"sed" command examples

Remove every 2nd line:

sed 'n;d' input.txt

Remove 8th line:

sed 8d input.txt

Remove all empty lines:

sed /^$/d input.txt

Print each 10th line:

sed -n '0~10p' input.txt

Remove all lines from the beginning until the first empty line, inclusively:

sed 1,/^$/d input.txt

Print all lines containing "John" pattern:

sed -n /John/p input.txt

Replace first occurrence of "Windows" word with "Linux" word in each line:

sed s/Windows/Linux/ input.txt

Replace all occurrences of "Old" word with "New" word in each line:

sed s/Old/New/g input.txt

Remove all spaces in the end of each line:

sed s/ *$// input.txt

Replace all occurrences of leading zeros with a single zero symbol:

sed s/00*/0/g input.txt

Remove all lines containing "TRACE" word:

sed /TRACE/d input.txt

Remove all "ABC" words from each line:

sed s/GUI//g input.txt

Remove all lines before the first occurrence of "2014-05-09" and save results into the same file:

sed -i '0,/2014-05-09/d' gc-details.log

Replace multiple patterns using regular expressions and groups:

echo "I'm 20 years old and I live on Earth" | sed -r "s/(.*)20(.*)Earth/\1500\2Mars/"
I'm 500 years old and I live on Mars

Number each line with tab separator:

sed = input.txt | sed 'N;s/\n/\t/'

Number each non-empty line with space separator:

sed '/./=' input.txt | sed '/./N; s/\n/ /'

If a line end with "\" symbol then remove "\" and append next line:

sed -e :a -e '/\\$/N; s/\\\n//; ta' input.txt

If a line starts with "=" symbol then append the line to a previous one and replace "=" with space:

sed -e :a -e '$!N;s/\n=/ /;ta' -e 'P;D' input.txt

Add an empty line after each line:

sed G input.txt

Add an empty line after each line, avoid multiple empty line:

sed '/^$/d;G' input.txt

Add two empty lines after each line:

sed 'G;G' input.txt

Add an empty line before each line containing "RED" word:

sed '/RED/{x;p;x;}' input.txt

Add an empty line after each line containing "GREEN" word:

sed '/GREEN/G' input.txt

Add empty lines before and after each line containing "GO" word:

sed '/GO/{x;p;x;G;}' input.txt