced coders. In it's current format, I can't approve with good conscience.

I think complaints about the syntax being messy are really about the array manipulation functions being inherently hard to format nicely.

To be clear I agree that the proposed example isn't great, nested closures are never going to win any readabillity prizes. If however we look at any non-array based manipulation I think the readabillity is objectively better:

$name = 'my_user_name'
    |> fn (string $string): string => str_replace('_', ' ', $string)
    |> strtolower(...)
    |> ucwords(...)
    |> trim(...);

Or without first class callables:

$name = 'my_user_name'
    |> fn (string $string): string => str_replace('_', ' ', $string)
    |> fn (string $string): string => strtolower($string)
    |> fn (string $string): string => ucwords($string)
    |> fn (string $string): string => trim($string);

Bonus: this also adds runtime type checks to each step. Eg strreplace returns string|array

I can't imagine anyone would think this is better:

$name = trim(
    ucwords(
        strtolower(
            str_replace('_', ' ', 'my_user_name')
        )
    )
);

With First-class callable syntax available since 8.1, it would now be possible to write it as below, which is much better then string names of functions:

$result = "Hello World"
    |> htmlentities(...)
    |> str_split(...)
    |> fn($x) => array_map(strtoupper(...), $x)
    |> fn($x) => array_filter($x, fn($v) => $v != 'O');

I see no immediate benefit of the proposed solution over the userland implementations. The RFC mentions a shopping cart example, but I don't think that's cleaner than using league/pipeline or Laravel's pipeline.

It's a bit messy for the simpler examples as well.

For me, the most important argument is that the pipeline pattern is a tried and tested pattern, that this RFC builds upon. A couple of examples:

• league/pipeline
• laravel's pipeline

This RFC adds syntax to make using these kinds of pattern much more convenient.

On top of that, there's the argument that multiple modern languages support a pipe operator:

• Hack
• Haskall
• Ruby
• Elixir
• And more

Finally, I've had numerous occasions where a pipe operator would simplify my own code — I have more than a handful real life cases where this would be useful.

I've written DefaultImplementation traits to implement common interface functionality one too many times. Would love to see this as a feature.

I think the potential confusion this adds is not worth the small amount of extra code that would need to be written to use an anonymous function

mmh, seems to me that just replacing function by fn and having such a different behaviour is risky

Auto capture by-value is clear enough with a single expression closure, but I think for multiple statement functions it could be confusing to have to keep track of scope. Being explicit (with use()) makes it simple and concise.

If this were to pass, I think the syntax should continue to include the arrow to signify the auto capture by-value intention. Just like in JavaScript, the arrow is used whether single expression or multiple, and the use of brackets is what separates them.

The 'use' statement clarifies the scope for me. So a proposal like this could have the side effect of mixing scope which would lead to a confusing code.

Technically writing a trait isn't a showstopper, but it adds cognitive load because the developer now must know/remember using this Interface requires adding this Trait. PHP should focus more on development experience!

It will solve having to create traits to add a default implementation when creating interfaces and keeping it nicely together improving the DX

Creating traits for default implementation is just a pain. I want syntactic sugar

I wrote down some thoughts on this RFC on my blog. I think it's worth rethinking our current definition of what "an interface" is. Especially since many languages are interface default methods as their way of multi-inheritance.