Command line: create file with new filename if the filename already exists

Question

I was wondering if it was possible have a command that creates one file, but every time the command is executed, it wouldn't overwrite the file created in the previous execution.

For example: touch test1.txt would create 1 file called test1.txt. But the next time I execute it, I would like the new file to be called test2.txt, or something like that. So without overwriting the already existing file. In a way it could be executed multiple times without problems.

Don't misunderstand me, I'm not trying to create multiple files with one command.

Thanks in advance!

Answer 1

The easiest solution would be to add a timestamp to the filename and not use a single digit.

The easiest method to create an empty file would be touch test$(date +%Y%m%d-%H%M%S) and that would result in a file named test20110802-170410. A 2nd time test* will get a newer timestamp so it will result in 2 files.

Answer 2

I doubt a general command exist to do that, but you can think of something like this:

create() {
  read prefix number suffix < <(sed -r 's/(.*)([0-9]+)\.(.*)$/\1 \2 \3/' <<<"$1")
  while true; do
    file="$prefix$number.$suffix"
    if [[ -e "$file" ]]; then
      ((number++))
    else
      touch "$file"
      break
    fi
  done
}

The input parameter to the function is split in prefix, number, suffix, then until the file exists, the number is incremented. Found a free slot, the file is created with touch.

The split mechanism should be adapted to your needs, and various error check should be added.

Answer 3

If sequence numbers are not required, then the easiest way is to use mktemp command.