Print only if line matches

With the following input (test.txt):

Sacramento   California   USA
Detroit      Michigan     USA
Vancouver    BC           Canada

Using awk:

By default, awk prints the current line on true conditions:

# Print only lines in USA
awk '$3 == "USA"' test.txt
Sacramento   California   USA
Detroit      Michigan     USA

Or perform a regex match:

awk '$3 ~ /[A-Z]{3}/' test.txt
Sacramento   California   USA
Detroit      Michigan     USA

To only print certain columns of matching lines:

# Print only city names in USA
awk '$3 == "USA" { print $1 }' test.txt
Sacramento
Detroit

Using sed:

• -n means do not print by default
• /p command prints the matching line

Print entire matching line:

# Print only lines containing "USA"
sed -n -e '/USA/p' test.txt
Sacramento   California   USA
Detroit      Michigan     USA

Replace some of the text before printing only matching lines:

# Replace 'i' with 'x' on lines which contain USA
sed -n -e '/USA/s/i/x/gp' test.txt
Sacramento   Calxfornxa   USA
Detroxt      Mxchxgan     USA

Print all lines, but only replace on matching lines

sed -e '/USA/s/i/x/g' test.txt
Sacramento   Calxfornxa   USA
Detroxt      Mxchxgan     USA
Vancouver    BC           Canada

Notice the regex (/USA/) just precedes the s/ portion of a normal sed replacement.

#bash #sed #awk #tips