Summary and Schedule

In science, we rely on establishing the behaviour and accuracy of experimental devices before making actual measurements, for example through calibration. Equally, theoretical calculations make predictions that inform experiment design and data analysis decisions, and thus we also rely on establishing the correctness of these calculations. The same is true of the software we use to simulate, take, and analyse the data collected through experiments - it is an theoretical calculation or experimental device just as much as any other. We must therefore throughly test our software’s accuracy and reliability in order that the scientific results we derive from its use are not compromised.

In this lesson, we introduce methods and tooling for unit testing C++ code. Our “units” here are the lower level components of the code, specifically individual functions and/or classes. By defining test cases, test suites and test fixtures, we’ll see how we can integrate testing as a standard part of the software development workflow. The GoogleTest framework will be introduced to build these elements, a commonly used testing system in both science and industry that saves us a lot of otherwise boilerplate coding. We’ll finish by looking at how unit testing provides a foundation for higher level integration and regression testing, and other useful tooling for aspects unit testing cannot catch.

The actual schedule may vary slightly depending on the topics and exercises chosen by the instructor.

Software installation

These instructions set out how to obtain and install the software on Linux or macOS. It is assumed that you have:

  • access to the Bash or Zsh shell on a fairly modern Linux or macOS system
  • a text editor such as Vim or Emacs, or an IDE such as VSCode
  • sufficient disk space (~300MB) to store the software and outputs

You do not need root/administrator access.

C++ Compiler

As we will be coding in C++ (and the C++17 standard), we will need a compiler. We’ll take this from the system for reliability, so it can be installed through the appropriate package management system.

Install GCC through your system’s package manager.

Install Xcode from the App Store.

This lesson has not been designed to run on Windows. We would recommend using the Windows Subsystem for Linux, and following the instructions for Linux in the rest of the lesson.

Pixi

As we move through the course, we will need additional tools to help us write, build, and run our tests efficiently. To do this, we will use the Pixi package manager to provide an isolated development environment with the software we need.

Open a terminal and run:

BASH 

curl -fsSL https://pixi.sh/install.sh | sh

or, if you don’t have curl available:

BASH 

wget -qO- https://pixi.sh/install.sh | sh

More details on what these do are on Pixi’s installation page.

If you have Homebrew installed, then it’s recommended to install Pixi with this. Open a terminal and run:

BASH 

brew install pixi

Otherwise, you should install Pixi directly:

BASH 

curl -fsSL https://pixi.sh/install.sh | sh

More details on what these do are on Pixi’s installation page.

Discussion

Why Pixi?

If you’re familiar with installing software, you might be asking why we’re using Pixi, another package manager, on top of the system ones (or similar tools like Homebrew). The answer is in two parts:

  1. Consistency: we can lock in the versions of the tools used to avoid issues in the exercises
  2. Simplified workflow: We’ll use Pixi’s tasks to simplify the configure/build/test workflow so we can concentrate on the testing concepts rather the tooling.

It is also very similar in concept to (and uses packages from) Conda which you may be familiar with already.

Software Setup

Open a terminal and check that you have the following commands available:

BASH 

g++ --version

BASH 

clang++ --version

and

BASH 

pixi --version

Download the project zip file and unzip it somewhere suitable. Change into the the unpacked ccptest-test/ base directory and try setting up the Pixi environment (NB: we have shown $, the prompt, to distinguish the commands to be run from the output you should see. You don’t need to type this!):

BASH 

$ cd ccptest-test
$ ls
CMakeLists.txt pixi.lock      README.md      test
LICENSE.md     pixi.toml      src
$ pixi shell

This will download the packages needed and drop you into a shell with the environment setup to use these. Depending on your network connection and system, this may take several seconds. You should see your prompt changed to reflect this new environment:

BASH 

$ pixi shell

(ccptepp-test) $

To check you have the tools needed, run the following commands and check they complete with the following information:

BASH 

(ccptepp-test) $ cmake --version
cmake version 4.2.0

CMake suite maintained and supported by Kitware (kitware.com/cmake).

(ccptepp-test) $ ninja --version
1.13.2

(ccptepp-test) $ gcovr --version
gcovr 8.2

Copyright (c) 2013-2024 the gcovr authors
Copyright (c) 2013 Sandia Corporation.
Under the terms of Contract DE-AC04-94AL85000 with Sandia Corporation,
the U.S. Government retains certain rights in this software.

To exit the Pixi environment, use exit to get back to your default terminal environment:

BASH 

(ccptepp-test) $ exit

Saving session...
...saving history...truncating history files...
...completed.
$

You can also simply close the terminal.

More detailed information on Pixi use will be provided in each exercise as needed, though in all cases we will be working in the directory ccptepp-test/.