Shows a sorted list (decreasing by size) of objects in a repository's full commit history.

git-big-objects.sh

#!/bin/bash

# Shows you the largest objects in your repo's entire history. Just call this script after
# you cd into the repo you'd like to run it on.
# 
# To save the results to a file, just redirect stderr to /dev/null, and then redirect
# stdout to the filename of your choice. Here's an example:
# 
# ./git-big-objects.sh 2>/dev/null > sorted-git-objects.txt
 
declare -a on_exit_items

function abort_all()
{
    # Get our process group id
    PGID=$(ps -o pgid= $$ | grep -o [0-9]*)
    # Kill it in a new new process group
    setsid kill -- -$PGID &>/dev/null
    on_exit 1
}

function on_exit()
{
    # kill any child processes that we started
    for i in "${on_exit_items[@]}"; do
        eval $i &>/dev/null
    done
    exit $1
}

function add_on_exit()
{
    local n=${#on_exit_items[*]}
    on_exit_items[$n]="$*"
}

trap abort_all SIGINT EXIT

# create some temporary files and mark them for deletion in the trap callbacks
ALL_SHAS="$(mktemp)"
add_on_exit rm -rf ${ALL_SHAS}
BIG_OBJS="$(mktemp)"
add_on_exit rm -rf ${BIG_OBJS}

(>&2 echo "creating list of repository objects")

git rev-list --objects --all | sort -k 2 2> /dev/null > "${ALL_SHAS}" &
git gc &> /dev/null && git verify-pack -v .git/objects/pack/pack-*.idx | egrep "^\w+ blob\W+[0-9]+ [0-9]+ [0-9]+$" | sort -k 3 -n -r 2> /dev/null > "${BIG_OBJS}" &

# wait for the previous 2 command to finish
wait

# show all objects from biggest to smallest
for sha in $(cut -f 1 -d\  < "${BIG_OBJS}"); do
    echo $(grep ${sha} "${BIG_OBJS}") $(grep ${sha} "${ALL_SHAS}") | awk '{print $1,$3,$7}'
done;