Link Collector
Prevent duplicate page links

How It Works

The LinkCollector scans the page content before Twig/Markdown processing and collects every internal link slug.

Raw Content:
  "Check out [our Alps tours](/alps/hiking) and..."

  ↓ LinkCollector scans content
  → Finds: "/alps/hiking"
  → Registers slug in collector

  ↓ Twig executes (pages_list, custom functions)
  → Can now exclude "alps/hiking" from results

  ↓ Final output
  → No duplicate links

Supported Link Formats

The collector detects internal links in these formats:

Not collected:

• External links (https://example.com)
• Relative links (relative-page without leading /)
• Anchor-only links (#section)

Usage with pages_list

{# Content contains [See our hiking tours](/hiking): the hiking page is left out #}
{{ pages_list('children', 6, excludeAlreadyLinked: true) }}

A list using the parameter also registers the cards it renders, so several lists on the same page never show the same page twice:

{{ pages_list('taxonomy:destination', 4, excludeAlreadyLinked: true) }}
{# a card rendered above will not come back below #}
{{ pages_list('children', 6, excludeAlreadyLinked: true) }}

Registering is opt-in, like filtering: a pages_list() without the parameter renders normally and leaves the collector untouched — a full listing followed by an excluding carousel keeps working.

An excluding list still renders its full max: the exclusion is done by the query, so max counts the pages that survived it — the list backfills with the next matching pages instead of coming up short.

A paginated list only registers the cards of the pager page it renders; the other pager pages are not linked, so they stay listable.

Usage with pages() Function

For more control, filter pages() with exclude_linked():

{% set allPages = pages(host, 'taxonomy:travel') %}
{% set uniquePages = exclude_linked(allPages) %}
<ul>
{% for page in uniquePages %}
  <li><a href="{{ pw_path(page) }}">{{ page.title }}</a></li>
{% endfor %}
</ul>

Available Twig Functions

Checking Individual Slugs

{% for page in pages(host, 'related') %}
  {% if not is_slug_linked(page.slug) %}
    <a href="{{ pw_path(page) }}">{{ page.title }}</a>
  {% endif %}
{% endfor %}

Debugging Collected Links

{# See what links were collected from your content #}
{{ dump(linked_slugs()) }}
{# Output: ['alps/hiking', 'beach-trips', 'contact'] #}

Usage in Custom Twig Functions (PHP)

A custom Twig function returning pages can inject LinkCollectorService:

use Pushword\Core\Service\LinkCollectorService;

final class MyExtension
{
    public function __construct(
        private readonly LinkCollectorService $linkCollector,
    ) {
    }

    #[AsTwigFunction('my_product_list', isSafe: ['html'])]
    public function renderProductList(): string
    {
        $pages = $this->getProductPages();

        // Exclude pages already linked in content
        $pages = $this->linkCollector->excludeRegistered($pages);

        // Render...
    }
}

LinkCollectorService API

// Register a slug manually
$linkCollector->registerSlug('my-page');

// Register from a Page entity
$linkCollector->register($page);

// Check if a slug is registered
$linkCollector->isSlugRegistered('my-page'); // true/false

// Filter an array of pages
$uniquePages = $linkCollector->excludeRegistered($pages);

// Get all registered slugs
$slugs = $linkCollector->getRegisteredSlugs(); // ['slug' => true, ...]

Runtime

• The filter runs once per page render, before Twig, without database queries; excluding lists then add their own rendered cards as Twig executes.
• The collector resets on each HTTP request and between two pages of a static export.
• It only collects: content is never modified, and nothing is filtered unless you use excludeAlreadyLinked: true or exclude_linked().