The choice to use a date based assignment is better for organization and it should lead to no duplication, but it would not hurt to build error handling for duplicate names as a cautionary tool. Do you know how to use bash to check for existing files?

Then here is a good example of filename error handling and implementing a formatted sequence number to force creation of a new valid name.

#!/bin/bash



filedir="${HOME}/testscripts/emptydir/"

filename="testfile"

fileext=".txt"



# format the date as year-month-date so it can be easily sorted

dt=$(date +"%y-%m-%d")



# loop through from 1 to 999

#  adding each three digit number to the filename

#  until a number that does not exist if found,

#  then make the file with the available name



# use the seq command to increment a number, it is being formatted

#  as a 3 digit number from 1 to 999

for seqnum in $(seq -f "%03g" 1 999)

do

  # set a filename including the directory, base name,

  #  date, sequence number and the extension

  fullfile="${filedir}${filename}-${dt}-${seqnum}${fileext}"



  # test the see if the file exist, first test to see

  #  if it does not exist

  #

  # the ! means NOT

  # the -f tests to see if the file exists

  # so "! -f" means check to see if the file does not exist

  #

  if [ ! -f ${fullfile} ]

  then

    # since the file does not exist, make it

    touch ${fullfile}

    echo "  ${fullfile} has been created"



    # exit the loop since the file has been created

    break

  fi

done

Honestly it does not match the proper printf floating point syntax, I have been using it for a while after I found it on the net. I am assuming that the synax is stating to have leading zeros while forcing whole numbers to be three characters long.