What does git rebase mean?

git rebase means: redefining the repository status of the branch; when performing the rebase operation, git will extract the modifications on the branch to be rebased from the common ancestor of the two branches, and then rebase the branch to be rebased. The branch points to the latest commit of the base branch, and finally the changes just extracted are applied to the back of the latest commit of the base branch.

The operating environment of this tutorial: Windows 7 system, Git version 2.30.0, Dell G3 computer.

git rebase, as the name suggests, is to redefine (re) the starting point (base), that is, to redefine the repository status of the branch.

1. Submission node diagram

First, let’s feel what rebase is doing through a simple submission node diagram

Two branches, master and feature, where feature is at submission point B The branch pulled from master

has a new commit M on master, and two new commits C and D on feature

At this time, switch to the feature branch. Executing the following command is equivalent to merging the master branch into the feature branch

git checkout feature
git rebase master

//这两条命令等价于git rebase master feature

The following picture is the commit node diagram after rebasing. Explain its working principle:

• feature: base branch to be changed, current branch
• master: base branch, target branch

Official explanation: When performing a rebase operation, git will start from The common ancestor of the two branches begins to extract the modifications on the branch to be rebased, then points the branch to be rebased to the latest commit of the base branch, and finally applies the modifications just extracted to the back of the latest commit of the base branch.

Explanation with examples: When executing git rebase master on the feature branch, git will extract the modifications on the feature branch starting from the common ancestor B of master and featuer, that is, the two commits C and D, extract first arrive. Then point the feature branch to the latest commit of the master branch, which is M. Finally, the extracted C and D are connected to M, but this process is to delete the original C and D and generate new C’ and D’. Their submission content is the same, but the commit id is different. Naturally, feature eventually points to D’.

Popular explanation (important!) : rebase, rebase, can be directly understood as changing the base. The feature branch is a branch pulled from B of the master branch, and the base of feature is B. And if master has a new commit after B, it is equivalent to using the new commit on master as the new base of the feature branch. The actual operation is to save the submissions of the feature after B, then delete the original submissions, then find the latest submission location of the master, and then connect the saved submissions (new node with new commit id), so that the base of the feature branch is quite Yu became M instead of the original B.

There is also a very straightforward explanation. The key to the rebase command is actually to understand the "base". git rebase is to commit the difference between the base branch and the current branch (after the bifurcation point) ) is obtained, and then resubmit the differential submissions one by one after the latest submission point of the "base branch", and finally point the HEAD pointer of the current branch to the latest submission point

---

2. Actual git submission example

The submission record is constructed according to the above diagram, as shown in the following figure: (ABM is the master branch line, and ABCD is the feature branch line. Here, the master changes color and branches out. This does not affect understanding. It means two Just two lines per branch!)

At this time, execute git rebase master on the feature branch

After the rebase is completed, ABCD is the original feature branch line , ABMC'D' is the new feature branch line, ABM is the master branch line (no change)

---

3. Recommended usage scenarios

There are so many things to do, but this is actually the most important thing. Different companies and situations have different usage scenarios, but in most cases the recommendations are as follows:

• Use rebase when pulling the latest code from the public branch, that is, git pull -r or git pull -- Rebase, but one disadvantage is that after rebase, I don’t know which branch my current branch was first pulled from, because the base has changed. (If you use merge, there will be a meaningless commit record "Merge... to...")

• When merging code on a public branch, use merge. (If you use rebase, then if other developers want to see the history of the main branch, it will not be the original history. The history has been tampered with by you)

Recommended study: "Git Tutorial》

The above is the detailed content of What does git rebase mean?. For more information, please follow other related articles on the PHP Chinese website!