- Loading...
$ git config --global user.name "Full Name"
$ git config --global user.email "full.name@company.com"
$ git config --global core.editor "vim"
$ git config --global http.proxy 'PROXY-HOSTNAME:PROXY-PORT'
Substitute PROXY-HOSTNAME
with the hostname of the HTTP(S) proxy server, and PROXY-PORT
with the port of the HTTP(S) proxy server. When using the credential manager on Windows, the proxy must be specified scheme first (e.g. http://).
$ git config --global alias.<name> '<command>'
For example, if you want to type git co
instead of git checkout
you would run:
$ git config --global alias.co 'checkout'
Glad you asked! Please see the Aliases page.
$ git clone <URL>
Make sure you have uploaded an SSH key to your GitHub account and then run:
$ git clone git@github.com:path/to/repo.git
For example, to clone the jdk repository over SSH:
$ git clone git@github.com:openjdk/jdk.git
If you don't want to type git@github.com
every time, add an entry to your SSH configuration.
A common way of structuring local repositories is by following the domain name and path convention. For an example, see below:
$ tree . ├── git │ └── github.com │ ├── edvbld │ │ ├── jdk │ │ └── loom │ └── openjdk │ ├── jdk │ ├── loom │ └── valhalla └── hg └── hg.openjdk.java.net ├── code-tools │ └── jtreg └── jdk └── jdk
Yes, this is supported, supply a second argument to git clone
stating the directory name of the local repository:
$ git clone <URL> <DIRECTORY>
$ git checkout -b <NAME-OF-BRANCH>
If you are using Git version 2.24 or newer you can also use the command: git switch --create <NAME-OF-BRANCH>
A common way of structuring your local branches is to name them after the issue they correspond to, for example JDK-8237566.
$ git branch JDK-8237566 * JDK-8149128 JDK-8077146 master
The currently checked out branch has an asterisk ("*") next to it.
$ git log --format=oneline --graph --all
$ git checkout <NAME-OF-OTHER-BRANCH>
If you are using Git version 2.24 or newer you can also use the command: git switch <NAME-OF-OTHER-BRANCH>
Yes, this is supported. If you prefer to have one local repository per issue you are working on, then you would have a local repository and a local branch per issue you are working on. For example:
$ git clone https://github.com/<USERNAME>/jdk JDK-8123456 $ cd JDK-8123456 $ git checkout -b JDK-8123456
To create the first local repository representing the work on issue JDK-8123456. To create a second local repository for JDK-8654321, run:
$ git clone https://github.com/<USERNAME>/jdk JDK-8654321 $ cd JDK-8654321 $ git checkout -b JDK-8654321
$ git push --set-upstream <REPOSITORY> <LOCAL-BRANCH>
For example, if the repository you want to push corresponds to the remote named origin and your local branch is named JDK-8123456, you would run the following command:
$ git push --set-upstream origin JDK-8123456
If you are using the Skara CLI tools then and have the branch you want to publish currently checked out, then you can just run the command git publish
$ git merge <OTHER-BRANCH>
$ git log --graph --pretty=oneline --first-parent --abbrev-commit master..HEAD
First ensure that you have configured your full name, email and editor. You must then add the files that you want to be part of the commit (even files that are already tracked by Git):
$ git add path/to/file1 path/to/file/2
You can inspect the changes that will be part of the commit by running the following command:
$ git diff --staged
You can then create the commit by running the following command:
$ git commit
The above command will open your editor where you willl the commit message. If the commit message only is a single line then you can pass it directly to commit command: git commit -m 'Your commit message'
.
$ git show <COMMIT>
Yes, but you will be mutating history. If you have already pushed the commit, then other people might depend on the commit you are about to mutate. If you have only pushed the commit to your personal fork and you are not currently collaborating with someone, then it is fine to mutate the history (you yourself are the only one who will be affected). If you have pushed the commit to a project repository, then it is not fine to mutate history, since you cause problem for other contributors.
In many cases, for example when working on a pull request, you can just make an additional commit and push it. The Skara integrate command will squash (collapse) the commits in the pull request into a single commit before integrating it, so the final result will be as if you only had made one commit.
If you still want to mutate history and update the last commit, first add the files you want to add to the commit:
$ git add path/to/file1 path/to/file2
Then run the following command to amend the HEAD
commit:
$ git commit --amend --no-edit
Go to https://github.com/settings/tokens and and click on "Generate new token"
A personal fork is a copy of another repository with one major difference:
Assuming you have a local clone of your personal fork and you are using the Skara CLI tools:
$ git sync
If you also want to sync your personal fork and your local clone of your personal fork and you are using the Skara CLI tools you can run:
$ git sync --fast-forward
If you do not have the Skara CLI tools installed and you have a local clone of your personal fork, you need to the following steps:
$ git remote add upstream <URL-FOR-ORIGINAL-OPENJDK-REPOSITORY> $ # for each branch you want to sync $ git checkout <BRANCH> $ git pull upstream <BRANCH> $ git push origin <BRANCH>
A pull request is a way to suggest that some changes from a personal fork should be incorporated into the original repository the personal fork was created from. Reviewers can comment upon and need to approve a pull request before it can be integrated. The concept of a pull request is very similar to OpenJDK's concept of "RFR" emails - both are used to suggest changes and offer a way for reviewers to provide feedback on the suggested changes.
A draft pull request is a pull request that is not yet ready for review. Draft pull requests are primarily used to share work in a early stage or run additional testing before declaring the pull request ready.
--draft
flag to gh pr create.
--draft
to flag to git pr create
.git pr set --no-draft.
git pr set --title='New title'.
The body, or description, of a pull request is the initial comment in the pull request. Can be compared to the body of the initial "RFR" email.
git pr set --body.
This depends on the project - some projects (typically those in a prototype phase) give project Committers direct push access, but most opt to use pull requests for contributions.
Yes, JDK Committers (and above) are allowed to push directly to the openjdk/jdk-sandbox repository. See the documentation for the how to name your branches to avoid conflicting with other Committers' branches.
$ ssh-keygen -t ed25519 -C github -f filename/for/the/private/key
The public key will be stored next to the private key and have the same as the private key but will have the suffix .pub
.
Note: pick a strong password for your key. You will not have to type your password every time you want to use your SSH key if you use the SSH agent.
You use the ssh-agent. An ssh-agent should already be running if you are using GNU/Linux, macOS or Windows 10 (since version 1803). To add a key to the ssh-agent
, run ssh-add
:
$ ssh-add /path/to/the/private/key
If you want the key to exist in the ssh-agent
only for a limited time, use the -t
(for timeout) option:
$ ssh-add -t SECONDS /path/to/the/private/key
$ ssh-add -l
$ ssh-add -d /path/to/public/key
Note: if the only difference in filenames between the private key and the public key is that the public key ends in .pub
, then you can use the path to private key in the above command as well.
Yes, by adding a host alias for github.com in your SSH configuration file ~/.ssh/config
(C:\Program Files\Git\etc\ssh\ssh_config
on Windows). Add the following lines to the SSH configuration file:
Host github.com github gh User git Port 22 Hostname github.com
You can now clone from GitHub using the aliases listed after Host
:
$ git clone github.com:openjdk/jdk $ git clone github:openjdk/jdk $ git clone gh:openjdk/jdk
Yes, by adding a ProxyCommand
directive in your SSH configuration file ~/.ssh/config
(C:\Program Files\Git\etc\ssh\ssh_config
on Windows). The value for ProxyCommand
is operating system dependent, since different programs are available on different operating systems. Please follow the instructions specific for your operating system below:
Add the following lines to ~/.ssh/config
:
Host github.com github gh User git Port 22 Hostname github.com ProxyCommand nc -X connect --proxy PROXY-HOSTNAME:PROXY-PORT %h %p
Substitute PROXY-HOSTNAME
with the hostname of the HTTP(S) proxy server, and PROXY-PORT
with the port of the HTTP(S) proxy server.
Add the following lines to ~/.ssh/config
:
Host github.com github gh User git Port 22 Hostname github.com ProxyCommand /usr/bin/nc -X connect -x PROXY-HOSTNAME:PROXY-PORT %h %p
Substitute PROXY-HOSTNAME
with the hostname of the HTTP(S) proxy server, and PROXY-PORT
with the port of the HTTP(S) proxy server. Note that the full path to nc
must be used since Homebrew might install the GNU version of nc
(which does not support the -X
flag).
Add the following lines to C:\Program Files\Git\etc\ssh\ssh_config:
Host github.com github gh User git Port 22 Hostname github.com ProxyCommand C:\Program Files\Git\mingw64\bin\connect.exe -H PROXY-HOSTNAME:PROXY-HOST %h %p
Substitute PROXY-HOSTNAME
with the hostname of the HTTP(S) proxy server, and PROXY-PORT
with the port of the HTTP(S) proxy server. Note that the program connect.exe
is part of Git for Windows, if you install Git via some other mechanism then you might have to install connect.exe
yourself.
Yes, add the following to your ~/.bashrc
:
git_branch() { BRANCH="$(git symbolic-ref --short HEAD 2>/dev/null)" if [ "${BRANCH}" != "" ]; then printf " (${BRANCH}) " else printf "" fi } PS1="${PS1}\$(git_branch)"
For more information on how to customize your Bash prompt, see the Arch Linux wiki for a good reference. For more Git information in your prompt, see contrib/completion/git-prompt.sh.
Yes, on all of GNU/Linux, macOS and Windows this is provided out of the box when you install Git. If you want to install auto-completion manually, run the following commands:
$ curl -o ~/.git-completion.bash https://raw.githubusercontent.com/git/git/master/contrib/completion/git-completion.bash $ echo 'source $HOME/.git-completion.bash' >> ~/.bashrc
Yes, add the following to your ~/.zshrc
:
autoload -Uz vcs_info precmd_vcs_info() { vcs_info } precmd_functions+=( precmd_vcs_info ) setopt prompt_subst PROMPT="$PROMPT\$vcs_info_msg_0_" zstyle ':vcs_info:git:*' formats ' (%b) ' zstyle ':vcs_info:*' enable git
For more information, see the Pro Git book on Zsh.
Yes, Zsh ships with auto-completion for Git out of the box. Enable auto-completion by adding the following line to ~/.zshrc
:
$ echo 'autoload -Uz compinit && compinit' >> ~/.zshrc
Yes, see the package Magit.
Yes, see the package Forge.
Yes, see the plugin vim-fugitive. If you want to see the files that have been changed compared to HEAD
in the "gutter" (sign column), see vim-gitgutter.
No.
Yes, Visual Studio code has built-in support for Git. For details, see the Visual Studio Code documentation.
Yes, see the plugin GitHub Pull Requests and Issues.