Collections & Helpers

A Collection is Laravel's wrapper around a PHP array with a hundred-odd chainable methods. Every Eloquent get() returns one, and collect([...]) turns any array into one. Add the Str and Arr helpers and Carbon dates, and you have the toolkit behind most day-to-day Laravel code.

This lesson shows the methods you'll use constantly, on the real task data behind this site. Code captioned with a file path is copied from this repository; blocks marked Illustrative — not in this app run on the same data but aren't in the codebase.

What a Collection is#

Illustrative — not in this app.

$tasks = Task::with('project')->get();     // Illuminate\Database\Eloquent\Collection
$names = collect(['Ava', 'Ben', 'Chloe']); // Illuminate\Support\Collection

$tasks->count();     // 20
$tasks->first();     // a Task model
$tasks->isEmpty();   // false
$tasks->toArray();   // plain PHP arrays

Methods like map() and filter() return a new collection and never modify the original, so you can chain them into a readable pipeline:

Illustrative — not in this app.

$overdueTitles = Task::with('project')->get()
    ->filter(fn (Task $task) => $task->due_on?->isPast() && ! $task->isComplete())
    ->sortBy('due_on')
    ->map(fn (Task $task) => "{$task->project->name}: {$task->title}")
    ->values();

Try the methods on live data#

The explorer loads this app's 20 seeded tasks from /api/demo/tasks and runs each method in JavaScript the same way Laravel would. Pick a method to see the PHP and its result.

map, filter and reject#

• map() transforms every item.
• filter() keeps items where the callback returns true; with no callback, it removes falsy values.
• reject() is the opposite of filter().

The demo query-count endpoint uses nested map() to build its response:

app/Http/Controllers/Api/QueryCountController.php

$rows = $projects->map(fn (Project $project) => [
    'project' => $project->name,
    'tasks' => $project->tasks->map(fn ($task) => [
        'title' => $task->title,
        'assignee' => $task->assignee?->name,
    ]),
]);

pluck, keyBy and mapWithKeys#

pluck() pulls one field out of every item; with a second argument it builds a key => value map. The CMS admin uses it to fill dropdowns:

app/Filament/Resources/PostResource.php

->options(fn () => BlogType::orderBy('name')->pluck('name', 'id'))

mapWithKeys() lets you choose the key and value yourself. The demo seeder uses it to create users and tags and keep them addressable by name:

database/seeders/TutorialDemoSeeder.php

$tags = collect(['Frontend', 'Backend', 'Bug', 'Design', 'DevOps'])
    ->mapWithKeys(fn (string $name) => [$name => Tag::firstOrCreate(
        ['slug' => Str::slug($name)],
        ['name' => $name],
    )]);

Later it picks some of them by key and plucks their ids, which is exactly what attach() needs:

database/seeders/TutorialDemoSeeder.php

$task->tags()->attach($tags->only($taskTags)->pluck('id'));

contains, search and first#

These answer "is it there?" and "where is it?". The seeder that added the tutorials section to this site's homepage uses both:

database/seeders/HomeTutorialsBlockSeeder.php

if (collect($blocks)->contains('type', 'tutorials')) {
    return;
}

$heroIndex = collect($blocks)->search(fn (array $block): bool => ($block['type'] ?? null) === 'hero');
$position = $heroIndex === false ? 0 : $heroIndex + 1;

Illustrative — not in this app.

$tasks->first(fn ($task) => $task->priority === TaskPriority::High); // first match or null
$tasks->firstWhere('title', 'Write API docs');                        // shorthand
$tasks->contains(fn ($task) => $task->due_on?->isToday());            // bool

groupBy, countBy and partition#

These split or summarise a collection, and come up whenever you build a report or a dashboard.

Illustrative — not in this app.

$byProject = $tasks->groupBy('project.name');          // ['Website Relaunch' => Collection, ...]
$perPriority = $tasks->countBy(fn ($t) => $t->priority->value); // ['high' => 5, 'medium' => 9, ...]
[$done, $open] = $tasks->partition(fn ($t) => $t->isComplete());

groupBy() accepts dot notation for nested fields and a closure for computed keys. partition() returns two collections, which array destructuring unpacks.

Sorting, sums and averages#

Illustrative — not in this app.

$tasks->sortBy('due_on');                  // ascending
$tasks->sortByDesc(fn ($t) => $t->tags->count());
$tasks->sortBy([['due_on', 'asc'], ['title', 'asc']]); // multiple columns

$projects->sum('tasks_count');
$projects->avg('open_tasks_count');
$tasks->max(fn ($t) => strlen($t->title));

each versus map#

each() runs a callback for its side effects and returns the same collection; map() builds a new one. Return false from each() to stop early. On large tables, use Model::query()->each() or lazy() so Eloquent fetches rows in chunks instead of all at once.

Illustrative — not in this app.

Task::query()->open()->lazy()->each(function (Task $task) {
    // processes one row at a time, 1000 per query behind the scenes
});

The Str helpers#

Illuminate\Support\Str covers the string jobs you'd otherwise write by hand. The admin turns titles into URL slugs as you type:

app/Filament/Resources/PageResource.php

$set('slug', Str::slug($state));

And the demo seeder builds email addresses and throwaway passwords:

database/seeders/TutorialDemoSeeder.php

['email' => Str::slug($name, '.').'@demo.example'],
['name' => $name, 'password' => Str::random(64)],

Try the most common ones on your own text:

Illustrative — not in this app. Fluent strings chain like collections:

Str::of('  Build pricing calculator in Vue  ')
    ->trim()
    ->lower()
    ->replace('vue', 'Vue')
    ->limit(20);   // 'build pricing calcul...'

The Arr helpers and data_get#

Arr works on plain arrays, especially nested ones such as config, JSON payloads and the CMS's blocks column.

Illustrative — not in this app.

$block = ['type' => 'hero', 'data' => ['title' => 'Demo Blog', 'image' => null]];

Arr::get($block, 'data.title');            // 'Demo Blog'
Arr::get($block, 'data.subtitle', '');     // '' (default)
Arr::only($block['data'], ['title']);      // ['title' => 'Demo Blog']
Arr::pluck($blocks, 'type');               // ['hero', 'tutorials', 'blog_types', 'blog_posts']
data_get($page->blocks, '0.data.title');   // works on arrays and objects

Dates with Carbon#

now(), today() and every date-cast model attribute are Carbon objects. Task::markComplete() stores the current time:

app/Models/Task.php

$this->update(['completed_at' => now()]);

and the date and datetime casts make $task->due_on and $task->completed_at Carbon instances, which is why the API resource can call ->toDateString() on them.

Illustrative — not in this app.

$task->due_on->isPast();                    // bool
$task->due_on->diffForHumans();             // '3 days from now'
now()->addWeek()->startOfDay();             // next week, 00:00
$task->completed_at->format('j M Y H:i');   // '16 Oct 2026 15:24'
Task::whereDate('due_on', '<', today())->open()->count();

Key takeaways#

• Eloquent queries return Collections; collect() wraps any array.
• Chain filter, map, sortBy and values into readable pipelines; each step returns a new collection.
• pluck, mapWithKeys, groupBy, countBy and partition cover most reporting and dropdown needs.
• Filter in SQL when you can, and use lazy() or each() on queries for large tables.
• Str handles slugs, random strings and formatting; Arr and data_get read nested arrays safely.
• Date casts give you Carbon objects, so date maths and formatting are one method call.