What is a monorepo?
A repository containing two or more logical projects (e.g., an iOS client and a web application).
These projects may be unrelated, loosely coupled, or connected by other means (e.g., through a dependency management tool).
A repository is large in several ways:
- Number of commits
- Number of branches and/or tags
- Number of tracked files
- Size of tracked content (measured by viewing the repository's .git directory)
Git sparse checkout
Cloning an entire Git repository, organized as a monorepo, is extremely inefficient.
Git provides a feature that allows you to retrieve only a portion of a repository. This allows for extremely efficient loading.
However, this only works with Git version 2.25 or later.
clone
Clone without files. Get only the git history without files.
git clone --filter=blob:none --no-checkout <GITHUB_URL>
spare init
Enables sparse checkout feature.
git sparse-checkout init
sparse set
Specifies which directory to retrieve.
server/scheduler retrieves the scheduler directory within the server directory.
The same applies to client/android.
git clone --filter=blob:none --no-checkout <GITHUB_URL>
Then it is structured as follows:
.
├── .git/ <-- (Hidden folder where the actual Git data is stored)
└── client/
| └── android/
| ├── (Android project files and folders...)
| └── ...
└── server/
└── scheduler/
├── (Scheduler-related files and folders...)
└── ...set doesn't load anything, it just specifies what to load. The actual loading happens in checkout.
checkout
Actually download and set up the directory specified above.
git clone --filter=blob:none --no-checkout <GITHUB_URL>
