Popping the smallest possible repo and problems with git subtree on small clone

Question

I am trying to use git subtree pull --squash to merge another repo as a subdirectory in my repository called docroot . This happens in the Jenkins workspace, and I'm trying to use a shallow clone of my repo because it is pushed to a remote repo to run the application. It’s great when cloned with a full story, so I try to keep it as small as possible when it is pushed out. When I try to release git subtree pull on a shallow clone of my repository (cloned using git clone --depth=1 ), I get this error:

Impossible squash-merge: docroot was never added.

If I do not use --squash , it starts, but there are conflicts that need to be resolved, which is not suitable for a CI job that needs to be run often. My goal is to pull out the smallest number of repos that can be saved to disk. Several instances of this application will be displayed, so I want to keep it as small as possible. I will also look at how to make the repo smaller, but the largest size reduction I've seen is possibly using a shallow clone. Maybe there is a better way to do this? Thanks in advance.

Answer 1

I propose to solve this problem using another repository, docroot , like the git submodule in the original repository. This assumes that repo docroot is independent and does not share code with the parent repository.

This way you can have a shallow clone of the parent repo and should not run into conflicts with the full extraction of docroot .