In order to run examples, you will need to Install NodeGit first.

Return to all guides


This guide explains how to clone a repository, and in the case of failure, attempt to open the existing path.

View example source

Requiring NodeGit

In the guides directory, we like to keep our NodeGit relative to the project root.

var NodeGit = require("../../../");

However, in your project you will most likely be using the following command:

var NodeGit = require("nodegit");

Clone URL

The first argument to the clone method is a URL.

In this example we’re going to clone one of our test repositories from GitHub. You could easily substitute this with any valid http or https Git repository URL.

var cloneURL = "";

Clone path

The second argument to the clone method is a path.

Ideally your application will clone a repository into the same folder path regardless of how or where you execute it from. Paths are relative to the current working directory in NodeGit, so you will need to normalize it first.

This is very simple in Node:

var localPath = require("path").join(__dirname, "tmp");

Now this tmp directory will be created along side your script, no matter how or where you execute it from.

Clone options

The third argument to the clone method is an optional simple object.

var cloneOptions = {};

If you are using HTTP the OS X issue below does not affect you.

GitHub certificate issue in OS X

Unfortunately in OS X there is a problem where libgit2 is unable to look up GitHub certificates correctly. In order to bypass this problem, we’re going to passthrough the certificate check.

Note: this is not a problem with Windows or Linux

cloneOptions.fetchOpts = {
  callbacks: {
    certificateCheck: function() { return 1; }

Invoking the clone method

You can easily invoke our top-level Clone as a function passing along the three aforementioned arguments.

var cloneRepository = NodeGit.Clone(cloneURL, localPath, cloneOptions);

Notice how we store the return value from Git.Clone. This is a Promise to represent the asynchronous operation. It offers finer control flow by allowing us to capture errors and fallback if there is a clone failure.

Handling clone failure

A naive way to handle a clone failure is to try opening the same path. Clones will most commonly fail when the directory already exists. We can define a function to attempt opening in this case.

var errorAndAttemptOpen = function() {

This will be called as part of the Promise resolution in the final step.

The Promise chain

Lastly in our clone operation, we’ll assemble a Promise chain to handle errors and work with the Git.Repository instance result.

  .then(function(repository) {
    // Access any repository methods here.
    console.log("Is the repository bare? %s", Boolean(repository.isBare()));