Read from various data sources.
Shape and optimize for your needs.
Store and secure in one of many available data sinks.
Transformations are reusable blocks of transformations that can be applied to a data frame.
The main goal of Transformations is to a gruop together a set of transformations.
Transformation needs to implement the Transformation interface.
interface Transformation
{
public function transform(DataFrame $dataFrame) : DataFrame;
}
composer.json
{
"name": "flow-php/examples",
"description": "Flow PHP - Examples",
"license": "MIT",
"type": "library",
"require": {
"flow-php/etl": "1.x-dev"
},
"minimum-stability": "dev",
"config": {
"allow-plugins": {
"php-http/discovery": false
}
}
}code.php<?php
declare(strict_types=1);
use function Flow\ETL\DSL\{concat_ws, data_frame, from_array, lit, ref, to_stream};
use Flow\ETL\Transformation;
require __DIR__ . '/vendor/autoload.php';
data_frame()
->read(
from_array([
['id' => 1, 'first_name' => 'John', 'last_name' => 'Doe'],
['id' => 2, 'first_name' => 'Jane', 'last_name' => 'Smith'],
['id' => 3, 'first_name' => 'Bob', 'last_name' => 'Johnson'],
['id' => 4, 'first_name' => 'Alice', 'last_name' => 'Williams'],
])
)
->with(
/**
* Create new column "name" by concatenating "first_name" and "last_name" columns.
*/
new class implements Transformation {
public function transform(Flow\ETL\DataFrame $dataFrame) : Flow\ETL\DataFrame
{
return $dataFrame->withEntry('name', concat_ws(lit(' '), ref('first_name'), ref('last_name')));
}
}
)
->collect()
->write(to_stream(__DIR__ . '/output.txt', truncate: false))
->run();
+----+------------+-----------+----------------+
| id | first_name | last_name | name |
+----+------------+-----------+----------------+
| 1 | John | Doe | John Doe |
| 2 | Jane | Smith | Jane Smith |
| 3 | Bob | Johnson | Bob Johnson |
| 4 | Alice | Williams | Alice Williams |
+----+------------+-----------+----------------+
4 rows
