Basic CMAKE project snippet

This is a basic Cmake project snippet for x86 linux based applications. It tries to cover most of the most important topics in project development in the minimal but functional form. This minimal approach helps with understanding the topic. The project intentionally contains some bugs in code to show that the verification works.

Folder structure

The project contains of three folders: doc, objLibEx and libEx.

doc

This folder contains the example of Doxygen documentation.

objLibEx

This folder contains an example of the Object library. It contains the main function but doesn't create an application. It is created by the top level CMakeLists.txt file instead.

libEx

This folder contains an example of the library, GoogleTests and library installation.

Development environment

The project can be compile using the Docker-based development environment.

Create Docker image

Important

• Export UID to expose the user id as an environment variable by calling export UID=${UID}¹.

Run the following command to compile and run the complete suite

docker compose build && docker compose up -d

Code Formatting

Code formatting is based on clang-format utility. It uses the ./.clang-format file containing the style configuration. The code formatting operation will not be triggered by build but can be manually/automatically triggered by the IDE.

Configuration in VS Code

Hit Ctrl+Shift+P and search for Open User Settings. Search for format and select Editor: Format On Save.

Static Analysis

Static analysis is done using cppcheck tool. Custom target scheck is handling the static analysis process. This target is a part of the ALL group which means that it'll be run automatically by make when make or make all command is executed.

Static Analysis Usage

• Run make scheck in the ./build folder to start static analysis manually.

SonarQube

SonarQube is a very usefull on-the-fly static analysis tool for your Visual Studio Code.

Unit Tests

Unit tests are using GoogleTest framework. This project uses an installed version of the GoogleTest framework.

Unit Tests Usage

• Run ctest -VV in the ./build folder to start tests.

Memory Leaks Detection

Memory leaks detection is made using Valgrind utility.

Memory Leaks Detection Usage

• Run ctest -T memcheck in the ./build folder to start memory check.

Code Coverage

Code coverage analysis is done using the gcov as the coverage testing utility and the gcovr for presentation.

Code Coverage Usage

• Navigate to the build folder.
• Run cmake -DCMAKE_BUILD_TYPE=Coverage .. to configure coverage build.
• Run make to build the application.
• Run ctest -VV to test the application.
• Run make coverage to get the coverage results.

Installation

Installation is done via the built-in CMake installation mechanism.

Installation Usage

• Navigate to the build folder.
• Run cmake -DCMAKE_BUILD_TYPE=Release .. to configure release build.
• Run make to build the release.
• Run sudo make install to install application.
• Use -DCMAKE_INSTALL_PREFIX=/new/prefix if you want to change the default /usr/local/ installation directory prefix.
• Use -DINSTALL_GTEST=OFF to prevent google tests installation with the application.

The example full build and installation:

mkdir build
cd build
cmake -DCMAKE_INSTALL_PREFIX=/tmp/test-install \
-DCMAKE_BUILD_TYPE=Release -DINSTALL_GTEST=OFF ..
make all
make install

AppImage

The AppImage creation process description can be found on the AppImage documentation pages. The best starting point for the CMake based projects is the build-with-cmake.sh script explained there.

The adapted version of this script has been added to this project as make-appimage.sh.

AppImage Generation

• Run ./make-appimage.sh to generate Appimage of this application.

Useful links

• tutorial
• moder-cmake
• cmake-examples

Footnotes

1. This should be done even if there's an automatic Bash UID read only variable present since it is ignored by the docker. ↩