How to create a project in GitLab using git?

4

We can easily create a project in GitLab and "synchronize" it with git :

$ git clone [email protected]:user/teste.git

But what about backwards?

Create the folder and the local project, and send as a "new" in GitLab ?

    
asked by anonymous 30.08.2018 / 19:45

1 answer

4

I think you're talking about sending set a remote repository to an existing project.

Zero Local Repository

I usually do this:

  • I create the project folder
  • Initialize the local repository inside the folder I just created. The command is git init
  • After that, add the files that will be sent in the first commit. Run git add . to select all files (optional, if you want to do one by one you can, just an example)
  • Commit. git commit -m "primeiro commit"
  • Add the place of origin. git remote add origin [email protected]:user/teste.git
  • Make the initial push to your server: git push -u origin master

On the fly, you are sending location data from an existing repository on your machine.

Summary of what I did above through BASH:

>>> mkdir projeto
>>> cd projeto
>>> git init 
>>> git add .
>>> git commit -m "primeiro commit"
>>> git remote add origin [email protected]:user/teste.git
>>> git push -u origin master

Local repository with existing source

If you are trying to use an existing repository that has another source, you will not be able to use the git remote add origin command. In this case, use git remote set-url origin [email protected]:user/teste.git , but only if you already have a repository with origin , and do not create one from scratch.

Then just do git push -u origin master .

Explaining the commands

git-init

Starts a git pository. When you run this command, you are saying that, from now on, that directory will have GIT version control.

git-add

Adds the contents of one or more files to the change index. When you do git add readme.txt , you are saying that changes made to readme.txt will be saved in the next commit.

I like to use git add and add the changes I made to a topic.

For example:

I created a button on the user registration page that verifies that the email is valid. For this, I created the button in criar-usuario.html and in criar-usuario.js I made the ajax call that will make this query.

You can do this to process the changes:

 >>> git add views/criar-usuario.html js/criar-usuario.js

Then when you do commit , only the two files will be added to the list of modifications.

When you use the dot ( git add . ), you are saying that all files will be added.

git commit -m

Process the changes reported in git add and adds to the change history. In the case above, the -m flag indicates that you are reporting the commit message directly in the command call. If you only use git commit , a window that usually uses vim or nano (common code editor on Linux) will open for you to put the commit message.

Commit messages need to describe what you did.

Using the previous example:

>>> git add js/cadastro-usuario.js views/cadastro-usuario.html
>>> git commit -m "Adicionando botão de consulta de e-mail"

git remote add origin url_do_repostiório

You are saying that the place where GIT changes will be saved remotely will be the url informed.

This command is generally used when you create a repository with git init , because it just adds version control. The git remote add origin tells the server where you will save the changes.

git push -u origin master

git push sends the commit information you created to the server. The commits are being added to the queue and you can see which ones were not sent through the command git status .

-u is usually used in the first commit and origin master indicates that branch of source will be master . master is usually the default% of GIT%.

I will not dwell too much on Branchs. Read about it here:

30.08.2018 / 20:22