Ruby script to cleanup large files from your git repository

git_cleanup.rb

#!/usr/bin/env ruby

# Ruby shows you the largest objects in your git repo's pack file & offers you to remove them automatically
# Based on http://stubbisms.wordpress.com/2009/07/10/git-script-to-show-largest-pack-objects-and-trim-your-waist-line/ by Antony Stubbs

# Use this to fetch all branches locally first:
# for branch in `git branch -a | grep remotes | grep -v HEAD | grep -v master`; do
#   git branch --track ${branch##*/} $branch
# done

require 'shellwords'

# list all objects including their size, sort by size, take top 20
puts " > Listing objects"
objects = `git verify-pack -v .git/objects/pack/pack-*.idx | grep -Ev "non delta|chain length|git/objects" | tr -s " " | sort -k3gr | head -n 20`.lines.map(&:chomp)
puts " > Building file index"
files = Hash[`git rev-list --all --objects`.lines.map {|l| l.chomp.split(' ', 2)}]

to_remove = []
puts " > largest files:"
objects.each do |object|
  sha, _, size, compressedSize = object.split(' ')
  file = files[sha]
  exists = File.file?(file)
  puts " %5d kB (compressed: %5d kB) - %s%s" % [size.to_i/1024, compressedSize.to_i/1024, file, ("*" if exists)]
  to_remove << file unless exists
end

print " > do you want to remove these #{to_remove.size} files (not present in current branch)? (y/N): "
if gets.chomp == 'y'
  files = to_remove.map {|f| Shellwords.escape(f)}.join(' ')
  cmd = "git filter-branch --tag-name-filter cat --index-filter 'git rm -r --cached --ignore-unmatch #{files}' --prune-empty -f -- --all"
  puts " > #{cmd}"
  system cmd
  cmd = "rm -rf .git/refs/original && git reflog expire --expire=now --all && git gc --aggressive --prune=now"
  puts " > #{cmd}"
  system cmd
  puts " > now you need to 'git push origin --force --all && git push origin --force --tags'"
  puts " > and tell your teammates to 'git rebase'"
else
  puts ' > ok, bye.'
end