Drupal Continuous Integration (European Drupal Days 2015)

© Ibuildings 2014/2015 - All rights reserved
#DrupalDaysEU
DRUPAL
CONTINUOUS INTEGRATION
by Claudio Beatrice

#DrupalDaysEU
Gold Sponsors

#DrupalDaysEU
Media Sponsors
Silver Sponsors

DRUPAL
CONTINUOUS
INTEGRATION
Introduction
Drupal Continuous Integration - By Claudio Beatrice - Copyright 2014http://tenwarp.com/

Continuous Integration
“Continuous Integration” is a
software development practice
where members of a team
integrate work frequently,
usually each person integrates at
least daily - leading to multiple
integrations per day. Each
integration is veriﬁed by an
automated build (including
tests) to detect integration
errors as quickly as possible.
— Martin Fowler

• automating repetitive tasks such as build, QA and
deploy
• fast and frequent integrations
• fast feedback loop, catch errors asap
• ability to deploy quickly and keep track of every
released set of code
ADVANTAGES
Drupal Continuous Integration - By Claudio Beatrice - Copyright 2014-2015http://tenwarp.com/

developer
commits triggers
conﬁg
static
analysis
unit /
functional
tests
runs
runs
runs
report
FAIL
deploy
smoke /
functional
tests
report
SUCCESS
SUCCESS
FAIL
CI
server
revision
control
log
logs
logs
logs
BUILD PROCESS EXAMPLE

TOOLS
• SCM: git, svn, cvs, …
• CI SERVER: jenkins
• BUILD TOOL: phing
• DRUPAL AUTOMATION: drush
• DEPENDENCIES MANAGEMENT: drush make & composer
• DRUPAL CDD TOOLS: features & migrate
• TESTING: PHPUnit & Behat

DRUPAL - I
Drupal, as opposed to “classic” frameworks such as
Symfony or Zend, does not make the setup and
maintenance a Continuous Integration cycle easy
due to the high coupling between data and
conﬁguration within its database.

DRUPAL - II
To overcome this problem it’s necessary to adopt a
strictly controlled, disciplined work methodology
and to make use of a set of tools that allow us to
transform conﬁguration into code, to easily deal
with content and to incrementally update
database’s schema.

WORKFLOW - I
A developer’s workﬂow consist not only in the plain
old features implementation, but also in the correct,
precise and complete export of all the element of
conﬁguration that are necessary to rebuild the
functionalities on several environments.
!
Key modules: Features and Strongarm.

WORKFLOW - II
It’s often necessary to setup some default content that
may vary during development and that might be
expanded or updated once the site has been
published. It’s therefore necessary to formalise in code
all the data import procedures in order to better
administrate them and to be able to replicate them at
any given time.
!
Key module: Migrate.

WORKFLOW - III
As soon as a self-contained task such a bug-ﬁx or a
new feature implementation has been completed,
it’s necessary to commit and push that unit of work
to the central repository in order to integrate, test
and deploy it automatically on the development
environment.
!
Key tools: git/svn.

WORKFLOW - IV
!
Writing automated tests is necessary in order to
guarantee the behaviour of a given implementation
and to make sure that future changes won’t break
the work already done, creating regressions and
dealing potential damage to users and clients.
!
Key tools: PHPUnit, Behat.

Q & A

EXERCISES
1. What are the main advantages of a Continuous Integration system?
2. When do Unit Tests are executed during the build? How about
functional? Why are they executed at different times?
3. Why is it important to write tests?
4. Which beneﬁts come from using a Source Control Management system?
5. Why is Drupal so hard -compared to other tools such as “classic”
frameworks- to use in a Continuous Integration environment?
6. What tools allow Drupal to be successfully employed in a Continuous
Integration development cycle and for what reasons?

DRUPAL
CONTINUOUS
INTEGRATION

Phing is a build system written in PHP that allows
developers to describe the automated operations
used during development, deployment and
maintenance using plain XML ﬁles.
Phing

• Simplicity: it’s based on well-known, declarative XML syntax
and features a very rich documentation
• Power: thanks to the big amount of available tasks, it is able
to perform almost anything you could do using a
command line interpreter
• Extensibility: as it is written in PHP, it’s easy to extend and
debug
• Automation: the whole application conﬁguration and
deployment process can be handled by build ﬁles
Phing
BENEFITS

• execute tests and log their results
• perform filesystem operations
• interact with source configuration management systems
• generate documentation
• transfer files from/to remote systems
• fire drush commands
• invoke shell commands
Phing
FEATURES

There are two available file types in Phing to configure
the build: .properties and .xml files. The former
contains a list of key/value variables, while the latter
contains the “instructions” to execute.
In a usual setup there are up to five different
environments: loc, dev, stage, debug and prod. Each
of these ones will be configured using a
dedicated .properties file in order to make the build
file environment-agnostic so to reuse most of its code.
Phing
MAIN FILES AND ENVIRONMENTS

Phing
PROPERTY FILE

Phing
BUILD FILE - PROJECT TAG
For every build ﬁle, there must be a single, unique,
root “project" tag. Within that "target", "task" and
“type" tags can be used.

Phing
BUILD FILE - TYPE TAG
Besides default types such as numbers or strings, it is
possible to deﬁne more complex, nested types. An
interesting feature of Phing’s types is the ability to be
referenced by id later.

Phing
BUILD FILE - TARGET TAG
“Targets” are collections of components that are
assigned a unique name within the project. It is possible
to invoke other targets within a target and it’s also
possible to set dependencies between different targets.

Phing
BUILD FILE - TASK TAG
“Task”s are responsible for the execution of specific units
of work and are usually configurable through attributes
and nested tags. A few examples of what tasks do could
be: files copying, archive compression or drush
commands execution.

Phing
EXERCISES
1. What are the main benefits of using a build system?
2. Why is Phing more convenient than a bunch of shell
scripts?
3. What files does Phing use to work? What’s their
use?
4. What are the fundamental elements of a build file?
What’s their use?

Drush
Drush is a command-line shell and scripting
interface for Drupal, a veritable Swiss Army knife
designed to make life easier for those who spend
their working hours hacking away at the command
prompt.
— Drush Team

Drush
• Drupal Power: ability to simply, swiftly and
programmatically execute Drupal-speciﬁc tasks
• Integration: working with other command-line
tools is very easy and effective
BENEFITS

Drush
• code and database backup, restore e sync
• cron tasks execution
• automatic installations
• variables, cache and logs CRUD operations
• modules and themes administration
• SQL commands execution
• users administration
• dependency management
• remote tasks execution
FEATURES

Drush
Drush make is a Drush extension that deals with
modules, themes, libraries and patches
management. It makes use of a manifest ﬁle to index
and download the required versions of the packages
used to build a Drupal website.
MAKE - I

Drush
MAKE - II

Drush
Drush Make allows developers to specify a
destination sub-folder using the “subdir” option.
We’ll leverage that to establish the convention of
having four sub-folders for modules: contrib,
custom, devel and features. This division will help in
keeping a cleaner division between the modules
and will save us the burden of versioning “contrib”
modules.
MAKE - III

Drush
Drush allows the developers to create a kind of
labels called Site Aliases: these are nothing but
conﬁguration groups that fall under an arbitrary
name and make the execution of some Drush
commands easier. In our case those aliases match
the environments deﬁned for our project, and will
look something like myproject.loc and
myproject.dev.
SITE ALIAS - I

Drush
SITE ALIAS - II

Drush
Drush looks for Aliases definitions in several locations:
1. in the path set in the “alias path” variable of the options array
found in the drushrc.php file
2. in the path specified in the —alias-path command-line option
3. in one of the following folders: “/etc/drush”, “$HOME/.drush” or
the current site’s “sites/all/drush”
In order to keep Drush Aliases clean and separated, we’ll use our
projects’ “sites/all/drush” as the go-to location to store their
configuration so not to pollute the global configuration.
SITE ALIAS - III

Drush
Drush implements two extremely useful commands
to keep ﬁles and databases in sync between hosts:
rsync e sql-sync.
• drush rsync @mysite.loc @mysite.dev
• drush sql-sync @mysite.loc @mysite.dev
SYNCHRONISATION

Drush
EXERCISES
1. What’s Drush?
2. Why is it convenient to use it?
3. What are its most important features?
4. What’s the use for a make ﬁle?
5. What’s a Drush Site Alias? What’s its use?

Jenkins
Jenkins is a Java application useful to manage the
execution of those repetitive tasks that are needed
to build a software project.

Jenkins
• automation of deployments procedures
• speed-up of repetitive tasks such as test runs or the
creation of build artifacts to deploy on our projects’
target environments
• logging and monitoring for a handful of code-related
stats such as test results or code quality analysis
• convenient web ui to conﬁgure all the needed jobs
BENEFITS

Jenkins
FEATURES
• integration with several source configuration management
systems
• tests results and code quality reports generation
• integration with several build systems
• jobs status notifications
• configurable web ui
• hundreds of plugins

Jenkins
JOB CREATION - I

Jenkins
JOB CREATION - II

Jenkins
DASHBOARD

Jenkins
ESERCIZI
1. What’s Jenkins?
2. What are its tasks?
3. What’s a Job?

Features
The Feature module is able to export parts of the
configuration found in the database into “ordinary”
Drupal modules. The ability to produce a set of files
containing a specific functionality starting from the
database makes them much easier to manage and
deploy across several systems and installations.

• Removes the need for manual, repetitive work that’s necessary to
reproduce all of a website’s features and traits by encapsulating
them in portable, standard Drupal modules;
• Exports directly to Drupal modules or, in other words, text ﬁles:
that means it’s easy to version them using a source control
management system, which in turn makes easy to keep the history
of changes and allows multiple people to work on the project
concurrently;
• Allows to transport functionalities across websites in an easy,
reproducible and automatic fashion.
Features
BENEFITS

Features
Local Stage
Dev
Prod

Features
CREATION - I

Features
CREATION - II

Features
ADMINISTRATION

Features
OPERATIONS - ENABLING A FEATURE
Enabling a Features-exported module executes the
same operations a “normal” module executes, such as:
• Schema installation;
• System Register and Cache update;
• Execution of hook_install();
• Execution of hook_enable().

Features
OPERATIONS - UPDATE & REVERT
Features Update: updating a feature
consists in the act of saving some of
the information found in the database
into a speciﬁc Feature Module.
Features Revert: reverting a
feature consists in the act of
writing the information found in a
Feature Module into the database.

Features
features-components (fc) List features components.
features-diff (fd) Show the difference between the default
and overridden state of a feature.
features-export (fe) Export a feature from your site into a
module.
features-list (fl) List all the available features for your
site.
features-revert (fr) Revert a feature module on your site.
features-revert-all (fra) Revert all enabled feature module on
your site.
features-update (fu) Update a feature module on your site.
features-update-all (fua) Update all feature modules on your site.
DRUSH COMMANDS

Features
• Thanks to the power and ﬂexibility granted by the
Entity API it’s possible to easily create exportable
Entities that are automatically picked up and
handled by Features;
• A lot of contrib modules are already using such
possibility in order to provide exportable
conﬁguration. Some examples could be the Rules,
Views or Context modules.
BEYOND THE DEFAULT OPTIONS

Features
• Create and maintain “vertical” features such as blog, forum and front page. Avoid
creating Feature Modules that group “exportables” by type, such as “all site’s views” or
“all site’s rules”;
• Do always use English and specific namespaces when naming components
(field_blog_topics instead of field_argomenti);
• Do not use abbreviations unless it’s strictly ncssry.
• Keep a standard format when naming fields: try not to mix casing, dashing,
underscores and so on;
• Define conventions. The Kit Project, although abandoned, is a good source of
inspiration;
• Strive to keep things clean and in order: try to always use Features for importing and
exporting all the developed functionalities.
BEST PRACTICES

Features
EXERCISES
1. What are the main goals of the Features module?
2. What components can it act upon?
3. What does feature-revert do?
4. What beneﬁts does Features bring?
5. What’s Strongarm used for?
6. Can you name a few modules that allow their
components to be exported by Features?

Migrate
The Migrate module makes a powerful and ﬂexible
object oriented framework for importing contents
into Drupal available to developers: it offers out-of-
the-box support for a handful of sources, such as
databases and ﬁle formats, and destinations, such as
nodes and taxonomy terms.

Migrate
At macro-level, there’s four interesting areas to consider when
working with Migrate:
• An API for deﬁning and managing the migration process;
• A set of Drush commands for managing the migrations, such as
import, rollback and status;
• An Admin UI featuring some of the features exposed by Drush;
• Some example modules that demonstrate a handful of
techniques used to deﬁne custom migrations.
FEATURES

Migrate
MigrateSource
id
userName
email
address
MigrateMap
sourceId
destinationId
MigrateDestination
uid
name
mail
UnmigratedDestination
address
ARCHITECTURE

Migrate
SOURCE
• Inherits from the abstract class
MigrateSource
• Provides an interface to the data set
• Iterates on the source’s rows
• Provides a list of ﬁelds

Migrate
DESTINATION
• Inherits from the abstract class MigrateDestination
(and usually from the more specific
MigrateDestinationEntity)
• Responsible for persisting data in a specific
Drupal’s content type such as user, node, comment,
file.

Migrate
MAPPING TABLE
• Inherits from MigrateSQLMap, which in
turn inherits from the abstract class
MigrateMap
• Describes the key formats for the source
and the destination
• Stores all the information that’s used to
import and rollback each and every
single record

Migrate
FIELDS MAPPING
• Allows to link each destination field with a source
field
• Allows to transform data during the migration to
fix or adapt to the destination requirements.
• Allows to exclude some fields from the migration

Migrate
Handlers are classes that allow to execute additional
operations during the import process.
They are usually used to support migrations where
some speciﬁc, non-standard data is connected to
Core Entities: most of the time this will be related to
custom ﬁeld types.
HANDLER

Migrate
There’s two kinds of Handler: Destination and Field.
The former is used to manipulate the Destination
Entity around saving (pre- and post- save) and to
declare additional ﬁelds exposed by the Destination
Entity.
The latter works much like the Destination Handler,
except its methods are invoked for each Field (as
opposed to each Entity).
HANDLER

Migrate
Sometimes there are pieces of content that hold
circular references to each other: for instance a
tutorial post “part 1” could contain a reference to the
“part 2” and vice-versa.
STUB
NODE
nid 1
title tutorial part I
body …
series_nids 2,3,4
NODE
nid 2
title tutorial part II
body …
series_nids 1,3,4

Migrate
During the import process the ﬁrst node won’t be
able to correctly link to the second as the (new) id is
still unknown: Stubs solve this problem by creating
and linking a temporary, empty node that only holds
an id. As soon as the import process gets to the
stubbed node, it will populate it with all the due
information.
STUB

Migrate
Most of the time migrations are deﬁned as static as that
source and the destination match one-to-one: one row,
one migration.
However, sometimes it’s necessary to “join” multiple
sources having the same structure into a single
destination: this kind of migration is deﬁned as dynamic.
An example could be a migration where you need to
import users from several Drupal sites into a new one.
DYNAMIC MIGRATIONS

Migrate
When using this functionality, Drupal will take care of
instantiating each migration multiple times. Besides that, a
dynamic migration differs from a static migration in the
following ways:
• its constructors accepts an array of arguments, where each
of its keys identiﬁes an instance of the migration class;
• it overrides the generateMachineName() method in order
to generate a distinct machine name for each instance, so
to be able to identify them.
DYNAMIC MIGRATIONS

Migrate
migrate-analyze (maz) Analyze the source fields for a migration.
migrate-audit (ma) View information on problems in a migration.
migrate-auto-register (mar) Register any newly-defined migration classes.
migrate-deregister Remove all tracking of a migration.
migrate-fields-destination (mfd) List the fields available for mapping in a destination.
migrate-fields-source (mfs) List the fields available for mapping from a source.
migrate-import (mi) Perform one or more migration processes.
migrate-mappings View information on all field mappings in a migration.
migrate-messages (mmsg) View any messages associated with a migration.
migrate-reset-status (mrs) Reset a active migration's status to idle.
migrate-rollback (mr) Roll back the destination objects from a given migration.
migrate-status (ms) List all migrations with current status.
migrate-stop (mst) Stop an active migration operation.
migrate-wipe (mw) Delete all nodes from specified content types.
DRUSH COMMANDS

Migrate
ADMINISTRATION
http://tenwarp.com/

Migrate
• Migrate D2D - set of utilities for migrating content from
other versions of Drupal (5, 6, 7) to Drupal 7;
• Migrate Extras - integrates several contrib modules
such as Rules or Pathauto;
• Commerce Migrate - Integrates Drupal Commerce’s
Entities;
• Wordpress/Typo3/PHPBB Migrate - set of utilities to aid
the import of content from other well-known platforms.

Migrate
EXERCISES
1. What’s the main goal of the Migrate module?
2. What could be the sources for a migration?
3. What’s the use for the mapping table?
4. What does migrate-rollback do?
5. When is appropriate to use Migrate?
6. What’s a dynamic migration? How is it used?
7. What’s a stub?

Behat
Behat is a testing framework written in PHP with the
purpose of aiding the writing of Acceptance Tests
using Behaviour Driven Development practices.

Behat
IT’S NOT A FUNCTIONAL
TESTING FRAMEWORK
It’s a tool to verify that a given
system satisﬁes the business
needs of the client.

Behat
Tests are written using a semi-formal language
called Gherkin.
In a nutshell, Behat maps Gherkin sentences -usually
written in English, but it could be any language- to a
class method using to regular expressions.
GHERKIN

Behat
GHERKIN FEATURE

Behat
StoryBDD
Behaviour Driven Development based on User
Stories -aka StoryBDD- is a practice that helps into
bringing a development team’s understanding of
business requirements to the same level of the
client’s.
It does so by setting a formal process for asking
questions about requirements and then expressing
the outcome in a speciﬁc story-format.

Behat
For each new feature, a developer should answer
the following business questions:
• Who will use it?
• What’s its purpose?
• How is that achieved?
StoryBDD

Behat
“I want users to access the site through a dedicated
form”
• Who will use it? a user an anonymous visitor
• What’s the purpose? it’s not stated
• How is that achieved? by using the login form
StoryBDD

“I want an anonymous visitor to be able to access the
site using the login form in order to access her
dashboard containing her aggregated data”
!
In order to see aggregated data on my dashboard
as a site visitor
I need to be able to login on the site
Behat
beneﬁt
role
condition
StoryBDD

Once a feature is deﬁned, you should start writing all
the scenarios that describe it, using the Context-
Action-Outcome pattern:
!
Given I am in on the homepage
When I login
Then I should see “Welcome Back!”
And I should be on my dashboard
Behat
SCENARIOS
context
action
outcome

Behat

Behat
BEST PRACTICES
And I should see “message” inside “.notifications > p:first-child”
!
It’s recommended to avoid technical details in your scenarios as they
increase the coupling between tests and code and make the stories
harder for non-technical readers -such as the client- to understand.
!
And I should see “message” inside notification area

Behat
MINK
Each test gets executed within a context where the
methods invoked by the expressions used in the scenarios
can access to a browser-abstraction API called Mink.
There are several drivers available for Mink such as Goutte,
a headless browser written in PHP that’s able to parse
HTML and XML and Selenium, a full-ﬂedged framework
able to programmatically control real browsers.
Oh and there’s a Drush Driver, too! :)

Behat
MINK CONTEXT

Behat
MINK BEST PRACTICES
The Page object pattern is a way of keeping context ﬁles clean
by separating UI knowledge from the actions and assertions.
In other words, page objects hide the UI and expose clean
services that can be used in the context classes: such
abstraction makes context classes much more clear and helps
avoiding duplication and improving maintainability.

Behat
MINK DRUSH DRIVER

Behat
EXECUTION EXAMPLE
Gherkin
Feature
Behat Mink
Goutte
Selenium Firefox
Test Results

Behat
EXERCISES
1. What’s Behat and what’s its main goal?
2. What’s StoryBDD?
3. What’s a scenario?
4. What languages can be used to write the tests?
5. What questions should be answered by a feature description?
6. What are Mink’s main tasks?
7. What browsers can be used run the tests? What are the main
differences between them?

PHPUnit
PHPUnit is a unit testing framework, it’s part of xUnit
family and it has become the de-facto PHP standard
over the years.

PHPUnit
In software engineering, unit testing is a method by
which individual units of code are tested to
determine if they are ﬁt for use.
A unit is deﬁned as the smallest part of an
application that can be tested in isolation: most of
the time that unit is a class or a method.
UNIT TESTING

PHPUnit
A SUT -an acronym for System Under Test- is deﬁned
as that subset of the software being developed that’s
exercised by a test. It can be a speciﬁc method, and
object or even a set of objects.
DEFINITIONS

PHPUnit
Writing automated tests (not necessarily unit tests only) sets some
goals:
• Improving the quality of the product and its code;
• Increasing developers’s understanding of the SUT;
• Decreasing the damage caused by unexpected bugs;
• Simplifying the creation, maintenance and execution of the tests;
• Documenting the behaviours of the applications’ components.
GOALS

PHPUnit
TYPES OF TESTS
Unit
Integration
Functional
Functional: they operate on the
system from the outside, often by
simulating/employing a browser.
Integration: they check how
multiple components of the
system cooperate when
exercised together.
Unit: they make sure that the
single units of code behave as
expects when exercised in
isolation.

PHPUnit
Should I test before or after development?
Traditionally tests -being them unitary or functional- have always
been executed once the implementation was completed.
Such approach has been upturned by the rise of agile
methodologies, where the emphasis is put on the ability to
respond to change quickly rather than sticking to a plan.
Writing tests ﬁrst makes code inherently testable and therefore
safer to modify and usually produces less verbose and over-
engineered implementations.
METHODOLOGIES

PHPUnit
Should I test outside-in or inside-out?
As soon as development begins, a very important
decision has to be made: should it start from the big
picture and “zoom in” or from one or some of the
details and “zoom out” in order to compose the
macro-functionality at the end?
METHODOLOGIES

PHPUnit
Should I test outside-in or inside-out?
By starting development(and testing) from the outside, the
developer forces himself to think as a client ﬁrst, a method that
helps in making requirements and (temporarily missing)
dependencies more explicit.
Developing “from the inside” on the other hand frees the
developer from the problem of not existing dependencies at
the cost of guessing what’s going to be the right set of things to
implement before reaching the requested level of functionality.
METHODOLOGIES

PHPUnit
Testing status or behaviour?
There are two main ways to verify the correctness of a SUT:
• Testing the state: it’s often sufﬁcient to verify that given an initial
state, once the system is exercised the ending state then
matches the expected one.
• Testing the behaviour: it’s sometimes necessary to look inside
the black box of the SUT in order to check how the system works
during the execution of the tests to make sure all the behavioural
expectations were satisﬁed.
METHODOLOGIES

PHPUnit
TOOLS
PHPUnit provides a number of tools and helpers to aid testing, such as:
• Assertions
• Fixtures
• Database testing helpers
• Stub and mock objects
• Code coverage analysis tools
• Several logging formats
• Integration with several functional testing tools such as Selenium

PHPUnit
EXAMPLE

PHPUnit
PHPUnit features about 80 assertions (of which 40 are
“straight” and 40 are “negated”). It’s up to the developer
to pick the best one for the test in order to have the most
expressive tests and the best error messages possible.
ASSERTIONS

PHPUnit
FIXTURES
When writing tests, one of the most tedious and time
consuming parts is the preparation of the environment
status before the actual execution happens: such status is
known as the ﬁxture of the test .
PHPUnit features quite a big set of methods used to aid in
the creation and sharing of these states, such as:
setUpBeforeClass(), setUp(), assertPreConditions(),
assertPostConditions(), tearDown(), tearDownAfterClass()
and onNotSuccessfulTest(Exception $e).

PHPUnit
DATA PROVIDERS
It can be useful to be able to run
multiple iterations of the same
test on a larger-than-one data
set. Data Providers allows for a
convenient way to do just that.
Such providers are nothing but
a method that returns an array
containing all the combinations
of arguments a tests will be
called with.

PHPUnit
TEST DOUBLES
It could be helpful -or necessary- to replace a SUT’s
dependencies with so-called “test doubles”, a kind of
objects that looks like the dependency but it’s
created with the sole purpose of satisfying testing
needs.
Such a replacement allows to isolate the SUT and
manipulate its inputs as needed, making the test
more effective and focused on the unit alone.

PHPUnit
TEST DOUBLES
There are various kinds of Test Doubles, each serving a specific
purpose. Meszaros and Fowler suggested the following five
definitions:
• Dummy objects are passed around but never actually used.
Usually they are just used to fill parameter lists;
• Fake objects actually have working implementations, but
usually take some shortcut which makes them not suitable
for production (an InMemoryTestDatabase is a good
example);

PHPUnit
TEST DOUBLES
• Stubs provide canned answers to calls made during the test, usually
not responding at all to anything outside what's programmed in for
the test;
• Spies are stubs that also record some information based on how they
were called. One form of this might be an email service that records
how many messages it was sent;
• Mocks are pre-programmed with expectations which form a
speciﬁcation of the calls they are expected to receive. They can throw
an exception if they receive a call they don't expect and are checked
during veriﬁcation to ensure they got all the calls they were expecting.

PHPUnit
TEST DOUBLES

PHPUnit
OPTIONS
PHPUnit features a long list of options to configure its behaviour that’s
accessible both from command line and from a configuration file.
The former is a convenient way to quickly get started and try thing out,
but on the long run the phpunit.xml file is probably a more convenient
choice to configure the tool.
It’s a good practice to commit a phpunit.xml.dist file and (git) ignore the
phpunit.xml one: by doing such thing the other developers will be able
to create a phpunit.xml file tailored to their own needs.
PHPUnit will automatically use phpunit.xml if present or phpunit.xml.dist if
not.

PHPUnit
OPTIONS BEST PRACTICES
It’s a good practice to commit a phpunit.xml.dist ﬁle
and (git) ignore the phpunit.xml one: by doing such
thing the other developers will be able to create a
phpunit.xml ﬁle tailored to their own needs, leaving
the phpunit.xml.dist targeted at automatic builds,
fore example.
PHPUnit is smart enough to automatically use
phpunit.xml if present or phpunit.xml.dist if not.

PHPUnit
There are more tools provided by the community than what’s implemented
by PHPUnit alone. Here’s a short list of some relevant ones:
• Faker - a library that generates fake-yet-plausible data for a handful of
typical use cases such as names, addresses, numbers, dates, images, etc…
• Mockery - a powerful mocks creation library;
• Prophecy - a new, strongly-opinionated mocks creation library;
• PHP-VCR - a library to record and replay HTTP calls in tests;
• TuTu - a simple and ﬂexible HTTP mocking tool.

PHPUnit
EXERCISES
1. What’s Unit Testing?
2. What are the main types of tests?
3. What are the benefits of writing the tests before the code?
4. What’s an assertion?
5. What’s a fixture?
6. What’s a data provider?
7. What’s the difference between a Stub and a Mock?
8. What’s the purpose of the phpunit.xml file?

Drupal Continuous Integration (European Drupal Days 2015)

More Related Content

What's hot (20)

Viewers also liked (6)

Similar to Drupal Continuous Integration (European Drupal Days 2015) (20)

More from Eugenio Minardi (20)

Recently uploaded (20)

Drupal Continuous Integration (European Drupal Days 2015)