Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:127375
Feedback-ID: ifab94697:Fastmail
Precedence: bulk
MIME-Version: 1.0
Date: Thu, 15 May 2025 14:14:48 +0200
To: "Stephen Reay" <php-lists@koalephant.com>,
 "Andreas Hennings" <andreas@dqxtech.net>
Cc: "Volker Dusch" <volker@tideways-gmbh.com>,
 "Matthew Weier O'Phinney" <mweierophinney@gmail.com>,
 "php internals" <internals@lists.php.net>,
 =?UTF-8?Q?Tim_D=C3=BCsterhus?= <tim@bastelstu.be>
Message-ID: <74fde04e-1cbf-423c-8b50-7d7eb156a056@app.fastmail.com>
In-Reply-To: <266FA35A-15B0-435E-BBFE-1C6926EB0B7E@koalephant.com>
References: 
 <CAB-BujZLV=dccE6WjBFaLMSM977=fmJOirHSTbhJ1GuZj6iPdQ@mail.gmail.com>
 <266FA35A-15B0-435E-BBFE-1C6926EB0B7E@koalephant.com>
Subject: Re: [PHP-DEV] [RFC] Clone with v2
Content-Type: multipart/alternative;
 boundary=a1c3b7ef47ff49c7835b54d079446da8
From: rob@bottled.codes ("Rob Landers")

--a1c3b7ef47ff49c7835b54d079446da8
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

On Thu, May 15, 2025, at 13:56, Stephen Reay wrote:
>=20
>=20
>=20
> > On 15 May 2025, at 16:44, Andreas Hennings <andreas@dqxtech.net> wro=
te:
> >=20
> > =EF=BB=BFOn Thu, 15 May 2025 at 08:24, Stephen Reay <php-lists@koale=
phant.com> wrote:
> > [..]
> >>=20
> >>=20
> >> I may be missing something here..
> >>=20
> >> So far the issues are "how do we deal with a parameter for the actu=
al object, vs new properties to apply",  "should __clone be called befor=
e or after the changes" and "this won't allow regular readonly propertie=
s to be modified".
> >>=20
> >> Isn't the previous suggestion of passing the new property arguments=
 directly to the __clone method the obvious solution to all three proble=
ms?
> >=20
> > What exactly should happen then?
> > Would the __clone() method be responsible for assigning those proper=
ties?
> > Or does the __clone() method get the chance to alter the values befo=
re
> > they are assigned?
> > (this would mean they have to be passed by reference)
> > I think this last option is the best, because the values in the array
> > can be changed without any readonly constraints.
> >=20
> > Another option I was thinking of would be to call __clone() after the
> > changes are applied, and pass both the original object and the array
> > of changes as first parameter.
> > But I think this is a dead end.
> >=20
> > -- Andreas
> >=20
> >>=20
> >> There's no potential for a conflicting property name, the developer=
 can use the new property values in the order they see fit relative to t=
he logic in the __clone call, and it's inherently in scope to write to a=
ny (unlocked during __clone) readonly properties.
> >>=20
> >>=20
> >>=20
> >> Cheers
> >>=20
> >> Stephen
> >>=20
> >>=20
> >>=20
> >=20
>=20
> I would suggest that the __clone method should be directly responsible=
 for making any changes, just as it is now when it comes to deep cloning=
 or resetting values.
>=20
> Yes I realise it means developers need to then opt in and provide the =
functionality to support `clone $foo with(bar: "baz")` or whatever synta=
x is used.=20
>=20
> If the properties are public properties, there's nothing stopping some=
one writing their own clone_with() in userland now;  If someone is using=
 readonly properties I'd suggest they want to specifically manage update=
s to those properties themselves anyway.=20
>=20
> Additionally it means "clone with" would be usable for non-public prop=
erties at the discretion of the developer writing their code.=20
>=20
> The mental model is also very clear with this: copy the object in memo=
ry, and then call __clone(), with the arguments passed to the clone acti=
on - which may be none in the case of code that doesn't accept any clone=
 arguments.  The only change from the current model is that it *may* be =
passing arguments.=20
>=20
> Cheers
>=20
> Stephen=20
>=20

When I was working on Records, one of the important things was the abili=
ty to "hook into" the update that was about to happen. For example, if y=
ou have a Money type, you'd want to be able to ensure it cannot be negat=
ive when updating via `with()`. This is super important for ensuring con=
straints are met during the clone.

=E2=80=94 Rob
--a1c3b7ef47ff49c7835b54d079446da8
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html><html><head><title></title></head><body><div>On Thu, May =
15, 2025, at 13:56, Stephen Reay wrote:</div><blockquote type=3D"cite" i=
d=3D"qt" style=3D""><div><br></div><div><br></div><div><br></div><div>&g=
t; On 15 May 2025, at 16:44, Andreas Hennings &lt;<a href=3D"mailto:andr=
eas@dqxtech.net">andreas@dqxtech.net</a>&gt; wrote:</div><div>&gt;&nbsp;=
</div><div>&gt; =EF=BB=BFOn Thu, 15 May 2025 at 08:24, Stephen Reay &lt;=
<a href=3D"mailto:php-lists@koalephant.com">php-lists@koalephant.com</a>=
&gt; wrote:</div><div>&gt; [..]</div><div>&gt;&gt;&nbsp;</div><div>&gt;&=
gt;&nbsp;</div><div>&gt;&gt; I may be missing something here..</div><div=
>&gt;&gt;&nbsp;</div><div>&gt;&gt; So far the issues are "how do we deal=
 with a parameter for the actual object, vs new properties to apply",&nb=
sp; "should __clone be called before or after the changes" and "this won=
't allow regular readonly properties to be modified".</div><div>&gt;&gt;=
&nbsp;</div><div>&gt;&gt; Isn't the previous suggestion of passing the n=
ew property arguments directly to the __clone method the obvious solutio=
n to all three problems?</div><div>&gt;&nbsp;</div><div>&gt; What exactl=
y should happen then?</div><div>&gt; Would the __clone() method be respo=
nsible for assigning those properties?</div><div>&gt; Or does the __clon=
e() method get the chance to alter the values before</div><div>&gt; they=
 are assigned?</div><div>&gt; (this would mean they have to be passed by=
 reference)</div><div>&gt; I think this last option is the best, because=
 the values in the array</div><div>&gt; can be changed without any reado=
nly constraints.</div><div>&gt;&nbsp;</div><div>&gt; Another option I wa=
s thinking of would be to call __clone() after the</div><div>&gt; change=
s are applied, and pass both the original object and the array</div><div=
>&gt; of changes as first parameter.</div><div>&gt; But I think this is =
a dead end.</div><div>&gt;&nbsp;</div><div>&gt; -- Andreas</div><div>&gt=
;&nbsp;</div><div>&gt;&gt;&nbsp;</div><div>&gt;&gt; There's no potential=
 for a conflicting property name, the developer can use the new property=
 values in the order they see fit relative to the logic in the __clone c=
all, and it's inherently in scope to write to any (unlocked during __clo=
ne) readonly properties.</div><div>&gt;&gt;&nbsp;</div><div>&gt;&gt;&nbs=
p;</div><div>&gt;&gt;&nbsp;</div><div>&gt;&gt; Cheers</div><div>&gt;&gt;=
&nbsp;</div><div>&gt;&gt; Stephen</div><div>&gt;&gt;&nbsp;</div><div>&gt=
;&gt;&nbsp;</div><div>&gt;&gt;&nbsp;</div><div>&gt;&nbsp;</div><div><br>=
</div><div>I would suggest that the __clone method should be directly re=
sponsible for making any changes, just as it is now when it comes to dee=
p cloning or resetting values.</div><div><br></div><div>Yes I realise it=
 means developers need to then opt in and provide the functionality to s=
upport `clone $foo with(bar: "baz")` or whatever syntax is used.&nbsp;</=
div><div><br></div><div>If the properties are public properties, there's=
 nothing stopping someone writing their own clone_with() in userland now=
;&nbsp; If someone is using readonly properties I'd suggest they want to=
 specifically manage updates to those properties themselves anyway.&nbsp=
;</div><div><br></div><div>Additionally it means "clone with" would be u=
sable for non-public properties at the discretion of the developer writi=
ng their code.&nbsp;</div><div><br></div><div>The mental model is also v=
ery clear with this: copy the object in memory, and then call __clone(),=
 with the arguments passed to the clone action - which may be none in th=
e case of code that doesn't accept any clone arguments.&nbsp; The only c=
hange from the current model is that it *may* be passing arguments.&nbsp=
;</div><div><br></div><div>Cheers</div><div><br></div><div>Stephen&nbsp;=
</div><div><br></div></blockquote><div><br></div><div>When I was working=
 on Records, one of the important things was the ability to "hook into" =
the update that was about to happen. For example, if you have a Money ty=
pe, you'd want to be able to ensure it cannot be negative when updating =
via `with()`. This is super important for ensuring constraints are met d=
uring the clone.</div><div><br></div><div id=3D"sig121229152">=E2=80=94 =
Rob</div></body></html>
--a1c3b7ef47ff49c7835b54d079446da8--