Filesystem tree content comparison using md5sum

md5sum-cmp.sh

#!/bin/bash
# Filesystem tree content comparison using md5sum
#
# Usage:
#    $ md5tree-cmd hash <folder 1> > my_hash_file_1
#    $ md5tree-cmp hash <folder 2> > my_hash_file_2
#
#    $ md5tree-cmp only <hash_file_1> <hash_file_2>
set -e
cmd=$1

function hash() {
    find "$PWD/$1" -type f -exec md5sum "{}" + 
}

function only() {
    A=$1
    B=$2

    tmp=$(mktemp -t md5tree-cmp)
    echo "* tmp: $tmp"
    Ahash="${tmp}-Ahash"  # Just the hashes in A, sorted
    Bhash="${tmp}-Bhash"  # Just the hashes in B, sorted
    Dhash="${tmp}-Dhash"  # The diff of the hashes
    Aonly="${tmp}-Aonly"  # Unmatched hashes
    Bonly="${tmp}-Bonly"  # 
    trap "tmp_cleanup $tmp" EXIT TERM INT
    
    echo "* A: $A"
    echo "* B: $B"
    
    echo "* Collect hashes from files"
    cat "$A" | cut -d' ' -f1 | sort > "$Ahash"
    cat "$B" | cut -d' ' -f1 | sort > "$Bhash"

    echo "* Diff hashes"
    diff -u "$Ahash" "$Bhash" > "$Dhash" || true
    
    echo "* Collect the adds/removes from the diff"
    cat "$Dhash" | grep "^\-[^-]" | sed -e 's/-//g' > "$Aonly"
    cat "$Dhash" | grep "^\+[^+]" | sed -e 's/+//g' > "$Bonly"
    echo " - Files only in A: $(cat $Aonly | wc -l)"
    echo " - Files only in B: $(cat $Bonly | wc -l)"
    
    echo "* Writing all lines only in A to: ${A}_ONLY ..."
    grep "$A" --file "$Aonly" > "${A}-ONLY"

    echo "* Writing all lines only in B to: ${B}_ONLY ..."
    grep "$B" --file "$Bonly" > "${B}-ONLY"
    
    echo "* Done"
}

function tmp_cleanup() {
    rm $1*
}

if [ "$cmd" == hash ]; then
    hash $2
elif [ "$cmd" == only ]; then
    only $2 $3
else
    echo "md5sum-cmp"
    echo "Usage:"
    echo "    $ md5sum-cmp hash <folder>  >  <output>"
    echo "    $ md5sum-cmp only <output1> <output2>"
fi