99
votes

Is there a Python way without using a subprocess to clone a git repository? I'm up for using any sort of modules you recommend.

10
gitpy, i guess it would be called - SilentGhost
@SilentGhost: you mean this gitpy? github.com/vmalloc/gitpy from ryaari.com/blog/?p=9 - VonC
Looks like there's GitPython (pypi.python.org/pypi/GitPython, gitorious.org/git-python) which I don't think has a clone method, but I'll bet you could add one... internally it's going to have to call git clone anyway though. - Cascabel
[Dulwich][1] is a pure-Python implementation of Git that does not fork at all. Be aware that it's still under development, so it may be buggy. [1]:samba.org/~jelmer/dulwich - Mark Lodato

10 Answers

71
votes

There is GitPython. Haven’t heard of it before and internally, it relies on having the git executables somewhere; additionally, they might have plenty of bugs. But it could be worth a try.

How to clone:

import git
git.Git("/your/directory/to/clone").clone("git://gitorious.org/git-python/mainline.git")

(It’s not nice and I don’t know if it is the supported way to do it, but it worked.)

147
votes

Using GitPython will give you a good python interface to Git.

For example, after installing it (pip install gitpython), for cloning a new repository you can use clone_from function:

from git import Repo

Repo.clone_from(git_url, repo_dir)

See the GitPython Tutorial for examples on using the Repo object.

Note: GitPython requires git being installed on the system, and accessible via system's PATH.

24
votes

My solution is very simple and straight forward. It doesn't even need the manual entry of passphrase/password.

Here is my complete code:

import sys
import os

path  = "/path/to/store/your/cloned/project" 
clone = "git clone gitolite@<server_ip>:/your/project/name.git" 

os.system("sshpass -p your_password ssh user_name@your_localhost")
os.chdir(path) # Specifying the path where the cloned project needs to be copied
os.system(clone) # Cloning
10
votes

Github's libgit2 binding, pygit2 provides a one-liner cloning a remote directory:

clone_repository(url, path, 
    bare=False, repository=None, remote=None, checkout_branch=None, callbacks=None)
10
votes

Here's a way to print progress while cloning a repo with GitPython

import time
import git
from git import RemoteProgress

class CloneProgress(RemoteProgress):
    def update(self, op_code, cur_count, max_count=None, message=''):
        if message:
            print(message)

print('Cloning into %s' % git_root)
git.Repo.clone_from('https://github.com/your-repo', '/your/repo/dir', 
        branch='master', progress=CloneProgress())
8
votes

For python 3

First install module:

pip3 install gitpython

and later, code it :)

import os
from git.repo.base import Repo
Repo.clone_from("https://github.com/*****", "folderToSave")

I hope this helps you

4
votes

With Dulwich tip you should be able to do:

from dulwich.repo import Repo
Repo("/path/to/source").clone("/path/to/target")

This is still very basic - it copies across the objects and the refs, but it doesn't yet create the contents of the working tree if you create a non-bare repository.

3
votes

Pretty simple method is to just pass the creds in the url, can be slightly suspect though - use with caution.

import os

def getRepo(repo_url, login_object):
  '''
  Clones the passed repo to my staging dir
  '''

  path_append = r"stage\repo" # Can set this as an arg 
  os.chdir(path_append)

  repo_moddedURL = 'https://' + login_object['username'] + ':' + login_object['password'] + '@github.com/UserName/RepoName.git'
  os.system('git clone '+ repo_moddedURL)

  print('Cloned!')


if __name__ == '__main__':
    getRepo('https://github.com/UserName/RepoYouWant.git', {'username': 'userName', 'password': 'passWord'})
3
votes

You can use dload

import dload
dload.git_clone("https://github.com/some_repo.git")

pip install dload
1
votes

This is the sample code for gitpull and gitpush using gitpython module.

import os.path
from git import *
import git, os, shutil
# create local Repo/Folder
UPLOAD_FOLDER = "LocalPath/Folder"
if not os.path.exists(UPLOAD_FOLDER):
  os.makedirs(UPLOAD_FOLDER)
  print(UPLOAD_FOLDER)
new_path = os.path.join(UPLOADFOLDER)
DIR_NAME = new_path
REMOTE_URL = "GitURL"  # if you already connected with server you dont need to give 
any credential
# REMOTE_URL looks "[email protected]:path of Repo"
# code for clone
class git_operation_clone():
  try:
    def __init__(self):
        self.DIR_NAME = DIR_NAME
        self.REMOTE_URL = REMOTE_URL

    def git_clone(self):

        if os.path.isdir(DIR_NAME):
            shutil.rmtree(DIR_NAME)
        os.mkdir(DIR_NAME)
        repo = git.Repo.init(DIR_NAME)
        origin = repo.create_remote('origin', REMOTE_URL)
        origin.fetch()
        origin.pull(origin.refs[0].remote_head)
  except Exception as e:
      print(str(e))
# code for push
class git_operation_push():
  def git_push_file(self):
    try:
        repo = Repo(DIR_NAME)
        commit_message = 'work in progress'
        # repo.index.add(u=True)
        repo.git.add('--all')
        repo.index.commit(commit_message)
        origin = repo.remote('origin')
        origin.push('master')
        repo.git.add(update=True)
        print("repo push succesfully")
    except Exception as e:
        print(str(e))
if __name__ == '__main__':
   a = git_operation_push()
   git_operation_push.git_push_file('')
   git_operation_clone()
   git_operation_clone.git_clone('')