Pomm2 Faker client

This package is a Pomm2 client to fill tables with data generated by Faker.

This package is still under development and has no tests hence there are bugs.

Installation

To install Pomm2Faker, just add the following line in the require-dev section of your composer.json file:

"pragmafabrik/pomm-faker": "1.0-dev"

and invoke the composer.phar update command.

Usage

Using the Client is fairly easy:

<?php

use PommProject\Foundation\Pomm;

use PragmaFabrik\Pomm\Faker\FakerPooler;

use Faker\Factory as FakerFactory;

$loader  = require __DIR__.'/vendor/autoload.php';
$pomm    = new Pomm(['my_db' => ['dsn' => 'pgsql://user:pass@host:port/db_name']]);

// register Faker Pooler to the session
$session = $pomm
    ->getDefaultSession()
    ->registerClientPooler(new FakerPooler(FakerFactory::create()))
    ;

 /*
Configure how we want data to be generated in the table `student`.
We do not want to insert any data for the `student_id` field nor the
`created_at`, they are both set by the database and will be returned by the
insert query.
We can use Faker's formatters directly using the `setFormatterType` method or
create a custom generator with the `setDefinition` method.
The untouched field `last_section` will be set to NULL by default.
*/
$session->getFaker('student')->getRowDefinition()
    ->unsetDefinition('student_id')
    ->setFormatterType('first_name', 'firstName')
    ->setFormatterType('last_name', 'lastName')
    ->setFormatterType('birthdate', 'dateTime', ['18 years ago'])
    ->setFormatterType('section', 'randomElement', [['A1', 'A2', 'A3']])
    ->setFormatterType('login', 'lastName')
    ->setDefinition('password', function($generator) {
        return sprintf("pass%s", $generator->format('password'));
    })
    ->unsetDefinition('created_at')
    ;

// Populate the database with 10 students and return the rows converted values.
$students = $session->getFaker('student')->save(10);

Complex structures and batch classes.

Sometimes, structures can be fairly complex with cross foreign key constraints. In this case, it is possible to use the ModelLayer to leverage Postgres's transaction mechanisms. This Pooler is provided by the ModelManager package.