[SOLVED] sed split substring that may or may not be present

Issue

Trying to split the column of text based on the text that may or may not be there.

sample file:

TEXT1D1NEWBWP210HTEXT2
TEXT1D1BWP210HTEXT2

expected output:

TEXT1D1NEWBWP210HTEXT2 NEWBWP 210H
TEXT1D1BWP210HTEXT2 BWP 210H

cmd used --- Expecting that "?" would check if the substring "NEW" is present and print in case it is there.

cat <text_file> | sed -e 's/.*\(\s*\)\(NEW\)\?\(BWP\)\([0-9]\+\)H.*/\0 \2\3 \4H/'

Output from the above cmd is

TEXT1D1NEWBWP210HTEXT2 BWP 210H
TEXT1D1BWP210HTEXT2 BWP 210H

Not sure what I am doing wrong here... :)

Solution

Using GNU or BSD sed for -E so you don't need all those backslashes (you were already using GNU sed for \s):

$ sed -Ee 's/((NEW)?BWP)([0-9]+)H.*/& \1 \3H/' file
TEXT1D1NEWBWP210HTEXT2 NEWBWP 210H
TEXT1D1BWP210HTEXT2 BWP 210H

The main problem with your regexp was the initial .* would consume the optional NEW if it was present.

Answered By - Ed Morton

Answer Checked By - Pedro (WPSolving Volunteer)