BeanFactory Cache

Intro

The BeanFactory is a great way to load and query for beans without using SQL.

For example we can load an Account with a known id:

$bean = BeanFactory::getBean('Accounts', '1234');

Unexpected behaviour

However this method can cause some unexpected behaviour. Let’s look at the following example: wzxhzdk:1

This produces the following output:

Acc1 Name: Bar
Acc2 Name: Bar

Acc1 Name: Bar
Acc2 Name: Bar
Acc3 Name: Foo

Enter the cache

So what is going on here?

Basically the getBean method is cached. getBean will cache at most 10 beans and drop some of the older entries based on usage. $acc2 is pulled from the cache and is therefore the same instance as $acc1. Any changes in one will be reflected in the other.

However when we load a bunch of beans our cached account drops out the BeanFactory Cache and when we fetch the bean into $acc3 we get a fresh instance from the database.