Creating empty directories in the Git repository is a daily business task, and everybody did it a few times.
The example from real life may be a directory for cache, logs, build artifacts, etc.
There are two different ways how to get the result: creating a .gitignore with specific content or creating
an empty file, usually .gitkeep in the directory. I will explain why the latter approach is better.
The first approach is to create a .gitignore file with the following content:
*
!.gitignore
According to the documentation Git will process .gitignore files moving from the root level to nested directories and overriding rules that coincide.
In other words, the lowest .gitignore file in the directory will always win.
The downside of having a few .gitignore files in the project is that it gets harder to fetch the full list of empty directories.
This may be especially a problem for the people who just started working on the project and don’t know the codebase well enough.
I find an approach with creating a .gitkeep to be more useful.
This approach allows having exactly one .gitignore file in the project, with as many empty directories as needed.
There is one additional line that excludes .gitignore everywhere:
# .gitignore
!.gitkeep
To create an empty directory one needs to put an empty .gitkeep file in it:
$ mkdir -p /path/to/the/directory
$ touch /path/to/the/directory/.gitkeep
Then, add the directory in question to the .gitignore:
# .gitignore
!.gitkeep
/path/to/the/directory/*
Pay attention to the * at the end – this will affect only files inside the directory, but not the directory itself.
The trick is that * does not cover so-called dotfiles and ignores .gitkeep.