# git-dumper

A tool to dump a git repository from a website.

## Usage

```
usage: git-dumper.py [options] URL DIR

Dump a git repository from a website.

positional arguments:
  URL                   url
  DIR                   output directory

optional arguments:
  -h, --help            show this help message and exit
  --proxy PROXY         use the specified proxy
  -j JOBS, --jobs JOBS  number of simultaneous requests
  -r RETRY, --retry RETRY
                        number of request attempts before giving up
  -t TIMEOUT, --timeout TIMEOUT
                        maximum time in seconds before giving up
```

### Example

```
./git-dumper.py http://website.com/.git ~/website
```

## Install the dependencies

```
pip install -r requirements.txt
```

## How does it work?

The tool will first check if directory listing is available. It it is, then it will just recursively download the .git directory (what you would do with `wget`).

If directory listing is not available, it will use several methods to find as many files as possible. Step by step, git-dumper will:
* Fetch all common files (`.gitignore`, `.git/HEAD`, `.git/index`, etc.);
* Find as many refs as possible (such as `refs/heads/master`, `refs/remotes/origin/HEAD`, etc.) by analyzing `.git/HEAD`, `.git/logs/HEAD`, `.git/config`, `.git/packed-refs` and so on;
* Find as many objects (sha1) as possible by analyzing `.git/packed-refs`, `.git/index`, `.git/refs/*` and `.git/logs/*`;
* Fetch all objects recursively, analyzing each commits to find their parents;
* Run `git checkout .` to recover the current working tree