Git：如何在不添加合并提交的情况下从另一个存储库中提取

Question

This is a project where multiple developers are working. Assume that all work is done on the master branch, and that there is a remote repo that has the following history today:

A --- B --- C

Today I went ahead and cloned this repository into one of my sandboxes (let's call this work_sb01 ). I did some work, and have committed my changes locally (there could be just one commit or multiple commits). The Git history for this sandbox looks as shown below:

A --- B --- C --- X

Two days from today, I clone the same repository into another sandbox (let's call it work_sb02 ). As expected, the repo HEAD has moved, so the history might now look something like this:

A --- B --- C --- D --- E

I need to be able to pull the changes that I made in work_sb01 and bring them into work_sb02 . I'm currently doing this with the following command:

cd <path_to_work_sb02>
git pull <path_to_work_sb01> master

This gets the job done, but it creates a merge commit. I have now a requirement to avoid adding merge commits into the project Git log. How can I pull the changes from work_sb01 and make them the most recent commit in the history? In other words, the local history for work_sb02 should look like the one shown below:

A --- B --- C --- D --- E --- X

Answer 1

Here is the right way to do this from work_sb02:

cd <path_to_work_sb02>
git remote add sb01 <path_to_work_sb01> 
git fetch sb01 
git checkout master 
git rebase --onto master master sb01/master

Answer 2

If doing in path_to_work_sb02 a pull --rebase path_to_work_sb01 does not give you the right order, you could... consider doing the opposite:

• pull --rebase path_to_work_sb02 from path_to_work_sb01 path.
• rename path_to_work_sb01 as path_to_work_sb02
• change its remote origin URL
• go on working from there.