Get the big blobs from your git repostory, in Python

git-diet.py

import subprocess, os
 
def get_big_blobs(git_repo, nitems=10):
    # Query the git repo for large items
    os.chdir(git_repo)
    big_blobs = subprocess.Popen(
        ('git verify-pack -v .git/objects/pack/pack-*.idx | '
         'grep -v chain | sort -k3nr | head --lines={1} -').format(git_repo, int(nitems)),
        shell=True,
        stdout=subprocess.PIPE)
    
    # Parse the result
    for line in big_blobs.stdout:       
        # Line gives us sha, object, uncomp_size, comp_size etc...
        tokens = line.split()  
        sha = tokens[0]
        uncompressed_size = int(tokens[2]) / 1024
        compressed_size = int(tokens[3]) / 1024
        
        # Find the items in the git repository tree
        item = subprocess.Popen('git rev-list --all --objects | grep {0}'.format(sha),
                     shell=True, stdout=subprocess.PIPE).stdout.read()
        item = item.strip('\n').split()[1]
        print uncompressed_size, item