Cleaning old CCLab Experiment Manager backups

clean_CCLabEM_backups.sh

#! \bin\bash

## This script is to help clean up the experiment manager backups
cd /media/PatricksHD/CC9_Backups

touch removal_list.txt

# Compressed tarballs of the pydio user data directory are taken twice a day with a cron job
# The filenames follow the naming convention pydio_userdata_backup_YYYY-MM-DD_HH-MM.tar.bz
# Thus, this script assumes that the date of a backup can be extacted by splitting
# the filename on underscores and extracting the fourth field, and that 2 files exist per date.

all_dates=$(find . -name 'pydio_userdata_backup*' | cut -d '_' -f 4 | sort -u)

# Loops over the dates we found backups for
for d in $all_dates; do

  # Find the names of files that have the current YYYY-MM-DD string in $d in their name
  x=$(find . -name "*$d*")

  ## If 2 backups from one date are found, proceed to compare them to see if they're identical.
  # There might not be 2, as older files might have already been deleted
  if [[ $(echo "$x" | wc -l) = 2 ]]; then

    # diff only reports when files are differ. Output will be empty when files are the same
    # -z returns true if the string is null, so this checks if the files are the same
    # according to diff.
    # If they are they same, put the first one in the list of files to be deleted
    if [[ -z $(echo "$x" | xargs -n 2 diff) ]]; then
      echo $x | cut -d' ' -f 1 >> removal_list.txt
    fi

  fi
done;

# Interactively ask if you want to delete each file
for f in $(cat removal_list.txt) ; do
  rm  -i "$f"
done

rm removal_list.txt