Introduction

For most projects, the source code is tracked using a version control system; it typically serves as input for the build system, which transforms it into objects, libraries, and executables. In certain cases, we use the build system to generate source code during the configuration or build steps. This can be useful to fine-tune the source code based on information gathered in the configuration step, or to automatize an otherwise error-prone mechanical generation of repetitive code. Another frequent use case for generating source code is to record information about the configuration or compilation for reproducibility. In this chapter, we will illustrate various strategies to generate source code, using the powerful tools provided by CMake.