CODEWITHSUNDEEP
awk
0stone0
0stone0

Reputation: 43999

Pad AWK columns

Trying to parse BSD top output to show only PID - COMMAND - MEM for a specific process;

$ top -l 1 | grep -E '%CPU\ |coreaudio'
PID    COMMAND          %CPU TIME     #TH    #WQ #PORTS MEM   PURG
354    com.apple.audio. 0.0  00:00.00 2      1   12     820K  0B
296    com.apple.audio. 0.0  00:00.03 2      1   38     2024K 0B
282    coreaudiod       0.0  03:25.05 94     2   736    21M   0B

Using to show only column $1 - $2

$ top -l 1 | grep -E '%CPU\ |coreaudio' | awk {'print $1" -- "$2'}
PID -- COMMAND
354 -- com.apple.audio.
296 -- com.apple.audio.
282 -- coreaudiod

Adding a 3th column messes with the 'columns' since the second columns is not being padded;

$ top -l 1 | grep -E '%CPU\ |coreaudio' | awk {'print $1" -- "$2" -- "$8'}
PID -- COMMAND -- MEM
354 -- com.apple.audio. -- 820K
296 -- com.apple.audio. -- 2024K
282 -- coreaudiod -- 21M

How would I 'pad' the 'column' to keep the 'layout' intact? Or should I use a different tool like ?

Note;

Using top -l 1 since I'm on a Mac

Upvotes: 1

Views: 176

Answers (1)

KamilCuk
KamilCuk

Reputation: 141135

You can pad the strings with some "constant" amount of spaces:

<<<$input awk '{printf "%-20s %-10s %-4s\n", $1, $2, $8}'
#                         ^^    ^^    ^ field width
#                        ^     ^     ^  left justify

You can use column:

<<<$input awk '{print $1, $2, $8}' | column -t

Upvotes: 2

Related Questions