How to evaluate a stream line by line

Question

I am trying to avoid creating any new files to store output in order to minimize the risk of overwriting something in a directory with the same name. I am trying to just evaluate each line in a stream with a pipe instead of outputting to a file and then using a while read line do done < file loop. Something like:

echo -e "1\n2\n3\n4\n5" | #evaluate current line separately#

Could I somehow read each line into an array and then evaluate the elements in the array? or is there a better way to avoid accidentally overwriting files?

Answer 1

In bash, the common way is to use the Process Substitution:

while read line ; do
    ...
done < <( commands producing the input)

Answer 2

You were halfway there...

echo -e "1\n2\n3\n4\n5" | while read line; do
    ...
done

Note that bash runs each part of the pipeline in a separate process, and any variables defined there will not persist after that block. (ksh93 will preserve them, as the loop will run in the current shell process.)

Answer 3

You can avoid overwriting files by using mktemp or tempfile to create temporary files with unique names. However, I would use process substitution as in choroba's answer.