# constructor hook - Thread: https://externals.io/message/63056 - Original archive: https://news-web.php.net/php.internals/63056 - Started by Rasmus Schultz on 2012-09-18T12:19:19+00:00 - Messages: 12 - Reply by email to internals@lists.php.net - More for agents: search with https://externals.io/search.md?q={query} · JSON API at https://externals.io/api · documentation at https://externals.io/llms.txt --- ## #63056 — Rasmus Schultz — 2012-09-18T12:19:19+00:00 URL: https://externals.io/message/63056#63056 Hey, I'm going to make this brief, because I suspect a lot of people are going to jump at the opportunity to cry bloody murder when I suggest this. I wonder if it would make sense to have a way to globally hook into __construct() - sort of like how you can hook into the autoloaders with `spl_autoload_register()` ... This could be useful for things like dependency injection and debugging. Lithium (for one) uses a unified constructor to accomplish something like this: http://lithify.me/docs/manual/lithium-basics/architecture.wiki That's bad for various reasons I won't delve into. Various DI containers use a replacement for the "new" keyword, e.g. DI::create('MyClass') which is bad (worse) for other reasons... So basically, what if you could solve these design patterns by hooking into constructors directly instead - something like: constructor_hook_register(function($object) { > echo 'an object of type '.get_class($object).' was constructed!'; > }); > $test = new Foo(); would print: an object of type Foo was constructed! One other area where this could help, is in event-driven systems, where this sort of thing could be used for "co-construction". Of course, you could accomplish the same thing in a more controlled fashion, by simply establishing conventions - for example, DI::inject(new My Class()) would work just as well, it's just inconvenient to type; on the other hand, it's explicit about what's happening. So I'm not at all demanding this feature - I'm merely posting this for debate. Thanks, Rasmus Schultz --- ## #63059 — johannes@schlueters.de — 2012-09-18T13:07:13+00:00 URL: https://externals.io/message/63056#63059 · In reply to: #63056 > Hey, > > I'm going to make this brief, because I suspect a lot of people are going > to jump at the opportunity to cry bloody murder when I suggest this. > > I wonder if it would make sense to have a way to globally hook into > __construct() - sort of like how you can hook into the autoloaders with > `spl_autoload_register()` ... > > This could be useful for things like dependency injection and debugging. I don't think it's a good thing to have. "new" right now gives you a clear idea whats going on. Magic behind the scenes is hard to debug. If you really want it: https://github.com/johannes/php-test-helpers johannes --- ## #63060 — Kris Craig — 2012-09-18T13:18:08+00:00 URL: https://externals.io/message/63056#63060 · In reply to: #63059 On Tue, Sep 18, 2012 at 6:07 AM, Johannes Schlüter wrote: > > > […] > HELP! BLOODY MURDER!!!! (Sorry). --Kris P.S. Why did this have to happen now?! He was just three days away from retirement.... P.P.S. Now I'm gonna find the *real* killer! P.P.P.S. (Sorry again). --- ## #63061 — Ferenc Kovacs — 2012-09-18T13:48:59+00:00 URL: https://externals.io/message/63056#63061 · In reply to: #63060 > On Tue, Sep 18, 2012 at 6:07 AM, Johannes Schlüter > wrote: > > > […] > going > > […] > > HELP! BLOODY MURDER!!!! > > > (Sorry). > > --Kris > > P.S. Why did this have to happen now?! He was just three days away from > retirement.... > > P.P.S. Now I'm gonna find the *real* killer! > > P.P.P.S. (Sorry again). > I find your recent activity on this list pretty unconstructive: http://www.mail-archive.com/internals@lists.php.net/msg61060.html http://www.mail-archive.com/internals@lists.php.net/msg60923.html http://www.mail-archive.com/internals@lists.php.net/msg60824.html http://www.mail-archive.com/internals@lists.php.net/msg60668.html the last/oldest not completely off-topic, but you still think that "What's my point? I'm not sure that I have one--" please don't burden the subscribers of the list with these kind of emails. thanks. -- Ferenc Kovács @Tyr43l - http://tyrael.hu --- ## #63062 — Edmond [HT] — 2012-09-18T15:08:41+00:00 URL: https://externals.io/message/63056#63062 · In reply to: #63056 Hello Rasmus! ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ RS> Hey, RS> I wonder if it would make sense to have a way to globally hook into RS> __construct() - sort of like how you can hook into the autoloaders with RS> `spl_autoload_register()` ... RS> This could be useful for things like dependency injection and debugging. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Programming language should have a clear behavior. But here is too much magic. Usually, DI objects are not created in a vacuum. They require a factory... may be service locator. Like this: http://snipt.org/vLae1/Traditional -- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Regards, Edmond -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= --- ## #63063 — Amaury Bouchard — 2012-09-18T15:14:27+00:00 URL: https://externals.io/message/63056#63063 · In reply to: #63056 Sounds to me like aspect-oriented programming, applied to object construction. Take a look at: https://github.com/AOP-PHP/AOP 2012/9/18 Rasmus Schultz : > Hey, > > I'm going to make this brief, because I suspect a lot of people are going > to jump at the opportunity to cry bloody murder when I suggest this. > > I wonder if it would make sense to have a way to globally hook into > __construct() - sort of like how you can hook into the autoloaders with > `spl_autoload_register()` ... > > This could be useful for things like dependency injection and debugging. > > Lithium (for one) uses a unified constructor to accomplish something like > this: > > http://lithify.me/docs/manual/lithium-basics/architecture.wiki > > That's bad for various reasons I won't delve into. > > Various DI containers use a replacement for the "new" keyword, e.g. > DI::create('MyClass') which is bad (worse) for other reasons... > > So basically, what if you could solve these design patterns by hooking into > constructors directly instead - something like: > > > […] > > > would print: > > an object of type Foo was constructed! > > > One other area where this could help, is in event-driven systems, where > this sort of thing could be used for "co-construction". > > Of course, you could accomplish the same thing in a more controlled > fashion, by simply establishing conventions - for example, DI::inject(new > My Class()) would work just as well, it's just inconvenient to type; on the > other hand, it's explicit about what's happening. > > So I'm not at all demanding this feature - I'm merely posting this for > debate. > > Thanks, > Rasmus Schultz --- ## #63064 — jpauli — 2012-09-18T16:22:37+00:00 URL: https://externals.io/message/63056#63064 · In reply to: #63063 > Sounds to me like aspect-oriented programming, applied to object construction. > Take a look at: https://github.com/AOP-PHP/AOP +1, this is the same case as AOP, but for constructor. Thus AOP may fit the need. Julien.P --- ## #63075 — Ralph Schindler — 2012-09-18T17:10:35+00:00 URL: https://externals.io/message/63056#63075 · In reply to: #63064 This is one of the main use cases of AOP. Demonstrated by this script: setArguments(array(new Bar)); } aop_add_before('Foo->__construct()', 'doDi'); $f = new Foo; // 'instance of Bar' -ralph > > > […] > > +1, this is the same case as AOP, but for constructor. > Thus AOP may fit the need. > > Julien.P --- ## #63121 — Rasmus Schultz — 2012-09-18T22:46:02+00:00 URL: https://externals.io/message/63056#63121 · In reply to: #63064 Good point - I agree. Thanks for taking the time to think about this! > On Tue, Sep 18, 2012 at 5:14 PM, Amaury Bouchard > wrote: > > […] > construction. > > […] > > +1, this is the same case as AOP, but for constructor. > Thus AOP may fit the need. > > Julien.P --- ## #63123 — Kris Craig — 2012-09-19T02:31:00+00:00 URL: https://externals.io/message/63056#63123 · In reply to: #63063 > Sounds to me like aspect-oriented programming, applied to object > construction. > Take a look at: https://github.com/AOP-PHP/AOP > > Yeah I wasn't aware of this, either, but I can think of several scenarios in which it'd be useful. I know this is PECL, but is this something that might be worth bundling? I'm still a bit new here so if that idea's heresy please let me know. If not, that would be a nice feature to have on a default installation I think. Thoughts? @Ferenc Yes congrats on cherry-picking a few recent posts that weren't Nobel Prize material (some of which did make constructive points when viewed in context, albeit with some creative flair). I'm not in the market for a nanny right now; I'll let you know if that ever changes. If you'd like to discuss your personal feelings about this further, please do so via private email so we don't spam the list. --Kris --- ## #63159 — jpauli — 2012-09-19T17:34:15+00:00 URL: https://externals.io/message/63056#63159 · In reply to: #63123 > > On Tue, Sep 18, 2012 at 8:14 AM, Amaury Bouchard wrote: > > > […] > Yeah I wasn't aware of this, either, but I can think of several scenarios > in which it'd be useful. I know this is PECL, but is this something that > might be worth bundling? I'm still a bit new here so if that idea's heresy > please let me know. If not, that would be a nice feature to have on a > default installation I think. Thoughts? Why not, 2 solutions here : - Bundle it with PHP sources - Merge it with core (trunk) I think that it may need more code to bind AOP functionnalities to Reflection though. Anyway, an RFC will be needed ;) Julien.Pauli > On Tue, Sep 18, 2012 at 8:14 AM, Amaury Bouchard > wrote: > > > […] > Yeah I wasn't aware of this, either, but I can think of several scenarios > in which it'd be useful. I know this is PECL, but is this something that > might be worth bundling? I'm still a bit new here so if that idea's heresy > please let me know. If not, that would be a nice feature to have on a > default installation I think. Thoughts? > > @Ferenc Yes congrats on cherry-picking a few recent posts that > weren't Nobel Prize material (some of which did make constructive points > when viewed in context, albeit with some creative flair). I'm not in the > market for a nanny right now; I'll let you know if that ever changes. If > you'd like to discuss your personal feelings about this further, please do > so via private email so we don't spam the list. > > --Kris --- ## #63172 — Kris Craig — 2012-09-19T18:37:02+00:00 URL: https://externals.io/message/63056#63172 · In reply to: #63159 > > > > […] > wrote: > > […] > that > > […] > heresy > > […] > > Why not, 2 solutions here : > - Bundle it with PHP sources > - Merge it with core (trunk) > > I think that it may need more code to bind AOP functionnalities to > Reflection though. > > Anyway, an RFC will be needed ;) > > Julien.Pauli > Agreed. --Kris