#!/bin/bash

#This script sanitizes a file in the following manner:
#it reads a pattern and a destFile
#then, all lines in destFILE that can be regexp'd to pattern are removed.
#optionally, a confirm variable is also specified; if so, the script does not prompt the user before overwriting destFILE

#this script takes the following input:
#sanitize.sh pattern destFile [confirm]

#make sure input is valid
[[ "$1" == "" || "$2" == "" ]] && printf "Incorrect input format.\nUse: \t sanitize.sh pattern destFile [confirm]\n" && exit
[[ ! -f "$2" ]] && printf 'destFile - %s - does not exist!\n' "$2" && exit


PATTERN=$1
destFile=$2
saveArray=()
discardArray=()
[[ "$3" =~ [yY][eE][sS]|[yY]|1|[tT][rR][uU][eE] ]] && confirm=0 || confirm=1

#go through each line, adding to either SAVE or DISCARD as necessary
while read -r line; do
    [[ ! $line =~ $PATTERN ]] && saveArray+=("$line") || discardArray+=("$line")
done < "$destFile"

#confirm output
if [[ $confirm -eq 1 ]]; then
  printf "Discarding the following:\n"
  printf '%s\n' "${discardArray[@]}"
else
  printf 'Discarding %d line(s).\n' "${#discardArray[@]}"
fi

#confirm the user is ok with this
response="y"
[[ $confirm -eq 1 ]] && read -r -p "Proceed? (y/N): " response
if [[ "$response" =~ [yY][eE][sS]|[yY] ]]; then
  echo "Removing unwanted lines..."
  printf '%s\n' "${saveArray[@]}" > "$destFile"
fi