git-fetcher v1.0.0-beta-13

#Usage

iojs fetch.js --help to show help
iojs fetch.js git@github.com:pawlik/repo_with_rich_history /home/work/i_will_work_here

Note

This code is still in a mess phase. I focused on making it usable ASAP, still need to tidy it up and figure out how to design automatic tests

Fetcher

The main goal of this tool is to provide a fast way for cloning big repositories. It's focus is to bring code as soon as possible, so you can continue in build script inside your repo (this is achieved by making so called shallow repo).

The problem with shallow repo is - it's pain in the ass for developers. Only few commits are available in git log (in case of fetcher only one), and git blame gives you wrong answers. So we try to unshallow repo for you.

How is it achieved

There are two scenarios.

1. very first run ever

• With the very first run it clones shallow repo from URL, then script exits. So you can call command to build your app.
• before the scripts exits, though, it calls detached process which
  • creates tmp mirror repo
  • unshallows this repo

  • reassigns remotes on just cloned repo, and unshallows it

2. next run

• your tmp mirror is here, so it just updates it
• clones your repo from tmp instead url
• reassigns remotes, so you can push your work to origin (instead of tmp mirror)

What it can do:

• when there's no tmp repo mirror it'll fetch shallow, repo, then create (also shallow) tmp repo. With --shallow it will then unshallow tmp, and then unshallow fetched app.

todo

See gh issues