1 Introduction to Git
Git is a free , open source distributed version control system for handling any project, small or large, with agility and efficiency.
1.1 The difference between Git and SVN
|
|
SVN |
GIT |
| Is it distributed |
non-distributed |
distributed |
| storage method |
file storage |
Metadata storage |
| branch |
A branch is just another directory |
pointer |
| global version number |
have |
without |
| Content integrity |
|
better than SVN |
2 Installation and configuration
| $ apt-get install git $ git config --global user.name “Your Name” $ git config --global user.email “[email protected]” |
3 local warehouse
The Git directory is divided into three areas:
working dir: the directory for operations
Temporary storage area (stage): In the .git directory, the add content is stored
Version area (HEAD): In the .git directory, information and content of each version are stored
3.1 Create a warehouse
3.2 Add files to the working directory
Add files on the working directory . Use git status to view the modification status
3.3 Add changes to the staging area
3.4 Submit changes to the version area
Use commit to submit changes, -m to bring submission instructions. After submitting, you can use the log to view the version record.
3.5 Modify the file again and submit
make changes to the file
Add and modify change1
Submit change change1
3.6 Undo changes
3.6.1 Restoring the workspace from the version area
Modify the workspace first, and then use the chekout -- <file> command to restore from the version area
3.6.2 Restoring the workspace from the staging area
Modify the file and add it to the temporary storage area, and then modify the file. At this time, you can use the chekout -- <file> command to restore the workspace from the temporary storage area.
3.6.3 Restoring the workspace and staging area from the version area
The command git reset --hard <commitid> restores the workspace and staging area from the version specified in the version area. <commitid> is the commit id, or version pointer.
Submit id: can be the first few digits of the id, but at least 4 digits
Pointer: HEAD current pointer
HEAD^ previous version; HEAD^^ represents previous version, and so on
HEAD~<N> previous versions
3.6.4 Summary
git reset --hard <commitid> restores the workspace and staging area from the version specified in the version area
git checkout -- <file> is to restore from the last add or commit
[Note] git checkout does not add --, it is to switch branches
3.7 Delete files
Use the git rm command to delete the file, and then git commit to submit it.
3.8 Branches
Branching in Git is done with pointers, so it is particularly efficient
3.8.1 Create a branch
git checkout -b <branch name> Create and switch to branch
git branch View the branch (* before the branch indicates the current branch)
3.8.2 Automatically merge branches
Make some changes on the dev branch first
merge dev on master
delete branch
3.8.3 Merge conflicts
If both branches have their own changes, the merge operation cannot be automated. Conflicts need to be resolved manually. There are two methods for merging branches: merge and rebase. Rebase is recommended.
3.8.4 merge
Modify readme.txt on dev
Modify readme.txt on master branch
Merge dev branch on master, prompt conflict
At this point, the readme.txt file becomes
Modify readme.txt to
Add and commit after resolving conflicts.
3.8.5 rebase
manual:
l Use git rebase to merge branches
l Modify the conflict. After the file is modified, use git add to add it
l Execute git rebase --continue to continue with the next conflict resolution until all conflicts are resolved
The branch before merging is as follows
After executing git rebase, each commit of the current branch will be cancelled, and they will be temporarily saved as pathch (in the .git/rebase directory), and then the branch will be updated to the new origin. Finally apply the saved patch to the branch.
The final effect, all commits are on a branch, as if no branch existed. The following figure shows the difference between rebase and merge.
4 Remote warehouse
4.1 Create a remote repository
4.2 Associate Remote Repository
4.3 Upload to remote warehouse
The format is git push –u <remote warehouse name> <local branch name>[:<remote branch name>]
[Note] The -u parameter is required for the first upload, and the -u parameter is not required for subsequent updates and modifications
[Note] If the remote branch name is the same as the local branch name, you can omit: and the following remote branch name
The dev branch can be seen on the remote repository.
4.4 Remote conflict
Because when a new remote warehouse was created, there were already changes (README.md).
First download with fetch, then merge, and finally push
| # git fetch origin master # git log -p master origin/master # git merge origin/master # git push origin master |
4.5 Fork branch
fork from an existing repository
4.6 Clone branch
The command git clone downloads the remote branch to the local. If it is not the master branch, you need to add -b <branch name>
4.7 Push to remote
Modify the local repository first
Use git push <remote name> <local branch name>:<remote branch name>
Changes can be seen on the remote repository
4.8 Pull request (PR)
To merge the branch modifications from the fork to the source branch, a pull request needs to be submitted, and the person in charge of the source branch agrees to merge it.
Click New Merge Request on the iSource page
Fill in PR information
Click Merge on the source branch to modify the automatic merge
5 Summary
5.1 General Process
5.2 Common git commands
git init
Turn the current directory into a manageable git repository and generate hidden .git files.
git add <file>
Add file modifications to the staging area.
git rm <file>
delete xx file
git commit -m “XX”
Commit changes -m is followed by comments.
git status
Check warehouse status
git diff <file>
View what the XX file has modified
git log
View version history
git reflog
View the version number id of the history
git reset --hard <commitid>
Recover from the version area. <commitid> can be a version pointer or a commit ID
HEAD current version
HEAD^ previous version. There can be multiple ^s here, each representing a rollback of a version
HEAD~<N> means roll back N versions
Commit ID can be the unique first few digits of Commit ID (at least 4 digits)
git checkout -- <file>
Undo all modifications made to the file in the workspace.
git checkout -b <branch name>
Create and switch to branch
git branch <branch name>
create branch
git branch
View all current branches
git branch -d <branch name>
delete the dev branch
git checkout <branch name>
switch to branch
git merge <branch name>
Merge the branch on the current branch
git stash
Hide the current work and resume work when the scene is restored later
git stash list
View a list of all hidden files
git stash apply
Recover hidden files, but the contents are not deleted
git stash drop
Delete Files
git stash pop
Recovering files also deletes files
git remote [-v]
View remote library name
-v list details
git remote add origin <remote repository>
Associate a remote library
git clone [-b <branch name>] <remote repository>
Clone from a remote repository.
-b <branch name> indicates which branch to clone. If not, the default master
git push [-u] <remote name> [<local branch name>][:<remote branch name>]
Push the local branch to the branch corresponding to the remote library
-u This parameter needs to be added for the first push, and it is not required in the future
If the remote branch name is omitted, it means that the local branch will be pushed to the remote branch with which there is a "tracking relationship" (usually the two have the same name)
If the local branch name is omitted, it means to delete the specified remote branch
Branch development principles:
If the local dev branch is modified, you need to switch to the master branch first, merge the modifications of the local branch back to the master (git merge dev), and then push the merged content to the master on the master; then switch back to the local branch, Then merge the contents of master back to the branch, and then continue to develop on the branch.
Development Process:
// view local branch
git branch
// Display all local and server branches
git branch -a
// Display the mapping relationship between the local branch and the server branch
git branch -vv
// switch branch (one -b parameter less than creating a branch)
git checkout {{branch_name}}
// Create a new branch, the code of the new branch comes from the current branch
git checkout -b [branch name]
// push the local branch code to the remote server, if the remote server does not have the branch, it will be created automatically;
git push origin [remote branch name]
// pull remote branch code to local current branch (recommended)
git pull origin master
If you only want to use git pull, you may need to establish a track relationship, then use
git branch --set-upstream-to=origin/<remote branch_name> <local branch_name> to
output the result: Branch h5_dev set up to track remote branch h5_dev from origin .
// Merge the local master branch to the current branch, such as currently in the dev branch, merge the code on the master to the dev branch
git merge master
// Merge the remote master branch to the current branch
git merge origin/master
// delete the local branch
git checkout {{another_branch}}
git branch -d {{local_branch_name}}
// delete remote branch
git push origin --delete {{branch_name}}
#Assume there are two local branches : master and dev, use dev for local development
# Step1. dev.commit
git status #Current branch , modified files
git diff #specific modification
git commit -m "{COMMIT_MESSAGE}" . # 提交
# Step2. master.update.merge-dev
git checkout master #Switch to master branch
git pull origin master:master # 更新master
git merge dev #merge the modifications of the dev branch
git push origin master:master #Submit code to remote Git repository
# Step3. dev.update
git pull origin master: dev #update dev
git checkout dev #Switch back to dev for the next development
Common commands:
- git config --global user.name
- git config --global color.status auto
- git config --global color.diff auto
- git remote -v// View the git path of the remote branch
- git clone [email protected]/tv.git
- git status // View the status of the current version (modified or not)
- git add xyz // Add the currently modified file to the staging area or track new files (git add. Add all modified files to the staging area)
- git commit -m '{{update_msg}}'
- git rm xxx // remove the file from the current tracking list and delete it completely
- git rm -cached xxx// delete only in the temporary storage area, keep the file in the current directory, no longer track
- git mv old_name new_name// rename file
- git log
- git log -n
- git log --stat
- git show b4f44fe
- git diff
Code volume statistics
git log --author="pWX490102" --since ==2018-01.26 --until=2018-03-24 --pretty=tformat: --numstat | awk '{ add += $1; subs += $2; loc += $1 - $2 } END { printf "added lines: %s, removed lines: %s, total lines: %s\n", add, subs, loc }' -
Force rollback command:
git reset --hard e1da2c37bdaf3df52b9cf105ec7a7bd83f488bc1 (last commit number)
Pull code from master branch to new branch
git branch --set-upstream-to=origin/master appMarket-pyq
Pull the remote branch down to the local and establish an association track:
git checkout --track origin/jiangbo_dev
Branch jiangbo_dev set up to track remote branch jiangbo_dev from origin.
Switched to a new branch 'jiangbo_dev'