Search Results

How about
sed -n '/ERROR/p; //,/^[0-9]\{4\}-/ {/^[0-9]\{4\}-/!p}' /tmp/testlog2.txtTested on Ubuntu linux 17.10

how about awk:

awk '/ERROR/' RS= sample.log

Glad you got what you need. You can see also how important a spot-on, precise, detailed specification is... once the problem is clear, help is near.

What's your bash version?

EDIT: Drop the spaces when assigning the variable:

outline="$outline"$'\n'"$line"

I don't think it is "a line", but several lines; at least it follows from his comment. So in any case, he must have some program which extracts a set of lines from the input file and writes these to...

It becomes a bit clearer now. You seem to read lines from some text file, which by default won't contain any <new line> chars unless certain merasures taken. You don't use e.g. "command substitution"...