Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:111731
To: Dik Takken <dik.takken@gmail.com>, php internals <internals@lists.php.net>
Thread-Topic: [PHP-DEV] Proposal: Adding functions any(iterable $input,
 ?callable $cb = null, int $use_flags=0) and all(...)
Thread-Index: AQHWfj2/bRooGr39CkeRSgZBGY4s+alRB8IAgAAT9Mw=
Date: Sun, 30 Aug 2020 20:43:03 +0000
Message-ID:
 <DM6PR07MB66183421EB52AD2928C6C365F9500@DM6PR07MB6618.namprd07.prod.outlook.com>
References:
 <DM6PR07MB66188EBF51D1BED2560BC889F9530@DM6PR07MB6618.namprd07.prod.outlook.com>,<89d77c3c-593b-b5dd-f39c-009575e295a5@gmail.com>
In-Reply-To: <89d77c3c-593b-b5dd-f39c-009575e295a5@gmail.com>
Accept-Language: en-CA, en-US
Content-Language: en-CA
x-ms-exchange-messagesentrepresentingtype: 1
x-tmn: [OPbw1OEy8wzh8rfRz58+1aQbCH4qsrvo+Mqh0yY4LSrEKB97LmB7zTAegOE1IJfa]
x-ms-publictraffictype: Email
x-incomingheadercount: 44
x-eopattributedmessage: 0
x-ms-office365-filtering-correlation-id: 8cd3bd63-d2c7-4153-7a8e-08d84d254acb
x-ms-traffictypediagnostic: BN3NAM04HT102:
x-microsoft-antispam: BCL:0;
x-microsoft-antispam-message-info:
 VpthplQQjnzWlJEEgDaCoYaz3jI7EnoUfHxD4i/ENzlTnzBMjUiuR4BI+nuQLDMmJfpK+3Ed3C2Kn7oEKtJLonjXKBroFzE7IbN8KzBxPkhwD6YrYIFpZj2bJDg17L4uMyaeOHa3bs065veIsKtHjBVy3dHdr7tBPJo1DLPpPuD7M/N76iqHdYp7ak7PZq212aR8BfB4QLwNv5+daKBzUXVG2NZRkXS7nJgm71/PlXOuvll4KHt7pU6d2MMXHVqN
x-ms-exchange-antispam-messagedata:
 NiuGb5fc4q2fV/eB0AG/rHjepcBJMJyXldhU2/6oFxfKSOYUNFrLIQnXrpRIS++Rj6BysguSk+wv1Lskekk5KZ2XGqoR09qOJCOQ4OxrNFh7MLCndT9YNpsqdBGsydNhgoeRw28Ho4YF5PUcYfjI3Tb4b73sVqpu74MPcFrXTaXnpqdYfTlE5dJaawbx0F3GFLwB3J+PVJLZ7+tCkhny3g==
x-ms-exchange-transport-forked: True
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-OriginatorOrg: hotmail.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-AuthSource: BN3NAM04FT014.eop-NAM04.prod.protection.outlook.com
X-MS-Exchange-CrossTenant-RMS-PersistedConsumerOrg: 00000000-0000-0000-0000-000000000000
X-MS-Exchange-CrossTenant-Network-Message-Id: 8cd3bd63-d2c7-4153-7a8e-08d84d254acb
X-MS-Exchange-CrossTenant-rms-persistedconsumerorg: 00000000-0000-0000-0000-000000000000
X-MS-Exchange-CrossTenant-originalarrivaltime: 30 Aug 2020 20:43:03.6303
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Internet
X-MS-Exchange-CrossTenant-id: 84df9e7f-e9f6-40af-b435-aaaaaaaaaaaa
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BN3NAM04HT102
Subject: Re: [PHP-DEV] Proposal: Adding functions any(iterable $input,
 ?callable $cb = null, int $use_flags=0) and all(...)
From: tysonandre775@hotmail.com (tyson andre)

Hi Dik Takken,=0A=
=0A=
> I would love to see this come to PHP. I also do a lot of Python=0A=
> development and I really like its operators module, which provides=0A=
> function equivalents to the intrinsic Python operators. Have a look:=0A=
> =0A=
> https://docs.python.org/3/library/operator.html=0A=
> =0A=
> Although the any() and all() functions are my favorites, having the full=
=0A=
> range of operator functions would be great. That could ultimately yield=
=0A=
> a proper counterpart to the range of array_* functions which support any=
=0A=
> iterable in stead of just arrays. Think of a keys() function that is a=0A=
> generalization of array_keys().=0A=
=0A=
Operator overloading was the subject of https://wiki.php.net/rfc/userspace_=
operator_overloading#vote=0A=
which had a 38-28 vote, which didn't meet the 2/3 voting threshold.=0A=
=0A=
> This reminds me of the iter library that Nikita created:=0A=
> =0A=
> https://github.com/nikic/iter=0A=
> =0A=
> So yes, this can also be done in user space. Having it built into the=0A=
language has advantages though:=0A=
> =0A=
> * High adoption rates, making lots of existing PHP code more concise=0A=
> * Possibly better performance=0A=
=0A=
Strangely, if I remember correctly, I've had better performance looping inl=
ine and calling closures than calling array_map=0A=
when there were a large number of elements. I think opcache's inferences an=
d the fact there's no switch=0A=
from internals back to the php vm was part of the reason for it.=0A=
(and the tracking of stack frames?)=0A=
=0A=
The readability's my main reason for making it native. I'd wondered if ther=
e'd be a benefit to preloading where we replace C functions with native php=
 functions (optionally JITted) if there's an expected benefit,=0A=
like HHVM does for parts of its standard library, but again out of scope.=
=0A=
=0A=
> Regarding performance: Since we already have opcodes for handling=0A=
> operators, it may be possible to extend these to consume arbitrary=0A=
> numbers of operands from iterables. Then, an operator function like=0A=
> any() would compile into its equivalent operator opcode. Finally, that=0A=
> opcode can probably be JIT compiled into a tight loop in machine code.=0A=
=0A=
That's definitely something I'd want to see for array_map/array_filter.=0A=
There's the question of whether it'd be reasonable to omit stack frames or =
whether it's practical to create fake frames with parameter values=0A=
for debug_backtrace().=0A=
=0A=
Larry Garfield also did some work on python-like list comprehensions,=0A=
though it's still in draft.=0A=
Automatically rewriting array_map(), array_filter(), all(), etc. to list co=
mprehensions and then to JITted code=0A=
could see a nice performance benefit in benefit, but this is out of scope o=
f the RFC I'm working on=0A=
See https://externals.io/message/104637=0A=
=0A=
> So a +1 from me for adding any() and all() but let us also consider the=
=0A=
> general problem of lacking iterable support in PHP. Perhaps that could=0A=
> be the subject of a followup RFC.=0A=
=0A=
Part of the issue is what the return type would be.=0A=
For example, reduce(iterable, callable, $initial =3D null) could work well,=
 but what should map() do=0A=
=0A=
`map(callable, iterable $values): iterable` - should it preserve keys, shou=
ld it return an array|Generator, array|CustomTraversable, something dependi=
ng on the object type, etc.=0A=
(Generator performance seems like it'd be worse than eager evaluation)=0A=
=0A=
Regards,=0A=
Tyson=