This document is for a version of CakePHP that is no longer supported. Please upgrade to a newer release!
Phinx is run using a number of commands.
The Init Command
The Init command (short for initialize) is used to prepare your project for
Phinx. This command generates the phinx.yml
file in the root of your
project directory.
$ cd yourapp
$ phinx init .
Open this file in your text editor to setup your project configuration. Please see the Configuration chapter for more information.
The Create command is used to create a new migration file. It requires one argument: the name of the migration. The migration name should be specified in CamelCase format.
$ phinx create MyNewMigration
Open the new migration file in your text editor to add your database
transformations. Phinx creates migration files using the path specified in your
phinx.yml
file. Please see the Configuration chapter
for more information.
You are able to override the template file used by Phinx by supplying an alternative template filename.
$ phinx create MyNewMigration --template="<file>"
You can also supply a template generating class. This class must implement the
interface Phinx\Migration\CreationInterface
.
$ phinx create MyNewMigration --class="<class>"
In addition to providing the template for the migration, the class can also define a callback that will be called once the migration file has been generated from the template.
You cannot use --template
and --class
together.
The Migrate command runs all of the available migrations, optionally up to a specific version.
$ phinx migrate -e development
To migrate to a specific version then use the --target
parameter or -t
for short.
$ phinx migrate -e development -t 20110103081132
Use --dry-run
to print the queries to standard output without executing them
$ phinx migrate --dry-run
The Rollback command is used to undo previous migrations executed by Phinx. It is the opposite of the Migrate command.
You can rollback to the previous migration by using the rollback
command
with no arguments.
$ phinx rollback -e development
To rollback all migrations to a specific version then use the --target
parameter or -t
for short.
$ phinx rollback -e development -t 20120103083322
Specifying 0 as the target version will revert all migrations.
$ phinx rollback -e development -t 0
To rollback all migrations to a specific date then use the --date
parameter or -d
for short.
$ phinx rollback -e development -d 2012
$ phinx rollback -e development -d 201201
$ phinx rollback -e development -d 20120103
$ phinx rollback -e development -d 2012010312
$ phinx rollback -e development -d 201201031205
$ phinx rollback -e development -d 20120103120530
If a breakpoint is set, blocking further rollbacks, you can override the
breakpoint using the --force
parameter or -f
for short.
$ phinx rollback -e development -t 0 -f
Use --dry-run
to print the queries to standard output without executing them
$ phinx rollback --dry-run
Note
When rolling back, Phinx orders the executed migrations using
the order specified in the version_order
option of your
phinx.yml
file.
Please see the Configuration chapter for more information.
The Status command prints a list of all migrations, along with their current status. You can use this command to determine which migrations have been run.
$ phinx status -e development
This command exits with code 0 if the database is up-to-date (ie. all migrations are up) or one of the following codes otherwise:
1: There is at least one migration left to be executed.
2: There was a migration run and recorded in the database, but the migration file is now missing.
The Breakpoint command is used to set breakpoints, allowing you to limit rollbacks. You can toggle the breakpoint of the most recent migration by not supplying any parameters.
$ phinx breakpoint -e development
To toggle a breakpoint on a specific version then use the --target
parameter or -t
for short.
$ phinx breakpoint -e development -t 20120103083322
You can remove all the breakpoints by using the --remove-all
parameter
or -r
for short.
$ phinx breakpoint -e development -r
Breakpoints are visible when you run the status
command.
The Seed Create command can be used to create new database seed classes. It requires one argument, the name of the class. The class name should be specified in CamelCase format.
$ phinx seed:create MyNewSeeder
Open the new seed file in your text editor to add your database seed commands.
Phinx creates seed files using the path specified in your phinx.yml
file.
Please see the Configuration chapter for more information.
The Seed Run command runs all of the available seed classes or optionally just one.
$ phinx seed:run -e development
To run only one seed class use the --seed
parameter or -s
for short.
$ phinx seed:run -e development -s MyNewSeeder
When running Phinx from the command line, you may specify a configuration file
using the --configuration
or -c
parameter. In addition to YAML, the
configuration file may be the computed output of a PHP file as a PHP array:
<?php
return [
"paths" => [
"migrations" => "application/migrations"
],
"environments" => [
"default_migration_table" => "phinxlog",
"default_database" => "dev",
"dev" => [
"adapter" => "mysql",
"host" => $_ENV['DB_HOST'],
"name" => $_ENV['DB_NAME'],
"user" => $_ENV['DB_USER'],
"pass" => $_ENV['DB_PASS'],
"port" => $_ENV['DB_PORT'],
]
]
];
Phinx auto-detects which language parser to use for files with *.yml
and *.php
extensions. The appropriate
parser may also be specified via the --parser
and -p
parameters. Anything other than "php"
is treated as YAML.
When using a PHP array, you can provide a connection
key with an existing PDO instance. It is also important to pass
the database name too, as Phinx requires this for certain methods such as hasTable()
:
<?php
return [
"paths" => [
"migrations" => "application/migrations"
],
"environments" => [
"default_migration_table" => "phinxlog",
"default_database" => "dev",
"dev" => [
"name" => "dev_db",
"connection" => $pdo_instance
]
]
];
Phinx can also be run inside of a web application by using the Phinx\Wrapper\TextWrapper
class. An example of this is provided in app/web.php
, which can be run as a
standalone server:
$ php -S localhost:8000 vendor/robmorgan/phinx/app/web.php
This will create local web server at http://localhost:8000 which will show current migration status by default. To run migrations up, use http://localhost:8000/migrate and to rollback use http://localhost:8000/rollback.
The included web app is only an example and should not be used in production!
Note
To modify configuration variables at runtime and override %%PHINX_DBNAME%%
or other another dynamic option, set $_SERVER['PHINX_DBNAME']
before
running commands. Available options are documented in the Configuration page.
Phinx can be used within your unit tests to prepare or seed the database. You can use it programatically :
public function setUp ()
{
$app = new PhinxApplication();
$app->setAutoExit(false);
$app->run(new StringInput('migrate'), new NullOutput());
}
If you use a memory database, you’ll need to give Phinx a specific PDO instance. You can interact with Phinx directly using the Manager class :
use PDO;
use Phinx\Config\Config;
use Phinx\Migration\Manager;
use PHPUnit\Framework\TestCase;
use Symfony\Component\Console\Input\StringInput;
use Symfony\Component\Console\Output\NullOutput;
class DatabaseTestCase extends TestCase {
public function setUp ()
{
$pdo = new PDO('sqlite::memory:', null, null, [
PDO::ATTR_ERRMODE => PDO::ERRMODE_EXCEPTION
]);
$configArray = require('phinx.php');
$configArray['environments']['test'] = [
'adapter' => 'sqlite',
'connection' => $pdo
];
$config = new Config($configArray);
$manager = new Manager($config, new StringInput(' '), new NullOutput());
$manager->migrate('test');
$manager->seed('test');
// You can change default fetch mode after the seeding
$this->pdo->setAttribute(PDO::ATTR_DEFAULT_FETCH_MODE, PDO::FETCH_OBJ);
$this->pdo = $pdo;
}
}