Run PHPUnit on multiple PHP versions using Docker

One of the biggest advantage using Docker is the ability to switch between tools versions. Personally I think the easiest way to demonstrate how you can benefit from this advantage is to run PHPUnit tests against different PHP versions.

Prepare the application

For this example I will use a package build on top of Respect Validation validator called flex-validator which offers a simpler interface to validate multiple inputs together and return all the errors at once.

We will clone the repository, install all it’s dependencies and finally run all it’s tests on 7.0 and 7.2 versions of PHP.

Clone the repo

$ git clone git@github.com:iulyanp/flex-validator.git

Install the dependencies using Docker

$ docker run --rm -v $(pwd):/var/www composer install

Running UnitTests

Now it’s time to run the tests on PHP 7.0 version.

$ docker run --rm -v $(pwd):/var/www -w /var/www php:7.0 vendor/bin/phpunit

PHPUnit 6.5.6 by Sebastian Bergmann and contributors.

.....................                                             21 / 21 (100%)

Time: 624 ms, Memory: 4.00MB

OK (21 tests, 46 assertions)

What happened?

• downloaded the php:7.0 docker image
• started a container from that image
• mapped the current folder into the /var/www directory -v $(pwd):/var/www
• made the /var/www directory the current one in the container -w /var/www
• run the vendor/bin/phpunit command inside the /var/www directory
• after tests executed the container closed and removed itself --rm

Or you can run it on PHP 7.2 version to see everything will work nicely with the latest PHP version.

$ docker run --rm -v $(pwd):/var/www -w /var/www php:7.2 vendor/bin/phpunit

PHPUnit 6.5.6 by Sebastian Bergmann and contributors.

.....................                                             21 / 21 (100%)

Time: 751 ms, Memory: 4.00MB

OK (21 tests, 46 assertions)

As you can see this is a very clean way to test our application or package on different PHP versions without installing them locally.

If you liked this post, you can share it on Twitter. Also you can follow me on Github or endorse me on LinkedIn.