I've been doing method chaining for quite some time, and it seems like the more I add to the dynamic approach of chaining, the more it gets like a big mess, such as adding a new feature to chains then has to work with all other types of chains.
Currently, I've dropped some older projects to re-write them in a more simpler fashion, what I'm working on currently is:
$psm->select('users')->where('id = 1')->run(function($row){ // do code }); Which works well, it creates statements dynamically (dynamic as in I could add another chain in the middle to create another clause in the statement.
What I'm wondering is if there's a certain structure to follow when creating chainable methods, because of course, you need to in the end return $this unless running the last method.
My current bare-bones structure in my head for creating chainable classes is as follows:
class foo { public $temporary = ''; public function add($string) { $this->temporary .= $string; return $this; } public function ret() { return $this->temporary; } } So, when running
$foo = new foo; echo $foo->add('hello ')->add('world!')->ret(); It outputs "hello world!" as you'd expect, but are there better approaches to creating chainable OOP methods? It just feels a bit hacky.
"Is there a certain way to chain methods in PHP?" refers to the safest and easiest to manage way of chaining methods. - This is important to me as when releasing my code I want it to be in the correct community-accepted format.
:-)