Allow constructor property promotion with final properties

[DanielEScherzer]

Description

The following code (https://3v4l.org/t2Pra)

<?php

class Demo {
    public function __construct(
        final public string $v
    ) {}
}

$d = new Demo( "FOO" );
var_dump( $d );

$r = new ReflectionProperty( Demo::class, 'v' );
var_dump( $r->isFinal() );

Resulted in this output:

Fatal error: Cannot use the final modifier on a parameter in /in/t2Pra on line 5

Process exited with code 255.

But I expected this output instead (from https://3v4l.org/Zus4v)

object(Demo)#1 (1) {
  ["v"]=>
  string(3) "FOO"
}
bool(true)

From discussion on #17861

This came up, and there were some concerns it could imply const parameters as in Java. So we just left it out.

[iluuu1994]

As mentioned in the PR itself, this came up in the discussion and was implicitly not part of the RFC. It's still reasonable to add, but there should be a separate discussion on the ML for it. That might be enough, but people might ask for an RFC (given the feedback, and the reason why we didn't add it in the first place).

[DanielEScherzer]

As mentioned in the PR itself, this came up in the discussion and was implicitly not part of the RFC. It's still reasonable to add, but there should be a separate discussion on the ML for it. That might be enough, but people might ask for an RFC (given the feedback, and the reason why we didn't add it in the first place).

Okay, so then can this please be re-opened as a feature request? I don't have the rights to re-open the issue. I'll send an email to the mailing list in the next few days

[DanielEScherzer]

Started mailing list discussion, https://externals.io/message/126475

[DanielEScherzer]

As mentioned in the PR itself, this came up in the discussion and was implicitly not part of the RFC. It's still reasonable to add, but there should be a separate discussion on the ML for it. That might be enough, but people might ask for an RFC (given the feedback, and the reason why we didn't add it in the first place).

Okay, so then can this please be re-opened as a feature request? I don't have the rights to re-open the issue. I'll send an email to the mailing list in the next few days

[bolding added] @iluuu1994 given the lack of objections on the mailing list (no one else has asked for an RFC) are you okay with this moving forward without an RFC? If yes, can you please removed the Status: Requires RFC label (I don't have permission to edit labels). If not, can you reply on the mailing list for visibility, and I'll start writing an RFC?

[iluuu1994]

@DanielEScherzer I think the wording in the e-mail could have been more explicit. Something like:

Unless somebody objects, maintainers have agreed to merge this PR in n weeks.

That leaves no room for interpretation, and it forces people who object to actually engage. Given that there is ample time until feature freeze, do you mind sending another e-mail? 2 weeks should be enough.

[iluuu1994]

are you okay with this moving forward without an RFC?

But yes, to answer that question, I don't object to move forward without an RFC if nobody else does.

[DanielEScherzer]

Sent an update, I'll be back in 2 weeks

[iluuu1994]

Thank you @DanielEScherzer!