git commit is used to record changes in a git repository.
Git is a distributed version control system. In order for any version control system to work, we need to have a way of indicating/marking/creating these various versions of a file. This is where a commit becomes useful.
A commit is the mechanism used to mark a specific point as a point of reference for the changes that have taken place. I.e, once you have made changes in your local directory and added them to the index, commit will allow you to have a marker identifying all the changes that were added to the index at that particular point in time. Besides creating a marker, a commit also automatically adds those changes to your local repository. In Git, a commit is a snapshot of the git repository at that particular point in time. The snapshot being the entire contents of every file in the repository. However, to be efficient, if a file has not changed, git just links to the previous version of the file instead of storing it again. But how do we create a commit?
git commit is used to record changes in a git repository by creating commits. For git commit to work, files to be committed must have been added to a special area called a index/staging area. This allows you to choose the specific files you want committed.
-
git add new-file- adds new file to the staging area -
git add modified-file- adds modified file to the staging area -
git rm file-to-be-removed- removes file from the local directory and staging area. -
git commit -a- automatically adds changes from the modified files to the staging area and also remove files from the staging area that have been removed from the local directory.
git commit - allows you to record changes to the files, that are in the staging area, to the local repository. It automatically starts the default editor and allows you to write your commit description in the editor. Thus it gives you the opportunity to write better commits which would be more meaningful to you later especially when you find yourself in a fix and would really want to know the exact changes you made in each commit. Usually, it is preferred to write a small summary of about 50 characters or less. Thereafter, you skip one line then list all the changes that are taking place in that commit. The commit is performed when you save and close the editor.
The terminal will then output
[master 6a683b1] <commit message>
1 file changed, 2 insertions(+), 2 deletions(-)
master stands for the branch on which the commit has been done.
6a683b1 are the first 7 characters for the hash(a 40 character name used to uniquely identify a commit).
<commit message> stands for your unique commit message.
Thereafter, the number of files, together with the changes performed on each are listed.
git commit file1
Assume you have created three files file1, file2 and file3, made changes in them and added them to the index as shown below.
touch file1 file2 file3
git add .
git commit file1
The command git commit file1 will only commit file1 while leaving the other changes eg. file2 and file3 in the staging area. Thus, it allows you to commit only certain changes from the staging area.
git commit -m "commit message" - performs a commit with the value in quotes as the commit message. You should edit it to be meaningful to you and all the members of your team so that they too get to understand the changes in that commit. However, you should note that git commit without the -m flag is preferred because it allows you to write your commit by automatically starting the default editor for you. Remember, in the editor, you have more space on which to write on.
git commit -q -m "commit message" - the -q or --quiet flag performs a commit while suppressing the commit summary message.
git commit --dry-run -m "commit message" - the dry-run option doesn't perform the commit but allows you to see what would happen if a real commit was done. It will show you the files that will be commited and the ones that will not. The files that will not be committed are the modified files and the untracked files. The modified files are the ones that had been added to the staging area before changes were made to them. The untracked files are the ones that have never been added to the staging area.
git commit --file=<file> - performs a commit while getting the commit message from the file <file>.
git commit --author=<author> - performs a commit while overriding the commit author with <author>.
git commit --date=<date> - performs a commit while overriding the author date with <date>.
git commit -e --file=<file> - performs a commit while allowing you to edit the messages taken from the file <file>.
Checkout the manual for a full list of options and flags.