Add opt-in unique indexing jobs to prevent Scout reindexing already queued models

[stevebauman]

Description:

I recently discovered that every time $model->save() is called on a searchable model (regardless of whether the model actually changed), Scout dispatches a MakeSearchable job. In write-heavy applications this means the queue fills with tons of MakeSearchable jobs for the same records, causing the same documents to be re-indexed more often than necessary. The same applies to RemoveFromSearch.

Because these jobs use SerializesModels, a queued job always re-fetches the latest model state when it runs -- so collapsing redundant jobs for the same records has no effect on what ultimately gets indexed. It just removes wasted work.

This PR adds opt-in unique variants of both indexing jobs:

• Laravel\Scout\Jobs\MakeSearchableUnique
• Laravel\Scout\Jobs\RemoveFromSearchUnique

Both extend their base job and implement ShouldBeUniqueUntilProcessing, sharing a UniqueByScoutKeys trait that derives the job's unique identity from the queueable model class plus its (sorted) Scout keys.

Usage:

Developers can opt-in by pointing Scout at the unique jobs (ex. in a service provider):

use Laravel\Scout\Scout;
use Laravel\Scout\Jobs\MakeSearchableUnique;
use Laravel\Scout\Jobs\RemoveFromSearchUnique;

Scout::makeSearchableUsing(MakeSearchableUnique::class);
Scout::removeFromSearchUsing(RemoveFromSearchUnique::class);

I've made this opt-in since this ShouldBeUnique assumes that the developer has a default cache store set up.

Let me know your thoughts! ❤️