php-expressive / iterators
A set of Iterator classes to assist in writing more expressive PHP code
Requires (Dev)
- phpunit/phpunit: ^10
README
PHP Expessive: Iterators
Introduction
What if you could write code in a way that reduced or even eliminated the need for "explanatory" comments? You know, the ones that say something like // This is what this code does
. If the code is complicated enough to warrant an explination, perhaps the code itself requires a revisit. It's estimated that software engineers spend as much as ten times as much time reading code than they do writing it. Shouldn't we make understanding our code easier and more enjoyable?
This library (and, eventually, the whole of the PHP Expressive initiative) aims to help improve the readability and understandability of our code; specificially in relation to using iterators and iterables (like array
). It is heavily insprired by the Iterator
trait in Rust (and even "borrows" some of its unit tests and documentation), but written for a PHP audience.
The goal of this library is not to stop you from writing for
or foreach
loops (although you will probably find yourself writing far fewer of them), but rather to increase the readability of your code. Even the most complicated loop can usually be broken down into smaller pieces that are more easily understood. By using the functionality of this library, you can reduce boiler-plate code, increase readability, and benefit from pre-written, pre-tested algorithms.
Motivating Example
Imagine you have an array of email addresses ($input_array
) and would like to trim this array to just the first 10 of these email addresses that are <something>@gmail.com
. Furthermore, you only need the name to the left of the @
(since they're all @gmail.com
anyway), and would prefer the names to be in lowercase. One of the most obvious ways of achieving this would be using a foreach
loop:
$num_chars_to_trim = strlen('@gmail.com'); $gmail_addresses = []; foreach ($input_array as $address) { $lower_address = strtolower($address); // We are assuming PHP 8.2, and ignoring RegEx functions for now if (str_ends_with($lower_address, '@gmail.com')) { $gmail_addresses[] = substr($lower_address, 0, -$num_chars_to_trim); if (count($gmail_addresses) == 10) { break; } } }
While this is a perfectly acceptable solution, if you were to look at it again with fresh eyes (imagine coming across this block of code in a few months), it would take some mental effort to understand what's going on. In order to make the code more expressive (i.e., readable), let's factor some of the functionality into name functions.
$is_gmail_address = fn($address) => str_ends_with($address, '@gmail.com'); $num_chars_to_trim = strlen('@gmail.com'); $trim_gmail_com = fn($address) => substr($address, 0, -$num_chars_to_trim); $gmail_addresses = []; foreach ($input_array as $address) { $lower_address = strtolower($address); if ($is_gmail_address($lower_address)) { $gmail_addresses[] = $trim_gmail_com($lower_address); if (count($gmail_addresses) == 10) { break; } } }
That's a little better, but can we do better? It still seems like there's a lot going in the foreach
loop. We could refactor the inner-most if
check, moving the logic outside of the loop at the expense of potentially unnecessary processing.
// ... As before $gmail_addresses = []; foreach ($input_array as $address) { $lower_address = strtolower($address); if ($is_gmail_address($lower_address)) { $gmail_addresses[] = $trim_gmail_com($lower_address); } } $gmail_addresses = array_slice($gmail_addresses, 0, 10);
The foreach
is a bit easier to read now, but if our original list has hundreds or thousands of items, we're processing all of them, even after we have the 10 we want. That's wasteful.
Let's look at PHP's Iterator
classes to see if they can help
// ... As before $input_iterator = new \ArrayIterator($input_array); $lowercased = new \TransformIterator($filtered, fn($str) => strtolower($str)); // Uh oh! $filtered = new \CallbackFilterIterator($input_iterator, $is_gmail_address); $transformed = new \TransformIterator($filtered, $trim_gmail_com); // Uh oh, again! $gmail_addresses = new \LimitIterator($transformed, limit: 10); $gmail_addresses = iterator_to_array($gmail_addresses);
This looks promising, but PHP doesn't have a TransformIterator
. Let's write one quick:
class TransformIterator implements Iterator { public function __construct(private iterable $iterator, private \Closure $f) {} public function current(): mixed { return ($this->f)($this->iterator->current()); } public function key(): mixed { return $this->iterator->key(); } public function next(): void { $this->iterator->next(); } public function rewind(): void { $this->iterator->rewind(); } public function valid(): bool { return $this->iterator->valid(); } }
Ok, that wasn't really that quick!
Looking back at previous code block, it does seem to be a bit more expressive. Reading from top to bottom, we can look at the imporant pieces: We're filtering the input array that are gmail addresses, transforming them by trimming gmail.com, limiting them to just 10 and then convering that to an array.
Let's think outside the box
The last sentence in the paragraph above certainly explains what we're trying to accomplish much more succictly than describing the loop in terms of foreach
and if
. However, what if we could make the code look even more like the description. Imagine the following.
// ... Same helper functions $gmail_addresses = // Note: This is *almost* correct $input_array ->transform(fn($str) => strtolower($str)) ->filter($is_gmail_address) ->transform($trim_gmail_com) ->take(10) ->to_array();
Wow! That code almost reads like a sentence! Now that's expressive! In fact, that's exactly what you can do with the Iterators
in this library, with one small modification. We first need to make an Iterator
out of the $input_array
which we can do like this
$gmail_addresses = // We just wrap $input_array in iter() to make it an Iterator iter($input_array) ->transform(fn($str) => strtolower($str)) ->filter($is_gmail_address) ->transform($trim_gmail_com) ->take(10) ->to_array();
Version
This library follows Semantic Versioning 2.0 and is currently in its initial developement phase (pre v1.0.0). However, every effort has been made to make it useable and helpful. If you'd like to help make it better, please see the CONTRIBUTING guide for how you can help!
Current (and Planned) Features
TODO
Changelog
Please see the Changelog to see the new features, bug fixes and deprecations for each release.
Documentation
You can look at the API documentation to see all of the currently implemented features and lots of examples of how to use them in your code.