sed

Using sed to extract text from files

sed -n -r 's/.*?something (.*?) here.*/\1/p'

Notes:

• sed will take its input from stdin
• sed regex does not support \d , use [0-9] instead to match a digit

A breakdown of the sed "script" above:

• -n : suppress printing of pattern space
• -r : use regular expressions in the sed "script"
• s/ : search and substitute
• .*? : discard everything before the matching pattern
• (.*?) : capture group for the text we want to extract
• .* : discard everything after the matching regex pattern
• \1 : the value of the capture group
• /p : prints the matched capture group (represented by \1 )