Contents

#!/usr/bin/env ruby

class GitCommitSize
  attr_reader :list

  def initialize args=nil
    @commits = `git rev-list --all #{args}`.split
    @list    = {}
  end

  def blobs_in_commit treeish
   `git diff-tree -r -c -M -C --no-commit-id #{treeish}`.
     split("\n").
     map { |blob| blob.split[3] }
  end

  def size_in_bytes obj
    return 0 if obj =~ /\A0+\z/
    `git cat-file -s #{obj}`.to_i
  end

  def bytes_in_commit treeish
    blobs = blobs_in_commit treeish
    total = blobs.map { |blob| size_in_bytes blob }.reduce(:+)
  end

  def walk
    @commits.map { |c| @list[c] = bytes_in_commit c }
  end

  def sorted
    @list.reject { |k, v| v.nil? }.sort_by { |k,v| v }
  end

  def avg
    (@list.values.compact.reduce(:+).to_f / @commits.count).round
  end

  def median
    list = @list.values.map(&:to_i).sort
    n = list.count
    case @list.count.odd?
    when true
      @list[(n+1)/2]
    when false
      ( list[n/2-1] + list[n/2] ) / 2.0
    end
  end
end

if __FILE__ == $0
  git = GitCommitSize.new
  count = git.commits.count
  puts "Walking #{count} commits."
  git.walk
  result = git.sorted
  numsep = /(?<=\d)(?=(?:\d{3})+(\.\d+)?\z)/
  pp result
  puts "Average: %s bytes" % git.avg.to_s.gsub(numsep, ?,)
end

Version data entries

1 entries across 1 versions & 1 rubygems

Version	Path
git_commit_size_scraper-1.0.0	lib/git_commit_size_scraper.foo