PHP Luminova: Database Seeding Management
Luminova simplifies database management with the novakit command and its associated methods. This ensures an easy process for executing seeders and convenient table truncation.
Database seeding is a process that allows you to populate your database with initial data, test data, or additional data required for your application. It is an essential part of database management, especially during development and testing phases. Luminova offers built-in support for database seeding through the novakit
command line tool, making it easy to manage and apply seed data in a structured and repeatable manner.
How Seeding Works
Seeder Classes: Seeders are classes that extend the
\Luminova\Database\Seeder
abstract class. Each seeder class contains arun
method where you define the data to be inserted into your database tables. This structure helps maintain clean and organized code.Run Method: The
run
method within a seeder class is where the actual seeding logic resides. Here, you can specify which tables to populate and the data to be inserted. This method ensures that your seed data is applied consistently every time it is executed.Calling Other Seeders: Within a seeder's
run
method, you can call other seeders using theinvoke
method. This allows you to break down complex seeding operations into smaller, manageable tasks and maintain a clear seeding hierarchy.Seeder Setup: Create a class that extends
\Luminova\Database\Seeder
. Ensure all seed classes are placed in the/app/Database/Seeders/
directory for proper run.Seeding Management: Luminova provides
novakit
commands to run, rollback, and manage seeders. These commands ensure that your seed data is applied in the correct order and allow you to easily revert changes if necessary.
- Class namespace:
\Luminova\Database\Seeder
- This class is an Abstract class
Methods
See documentation reference for Database Builder.
run
Run database table seeder.
public run(\Luminova\Database\Builder $builder): void
Parameters:
Parameter | Type | Description |
---|---|---|
$builder | \Luminova\Database\Builder | Database builder instance. |
Run Example
This example demonstrates how to create a seed
class to insert data into a database table. You can implement various insertion methods based on your needs, such as using the insert
method for array-based insertion or the exec
method for SQL query string insertion.
<?php
namespace App\Database\Seeders;
use \Luminova\Database\Seeder;
use \Luminova\Database\Builder;
class UsersTableSeeder extends Seeder
{
public function run(Builder $builder): void
{
$builder->table('users')->insert([
'name' => 'Peter',
'phone' => '080673653',
'email' => '[email protected]',
]);
}
}
invoke
Invokes another seeder within the current seeder's run method.
final protected invoke(class-string<\Luminova\Database\Seeder> $seed): void
Parameters:
Parameter | Type | Description |
---|---|---|
$seed | class-string<\Luminova\Database\Seeder> | The seed class name. |
Note: When you use the
invoke
method to specify another seeder class, it will also be executed during the seeding process or when rolling back seeds.
Seeding Command
To executes all pending database seeders.
php novakit db:seed
To executes a specific pending database seeder class.
php novakit db:seed --class=TestSeeder
Command Options
-c, --class
Description: Specify the seeder class to run.
Usage:
php novakit db:seed --class=TestSeeder
-t, --table
Description: Specify the seeder class table name to automatically create a backup which can be restored back if rollback fails.
Usage:
php novakit db:seed --table=foo
-n, --no-backup
- Description: Run seeder without creating a backup. This should be used with caution as it will create not backup to rollback to.
- Usage:
php novakit db:seed --no-backup
-r, --rollback
- Description: Rollback the seeder to the previous version. This allows you to revert changes that were applied.
- Usage:
php novakit db:seed --rollback
-i, --invoke
- Description: Run seeder and also invoke other invokable seeders classes.
- Usage:
php novakit db:seed --invoke
Truncate Documentation
The db:truncate
command is used to truncate a database table, effectively removing all records within the table. Unlike traditional deletion methods, truncation is a quick operation that empties the table without logging individual row deletions, making it significantly faster for large datasets.
Table Truncation
Truncation clears all records from the specified database table. It does not trigger the execution of seeders and does not update the seeder version after truncation. This means that if you intend to reseed after truncation, you must ensure that your seeder is updated accordingly.
Usage
To use the db:truncate
command, follow these examples:
Truncate a Specific Table
php novakit db:truncate --table=TestTable
This command will truncate the TestTable
in the database, removing all its records.
Options
-t, --table
: Specifies the name of the database table to truncate.-n, --no-transaction
: Runs the database truncation without wrapping it in a transaction. This option can be useful when you do not need the rollback capability typically provided by transactions.
Examples
To truncate a Table Without Transaction.
php novakit db:truncate --table=TestTable --no-transaction
In this example, the --no-transaction
option is used to truncate TestTable
without using a transaction, which can be beneficial in certain scenarios where transactions are not required or desired.
Considerations
- Seeder Locks: Truncation ignores seeder locks. It's essential to manage and update your seeders appropriately after truncating tables to avoid inconsistencies.
- Reseeding: After truncation, ensure your seeder class has been updated from the last version. Alternatively, use the
db:clear
command to clear the seeder if necessary before reseeding to avoid conflicts or errors.