python-git 3.9

Automate boring git tasks

# python-git

Automate the boring git stuff with python

## Motivation

Whenever I wanted to see the status of all my git repos I have to fire up the git-cmd.exe shell on windows, navigate to each folder and then do a git status. I have to do this both at home and at work.

But I got quickly tired of it. So I decided to make this tool to give me a quick report so I can see what is ahead and what’s behind and what’s ahead at a glance. In short, what needs attention so as to avoid those troubling merge conflicts.

## Requirements

The only requirements in terms of software is send2trash which helps take care of cleaning up stuff. Other thing you need is a computer with git either accessible from the command line (which means its in your system path) or as a standalone file somewhere in your system. If you’re working on PC without installation rights, you can use a portable git and python-git will work just fine.

You can get a portable git version from [here](https://git-scm.com/download/win)

Just unzip it and place it somewhere on your disk. Later (during initialization), you’ll need to tell python-git where this file is located.

## Installation

pip install https://codeload.github.com/immensity/python-git/zip/master

pip install python-git --upgrade

## Usage

Upon successful installation, the below command should return a blank screen

import pygit

To use python-git, you have to tell it exactly two things, depending on your system setup.

1. The location of your git repositories. You may do this by passing it a list of strings on the command line. Each string represents a full path name to single directory. You may also just provide a single directory which holds multiple git repositories and pygit will grab all the repositories for your.

1. The location of a git executable. This only applies if git is not accessible from your system cmd. That is, git is
2. not in your system path. More on this below.

If you have a master directory that holds multiple git repositories, pygit can also take the full path name of this master directory and then index the git repositories it finds there. It won’t index those directories that are not git repos.

It is also possible to tell pygit not to index certain directories by specifying the starting string of the directory name. This is referred to s a rule. Directories matching such rules will not be touched.

To initialize pygit, run

pygit.initialize()

In case things change (perhaps you moved folders around) and you want to reset your folders, just run the set_all() command again

To see all available repositories run

pygit.show_repos()

This command shows a list of all available repositories in the format

repository_id: repository_name: repository_directory_path

To load a particular repository use

pygit.load(repo_id_or_name)

where repo_id is a string-valued id assigned to that particular repo. The first value in the show_repos command’s output.

The load() command returns a Commands object for that repo, which provides a gateway for issuing git commands on the repository

Operations that can be performed on Commands object are shown below.

Commands().fetch() # perform fetch.

Commands().status() # see status

Commands().add_all() # stage all changes for commit

Commands().commit(message='minor changes') # commit changes. Press enter to accept default message

Commands().push() # perform push action

Commands().pull() # perform pull request

Commands().add_commit() # add and commit at once

### Batch Operations

Pygit provides some functions for performing batch operations on your repositories.

pygit.load_multiple(*args)

loads a set of repositories. You could decide to only load only 2 of 10 repositories. Perhaps you need to perform similar actions on those two alone. As an example

pygit.load_set(“2”, “5”)

returns a generator of Commands objects for repositories 2 and 5. Afterwards you can use a for loop to iterate over the repos like below

for each in pygit.load_set(“2”, “5”):

each.add_commit()

pygit.all_status()

performs a status command on all your repositories. The result is written to a text file. The text file opens automatically. The name of the file shows the date and time of the status request. All batch status request is written to its a separate file. Call it a snapshot of your repo status if you will Those items which are out of sync with their remote counterpart (by being ahead or being behind) are also highlighted as needing attention.

pygit.pull_all()

performs a pull request on all your repositories at once. Its return value is None.

pygit.push_all()

performs a push action on all your repositories at once. Its return value is None.

pygit.load_all()

returns a generator of Commands object for every repository.

## To do

Add git-bash.exe

Implement Commands.branch()

Write tests

Run test after importation to make sure every other thing works fine.

Define an update function that updates the repo dictionaries for the case when a new repo is added but the overall directory structure remains unchanged.

Git search pathsep

git check locations

C:Program FilesGitcmdgit.exe C:Program Files (x86)Gitcmdgit.exe C:Program FilesGitcmdgit.exe C:UsersChidimmaAppDataLocalProgramsGitcmdgit.exe

https://stackoverflow.com/questions/19687394/python-script-to-determine-if-a-directory-is-a-git-repository

http://gitpython.readthedocs.io/en/stable/

https://simpleisbetterthancomplex.com/tips/2017/08/11/django-tip-21-redirects-app.html

https://simpleisbetterthancomplex.com/tutorial/2017/08/20/how-to-use-celery-with-django.html

https://realpython.com/asynchronous-tasks-with-django-and-celery/