#!/bin/sh
# Delete files under current or given path which exist elsewhere as listed in the locate database
# Matches first by name, then by checksum (currently inefficiently via md5)
find "$@" -size +50M -type f | while read filepath
  do target="$(basename "$filepath" | synct-unarchive)"
  highlight "$filepath"
  locate -b "$target" | while read existing
    do test "$(realpath "$target")" != "$(realpath "$existing")" -a -f "$existing" -a "$(md5sum "$existing" | cut -d\  -f1)" = "$(md5sum "$filepath" | cut -d\  -f1)" && echo "Found duplicate at $existing" && rm -vi "$filepath" && break
    done
  done