This post seeks to discourage two patterns in any PHP code that may ever need to scale:<\/p>\n\n\n\n
Both of these may seem like innocent, sensible additions at one time without it being clear to developers how grave the consequences are down the line. (skip to a dynamic-property benchmark<\/a>)<\/p>\n\n\n\n This is fine for an introduction to OOP, but ignores real world needs like representing relational data. Consider that the consumer of a Company object may want to know the ids of employees at the company, and that those may need to be loaded from a database. To avoid repeated queries, it makes sense to cache that data. This leads to the addition of:<\/p>\n\n\n\n An unrelated feature may then require the ability to iterate over the attributes of a company, and perhaps only those that would be stored in the “companies” table of a SQL database. That becomes more complicated when the object’s set of properties now includes both attributes of the company itself (name, address), and something derived at runtime from another dataset (employee_ids).<\/p>\n\n\n\n This is by far the simplest solution in that it isolates attributes that need to be iterated over inside a data structure that’s already iterable.<\/p>\n\n\n\n The issue:<\/strong> Hash tables are memory hogs, and every instance of Company now requires a hash table.<\/p>\n\n\n\n To overly simplify things, if a class has n<\/em> properties, the interpreter can allocate space for n<\/em> pointers for each instance of the object. It then only needs a single list of symbols to map each to a particular pointer. Ignoring other data, such as property visibility, this can be illustrated roughly like so:<\/p>\n\n\n\n To find the baz property of object n<\/em>, the interpreter needs only to find that the index of baz is 2, then add 2b<\/em> to the first pointer-address of n<\/em> to find the pointer to the data. Most critically, the memory overhead of mapping the list of properties is only incurred once. That is, where m<\/em> is the size of the map, and b<\/em> is the size of the pointer, the total allocation is m<\/em> + 3nb<\/em>. It’s important to note that m<\/em> is larger than b<\/em>, making it minimally painful to scale n<\/em>.<\/p>\n\n\n\n When an associative $properties array is introduced, PHP implements this as a hashtable for every instance<\/em>. The data structure may be represented as:<\/p>\n\n\n\n Each of these encounters overhead that was only needed once in the previous example. Where s<\/em> is the size of one of these, the overhead becomes sn<\/em>. In plain language, this is the relevant comparison:<\/p>\n\n\n\n 1 big thing plus n<\/em> small things vs.<\/strong> n<\/em> big things<\/p>\n\n\n\n With all of this considered, it becomes immediately obvious why $properties arrays are a terrible idea for any sufficiently large set of objects.<\/p>\n\n\n\n The desire is to have an isolated, iterable container for a specific set of data related to the company. Another object that implements things like the Iterator interface is arguably the best way to handle this in PHP. A basic implementation is available on GitHub<\/a> as part of a test setup that demonstrates just how terrible the memory tradeoff of convenience can be (more on that later). To simply illustrate the solution at a high level:<\/p>\n\n\n\n This grants the desired isolation of attributes about a given company into $properties as well as providing convenient iteration and access, but it does so while avoiding the overhead of an array for each object because each $properties is an efficiently stored instance of a CompanyData class.<\/p>\n\n\n\n PHP is overly permissive and that can lead to destroying that memory advantage. Namely, the allowance of dynamic property assignment has a terrible side-effect.<\/p>\n\n\n\n ** This generated a notice prior to PHP 8.2, and the error_reporting default hid these notices until 8.0. As of 8.2, it generates a warning.<\/em><\/p>\n\n\n\n When PHP encounters this code, it can no longer rely on the map from properties to pointers that makes objects efficient because “color” is not in that map. Instead, a new map is created just for the $cow instance. This is effectively a return to having the overhead of a $properties array for every instance.<\/p>\n\n\n\nStarting Simple<\/h4>\n\n\n\n
class Company {\n public $name;\n public $address;\n}<\/code><\/pre>\n\n\n\n protected $employee_ids;\n public function employeeIds() {\n if (!isset($this->employee_ids) {\n $this->employee_ids = \/* fetch from DB *\/;\n }\n return $this->employee_ids;\n }<\/code><\/pre>\n\n\n\nA (Non-) Solution<\/h4>\n\n\n\n
class Company {\n \/\/ with the expectation that $this->properties['name']\n \/\/ replaces $this->name\n public $properties;\n}<\/code><\/pre>\n\n\n\nUnderstanding the Impact on Memory<\/h5>\n\n\n\n
\/\/ mapping symbols to indices\n{\n 'foo': 0,\n 'bar': 1,\n 'baz': 2\n}\n\n\/\/ 3 blocks of memory for object 1\n[*ptr_to_foo][*ptr_to_bar][*ptr_to_baz]\n\n\/\/ repeat for objects 2 to n-1...\n\n\/\/ 3 blocks of memory for object n\n[*ptr_to_foo][*ptr_to_bar][*ptr_to_baz]<\/code><\/pre>\n\n\n\n\/\/ $properties of object 1\n{\n 'foo': *ptr_to_foo_1,\n 'bar': *ptr_to_bar_1,\n 'baz': *ptr_to_baz_1\n}\n\n\/\/ repeat for objects 2 to n-1...\n\n\/\/ $properties of object n\n{\n 'foo': *ptr_to_foo_n,\n 'bar': *ptr_to_bar_n,\n 'baz': *ptr_to_baz_n\n}<\/code><\/pre>\n\n\n\nAn Efficient Solution<\/h4>\n\n\n\n
abstract class Data implements \\ArrayAccess, \\Iterator, \\Countable {\n \/\/ refer to the GitHub link above for the body\n}\n\nclass CompanyData {\n public $name;\n public $address;\n}\n\nclass Company {\n \/\/ $properties is an instance of CompanyData\n public $properties;\n}\n\n$c = new Company();\n\n\/\/ works thanks to ArrayAccess interface\n$c->properties['name'] = 'Widgets Unlimited';\n$c->properties['address'] = '123 Main St.';\n\n\/\/ works thanks to Iterator interface\nforeach ($c->properties as $prop => $value) {}<\/code><\/pre>\n\n\n\nDanger Ahead<\/h4>\n\n\n\n
class Animal {\n public $type;\n}\n\n$cow = new Animal();\n\/\/ PHP allows this**\n$cow->color = 'brown';<\/code><\/pre>\n\n\n\n