bennett-treptow / laravel-migration-generator

Generate migrations from existing database structures

Stars
207

Laravel Migration Generator

Latest Version on Packagist

Generate migrations from existing database structures, an alternative to the schema dump provided by Laravel. A primary use case for this package would be a project that has many migrations that alter tables using ->change() from doctrine/dbal that SQLite doesn't support and need a way to get table structures updated for SQLite to use in tests.
Another use case would be taking a project with a database and no migrations and turning that database into base migrations.

Installation

composer require --dev bennett-treptow/laravel-migration-generator
php artisan vendor:publish --provider="LaravelMigrationGenerator\LaravelMigrationGeneratorProvider"

Lumen Installation

composer require --dev bennett-treptow/laravel-migration-generator

Copy config file from vendor/bennett-treptow/laravel-migration-generator/config to your Lumen config folder

Register service provider in bootstrap/app.php

$app->register(\LaravelMigrationGenerator\LaravelMigrationGeneratorProvider::class);  

Usage

Whenever you have database changes or are ready to squash your database structure down to migrations, run:

php artisan generate:migrations

By default, the migrations will be created in tests/database/migrations. You can specify a different path with the --path option:

php artisan generate:migrations --path=database/migrations

You can specify the connection to use as the database with the --connection option:

php artisan generate:migrations --connection=mysql2

You can also clear the directory with the --empty-path option:

php artisan generate:migrations --empty-path

This command can also be run by setting the LMG_RUN_AFTER_MIGRATIONS environment variable to true and running your migrations as normal. This will latch into the MigrationsEnded event and run this command using the default options specified via your environment variables. Note: it will only run when your app environment is set to local.

Configuration

Want to customize the migration stubs? Make sure you've published the vendor assets with the artisan command to publish vendor files above.

Environment Variables

Key Default Value Allowed Values Description
LMG_RUN_AFTER_MIGRATIONS false boolean Whether or not the migration generator should run after migrations have completed.
LMG_CLEAR_OUTPUT_PATH false boolean Whether or not to clear out the output path before creating new files. Same as specifying --empty-path on the command
LMG_TABLE_NAMING_SCHEME [Timestamp]_create_[TableName]_table.php string The string to be used to name table migration files
LMG_VIEW_NAMING_SCHEME [Timestamp]_create_[ViewName]_view.php string The string to be used to name view migration files
LMG_OUTPUT_PATH tests/database/migrations string The path (relative to the root of your project) to where the files will be output to. Same as specifying --path= on the command
LMG_SKIPPABLE_TABLES migrations comma delimited string The tables to be skipped
LMG_SKIP_VIEWS false boolean When true, skip all views
LMG_SKIPPABLE_VIEWS '' comma delimited string The views to be skipped
LMG_SORT_MODE 'foreign_key' string The sorting mode to be used. Options: foreign_key
LMG_PREFER_UNSIGNED_PREFIX true boolean When true, uses unsigned variant methods instead of the ->unsigned() modifier.
LMG_USE_DEFINED_INDEX_NAMES true boolean When true, uses index names defined by the database as the name parameter for index methods
LMG_USE_DEFINED_FOREIGN_KEY_INDEX_NAMES true boolean When true, uses foreign key index names defined by the database as the name parameter for foreign key methods
LMG_USE_DEFINED_UNIQUE_KEY_INDEX_NAMES true boolean When true, uses unique key index names defined by the database as the name parameter for the unique methods
LMG_USE_DEFINED_PRIMARY_KEY_INDEX_NAMES true boolean When true, uses primary key index name defined by the database as the name parameter for the primary method
LMG_USE_DEFINED_DATATYPE_ON_TIMESTAMP false boolean When false, uses ->timestamps() by mashing up created_at and updated_at regardless of datatype defined by the database
LMG_MYSQL_TABLE_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_TABLE_NAMING_SCHEME when the database driver is mysql.
LMG_MYSQL_VIEW_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_VIEW_NAMING_SCHEME when the database driver is mysql.
LMG_MYSQL_OUTPUT_PATH null ?boolean When not null, this setting will override LMG_OUTPUT_PATH when the database driver is mysql.
LMG_MYSQL_SKIPPABLE_TABLES null ?boolean When not null, this setting will override LMG_SKIPPABLE_TABLES when the database driver is mysql.
LMG_MYSQL_SKIPPABLE_VIEWS null comma delimited string The views to be skipped when driver is mysql
LMG_SQLITE_TABLE_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_TABLE_NAMING_SCHEME when the database driver is sqlite.
LMG_SQLITE_VIEW_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_VIEW_NAMING_SCHEME when the database driver is sqlite.
LMG_SQLITE_OUTPUT_PATH null ?boolean When not null, this setting will override LMG_OUTPUT_PATH when the database driver is sqlite.
LMG_SQLITE_SKIPPABLE_TABLES null ?boolean When not null, this setting will override LMG_SKIPPABLE_TABLES when the database driver is sqlite.
LMG_SQLITE_SKIPPABLE_VIEWS null comma delimited string The views to be skipped when driver is sqlite
LMG_PGSQL_TABLE_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_TABLE_NAMING_SCHEME when the database driver is pgsql.
LMG_PGSQL_VIEW_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_VIEW_NAMING_SCHEME when the database driver is pgsql.
LMG_PGSQL_OUTPUT_PATH null ?boolean When not null, this setting will override LMG_OUTPUT_PATH when the database driver is pgsql.
LMG_PGSQL_SKIPPABLE_TABLES null ?boolean When not null, this setting will override LMG_SKIPPABLE_TABLES when the database driver is pgsql.
LMG_PGSQL_SKIPPABLE_VIEWS null comma delimited string The views to be skipped when driver is pgsql
LMG_SQLSRV_TABLE_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_TABLE_NAMING_SCHEME when the database driver is sqlsrc.
LMG_SQLSRV_VIEW_NAMING_SCHEME null ?boolean When not null, this setting will override LMG_VIEW_NAMING_SCHEME when the database driver is sqlsrv.
LMG_SQLSRV_OUTPUT_PATH null ?boolean When not null, this setting will override LMG_OUTPUT_PATH when the database driver is sqlsrv.
LMG_SQLSRV_SKIPPABLE_TABLES null ?boolean When not null, this setting will override LMG_SKIPPABLE_TABLES when the database driver is sqlsrv.
LMG_SQLSRV_SKIPPABLE_VIEWS null comma delimited string The views to be skipped when driver is sqlsrv

Stubs

There is a default stub for tables and views, found in resources/stubs/vendor/laravel-migration-generator/.
Each database driver can be assigned a specific migration stub by creating a new stub file in resources/stubs/vendor/laravel-migration-generator/ with a driver-prefix, e.g. mysql-table.stub for a MySQL specific table stub.

Stub Naming

Table and view stubs can be named using the LMG_(TABLE|VIEW)_NAMING_SCHEME environment variables. Optionally, driver-specific naming schemes can be used as well by specifying LMG_{driver}_TABLE_NAMING_SCHEME environment vars using the same tokens. See below for available tokens that can be replaced.

Table Name Stub Tokens

Table stubs have the following tokens available for the naming scheme:

Token Example Description
[TableName] users Table's name, same as what is defined in the database
[TableName:Studly] Users Table's name with Str::studly() applied to it (useful for standardizing table names if they are inconsistent)
[TableName:Lowercase] users Table's name with strtolower applied to it (useful for standardizing table names if they are inconsistent)
[Timestamp] 2021_04_25_110000 The standard migration timestamp format, at the time of calling the command: Y_m_d_His
[Index] 0 The key of the migration in the sorted order, for use with enforcing a sort order
[IndexedEmptyTimestamp] 0000_00_00_000041 The standard migration timestamp format, but filled with 0s and incremented by [Index] seconds
[IndexedTimestamp] 2021_04_25_110003 The standard migration timestamp format, at the time of calling the command: Y_m_d_His incremented by [Index] seconds

Table Schema Stub Tokens

Table schema stubs have the following tokens available:

Token Description
[TableName] Table's name, same as what is defined in the database
[TableName:Studly] Table's name with Str::studly() applied to it, for use with the class name
[TableUp] Table's up() function
[TableDown] Table's down() function
[Schema] The table's generated schema

View Name Stub Tokens

View stubs have the following tokens available for the naming scheme:

Token Example Description
[ViewName] user_earnings View's name, same as what is defined in the database
[ViewName:Studly] UserEarnings View's name with Str::studly() applied to it (useful for standardizing view names if they are inconsistent)
[ViewName:Lowercase] user_earnings View's name with strtolower applied to it (useful for standardizing view names if they are inconsistent)
[Timestamp] 2021_04_25_110000 The standard migration timestamp format, at the time of calling the command: Y_m_d_His
[Index] 0 The key of the migration in the sorted order, for use with enforcing a sort order
[IndexedEmptyTimestamp] 0000_00_00_000041 The standard migration timestamp format, but filled with 0s and incremented by [Index] seconds
[IndexedTimestamp] 2021_04_25_110003 The standard migration timestamp format, at the time of calling the command: Y_m_d_His incremented by [Index] seconds

View Schema Stub Tokens

View schema stubs have the following tokens available:

Token Description
[ViewName] View's name, same as what is defined in the database
[ViewName:Studly] View's name with Str::studly() applied to it, for use with the class name
[Schema] The view's schema

Example Usage

Given a database structure for a users table of:

CREATE TABLE `users` (
  `id` int(10) unsigned NOT NULL AUTO_INCREMENT,
  `username` varchar(128) COLLATE utf8mb4_unicode_ci DEFAULT NULL,
  `email` varchar(255) COLLATE utf8mb4_unicode_ci NOT NULL,
  `password` varchar(255) COLLATE utf8mb4_unicode_ci NOT NULL,
  `first_name` varchar(45) COLLATE utf8mb4_unicode_ci DEFAULT NULL,
  `last_name` varchar(45) COLLATE utf8mb4_unicode_ci DEFAULT NULL,
  `timezone` varchar(45) COLLATE utf8mb4_unicode_ci NOT NULL DEFAULT 'America/New_York',
  `location_id` int(10) unsigned NOT NULL,
  `deleted_at` timestamp NULL DEFAULT NULL,
  `remember_token` varchar(255) COLLATE utf8mb4_unicode_ci DEFAULT NULL,
  `created_at` timestamp NULL DEFAULT NULL,
  `updated_at` timestamp NULL DEFAULT NULL,
  PRIMARY KEY (`id`),
  KEY `users_username_index` (`username`),
  KEY `users_first_name_index` (`first_name`),
  KEY `users_last_name_index` (`last_name`),
  KEY `users_email_index` (`email`),
  KEY `fk_users_location_id_index` (`location_id`)
  CONSTRAINT `users_location_id_foreign` FOREIGN KEY (`location_id`) REFERENCES `locations` (`id`) ON UPDATE CASCADE ON DELETE CASCADE
) ENGINE=InnoDB AUTO_INCREMENT=1 DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci

A tests/database/migrations/[TIMESTAMP]_create_users_table.php with the following Blueprint would be created:

<?php

use Illuminate\Support\Facades\Schema;
use Illuminate\Database\Schema\Blueprint;
use Illuminate\Database\Migrations\Migration;

class CreateUsersTable extends Migration
{
    /**
     * Run the migrations.
     *
     * @return void
     */
    public function up()
    {
        Schema::create('users', function (Blueprint $table) {
            $table->increments('id');
            $table->string('username', 128)->nullable()->index();
            $table->string('email', 255)->index();
            $table->string('password', 255);
            $table->string('first_name', 45)->nullable()->index();
            $table->string('last_name', 45)->index();
            $table->string('timezone', 45)->default('America/New_York');
            $table->unsignedInteger('location_id');
            $table->softDeletes();
            $table->string('remember_token', 255)->nullable();
            $table->timestamps();
            $table->foreign('location_id', 'users_location_id_foreign')->references('id')->on('locations')->onUpdate('cascade')->onDelete('cascade');
        });
    }

    /**
     * Reverse the migrations.
     *
     * @return void
     */
    public function down()
    {
        Schema::dropIfExists('users');
    }
}

Currently Supported DBMS's

These DBMS's are what are currently supported for creating migrations from. Migrations created will, as usual, follow what database drivers Laravel migrations allow for

  • MySQL
  • Postgres
  • SQLite
  • SQL Server