Migrations and Seeding

Creating migrations

php vendor/bin/plato make:migration create_users
php vendor/bin/plato migrate
php vendor/bin/plato migrate:status
php vendor/bin/plato migrate:rollback

<?php

use plato\database\blueprint;
use plato\database\migration;
use plato\database\schema;

return new class extends migration {
    public function up(): void
    {
        schema::create('user', function (blueprint $table): void {
            $table->big_increments('id');
            $table->string('email', 190)->unique();
            $table->timestamps();
        });
    }

    public function down(): void
    {
        schema::drop_if_exists('user');
    }
};

Each migrate run creates one batch; rollback reverts the latest batch. The migrator uses a server-side lock to prevent concurrent deploys and loads all migration files before running any of them.

schema creates, alters, drops, renames, and inspects tables. blueprint supplies common columns, indexes, and modifiers. MySQL and ClickHouse grammars compile the DDL their engines support. Whether DDL is transactional depends on the database engine; do not assume a transaction can reverse structural changes.

Seeders

php vendor/bin/plato make:seeder base_roles
php vendor/bin/plato db:seed --class=base_roles

Seeders are explicit, repeatable data tasks. The framework has no model factory; domain fixtures and product defaults belong to the host application.