Get the X largest files from your repository

getLargestFilesListFromRepository.sh

#!/bin/sh
#
# Get the X largest files from your repository
# You have to pass the number of file you want
# The result is stored in the folder $folder
# foundList.txt is the list of largest files which are not deleted yet
# deletedList.txt is the list of largest files which are already deleted. They cannot be found in your workspace
#
#

# Folder to store the result and the temp files
folder="./cleanedFiles"
# The X first files to find
filesNumber=$1

# Is it a bare repository ?
if [ -e ".git" ];
then
    bareRepository=false
else
    bareRepository=true
fi
        
echo ""
echo "You're workin' on a git bare repository : $bareRepository"
echo "Search for the first $filesNumber largest files in the repository history, sorted by size"
echo ""

if [ "$#" != 1 ];
then
    echo "Error : only one argument is needed : the files number"
    exit 1
fi
if [ -z "$1" ];
then
    echo "Files number argument not supplied."
    exit 1
fi

rm -Rf $folder
mkdir -p $folder/temp
echo ""
echo "The folder $folder is created"
# List of SHA1 of all the files in the repo (available in the workspace and history)
git rev-list --objects --all | sort -k 2 > $folder/temp/allFilesSHA1.txt

echo ""
echo "Git garbage collector :"
git gc 

if [ "$bareRepository" = true ] ; then
    pathIdx="objects/pack/pack-*.idx"
else
    pathIdx=".git/objects/pack/pack-*.idx"    
fi

# First largest objects list
git verify-pack -v $pathIdx | egrep "^\w+ blob\W+[0-9]+ [0-9]+ [0-9]+$" | sort -k 3 -n > $folder/temp/allBigObjects.txt
cat $folder/temp/allBigObjects.txt | tail -$filesNumber > $folder/temp/bigObjects.txt
echo ""
echo "$filesNumber first largest objects list created"

# Largest files list 
for SHA in `cut -f 1 -d\ < $folder/temp/bigObjects.txt`; do
    echo $(grep $SHA $folder/temp/bigObjects.txt) $(grep $SHA $folder/temp/allFilesSHA1.txt) | awk '{ print $1,$3,$7 }' >> $folder/temp/largeFiles.txt
done;
echo ""
echo "Largest files list created"

# Largest files sorted list
cat $folder/temp/largeFiles.txt | sort -k2 -nr > $folder/temp/largeFiles.sorted.txt
echo ""
echo "Largest files sorted list created"

while read line; do
    path=$(echo $line | awk '{ print $3 }')
    
    git cat-file -e HEAD:$path > /dev/null 2>&1
    if [ $? -ne 0 ];
    then
        # File doesn't exist anymore in the last commit
        echo $line >> $folder/deletedList.txt
    else
        # File can be found in the last commit
        echo $line >> $folder/foundList.txt        
    fi
done < $folder/temp/largeFiles.sorted.txt 

echo "Open the foundList.txt and the deletedList.txt files in the $folder folder"