Back to [[git:main-git|git main page]]

If you clone a git repo inside you git, the best way to handle it is to use modules (sub modules).

==== Create a submodule ====

To define a submodule you can use :

<code bash>
git submodule add <git URL> <path>
</code>  
Where **git URL** is the URL of the git repo and path is where this git repo is stored in the main git.

For example we are working in the git repo **submodules_main** and we would like to use an external git at URL https://gitlab.com/bnzr/python_control_coppeliasim.git
<code bash>
mkdir external
git clone https://gitlab.com/bnzr/python_control_coppeliasim.git 
git submodule add https://gitlab.com/bnzr/python_control_coppeliasim.git external/python_control_coppeliasim/
</code>  

==== List all submodules ====

We can list the submodules with :
<code bash>
git submodule foreach --quiet 'echo $name'
</code>  

Information on submodules can be found in  **.gitmodules** and **.git/config** of the main project git

==== Content changes in submodules ==== 

If we add the changes, commit and push , the content of submodules will not been added in the main project git, but information on how to retrieve it will be stored. In our example we push the changes :
<code bash>
git add --all
git commit -m "add python_control_coppeliasim external git repo"
git push
</code>

==== Clone git repo with submodules ====

In another computer we can simply clone the main project repo and we will see that there is nothing in the submodule. The content of the submodules can be pulled by 
<code bash>
git clone <git URL> 
cd <git repo main folder>
git submodule update --init
</code>
If we know when cloning the main project git that there are submodules we can pull them at the start with **--recursive** keyword:
<code bash>
git clone <git URL> --recursive
</code>

if after pulling the main repo, some new submodules are created and if they are empty, their content can be retrieved with :
<code bash>
git submodule foreach git restore --staged .
git submodule foreach git restore .
</code>

==== Working with a particular submodule (not all) ====

If the main git repo contains many submodules, it can be interesting to work only on one particular submodule.
For example, if after cloning the main git repo the submodule contains nothing, it can be cloned independently of the others submodules. 
Suppose the path to submodule (from main repo path) is /path/to/mysubmodule, to clone it type :
<code bash>
git submodule init /path/to/mysubmodule
git submodule update -- /path/to/mysubmodule
</code>
If the submodule is not in the correct branch, these commands can be used to list the branches and set to the correct one (ex master)
<code bash>
git -C /path/to/mysubmodule branch
git -C /path/to/mysubmodule checkout master
</code>
In general a git **command** on submodule can be acheived from main git repo by :
<code bash>
git -C /path/to/mysubmodule command
</code>