awk ignores the record separator I set for the first record

Question

I am trying to change the record separator but awk does not use the record separator I set until the second record and beyond.

I get output like this:

ARRAY /dev/md4 level=raid1 num-devices=2 metadata=0.90 UUID=e85beb4f:84b05929:786a31c9:93e269d4
ARRAY /dev/md6 level=raid1 num-devices=2 metadata=0.90 UUID=4dbc659a:4d11f1be:6cbe4d26:c1372c53
ARRAY /dev/md5 level=raid1 num-devices=2 metadata=0.90 UUID=7d34fc96:1fa6c3c5:b9932435:ad126815
ARRAY /dev/md3 level=raid1 num-devices=2 metadata=0.90 UUID=862ae8d0:bd00b2fd:9e171b97:2937fc49
ARRAY /dev/md2 level=raid1 num-devices=2 metadata=0.90 UUID=03bb1b77:42d474f4:608c5276:cc04d84c
ARRAY /dev/md1 level=raid1 num-devices=2 metadata=0.90 UUID=6c14cddc:3ef74d2d:2f26fa3f:32cc274b
ARRAY /dev/md0 level=raid1 num-devices=2 metadata=0.90 UUID=76e50525:9ae71e74:9697a0e6:6f77e132

I need output like this:

ARRAY=/dev/md4
level=raid1
num-devices=2
metadata=0.90
UUID=e85beb4f:84b05929:786a31c9:93e269d4
ARRAY=/dev/md6
level=raid1
num-devices=2
metadata=0.90
UUID=4dbc659a:4d11f1be:6cbe4d26:c1372c53
ARRAY=/dev/md5
level=raid1
num-devices=2
metadata=0.90
UUID=7d34fc96:1fa6c3c5:b9932435:ad126815
ARRAY=/dev/md3
level=raid1
num-devices=2
metadata=0.90
UUID=862ae8d0:bd00b2fd:9e171b97:2937fc49
ARRAY=/dev/md2
level=raid1
num-devices=2
metadata=0.90
UUID=03bb1b77:42d474f4:608c5276:cc04d84c
ARRAY=/dev/md1
level=raid1
num-devices=2
metadata=0.90
UUID=6c14cddc:3ef74d2d:2f26fa3f:32cc274b
ARRAY=/dev/md0
level=raid1
num-devices=2
metadata=0.90
UUID=76e50525:9ae71e74:9697a0e6:6f77e132

When I try to format it awk does nothing with the first record but then correctly formats the rest of the recors.

Here is the command I am using with the output:

$ mdadm --misc -sD |sed 's/ /=/' |awk '{RS=" "}{print $0}'
ARRAY=/dev/md4 level=raid1 num-devices=2 metadata=0.90 UUID=e85beb4f:84b05929:786a31c9:93e269d4
ARRAY=/dev/md6
level=raid1
num-devices=2
metadata=0.90
UUID=4dbc659a:4d11f1be:6cbe4d26:c1372c53
ARRAY=/dev/md5
level=raid1
num-devices=2
metadata=0.90
UUID=7d34fc96:1fa6c3c5:b9932435:ad126815
ARRAY=/dev/md3
level=raid1
num-devices=2
metadata=0.90
UUID=862ae8d0:bd00b2fd:9e171b97:2937fc49
ARRAY=/dev/md2
level=raid1
num-devices=2
metadata=0.90
UUID=03bb1b77:42d474f4:608c5276:cc04d84c
ARRAY=/dev/md1
level=raid1
num-devices=2
metadata=0.90
UUID=6c14cddc:3ef74d2d:2f26fa3f:32cc274b
ARRAY=/dev/md0
level=raid1
num-devices=2
metadata=0.90
UUID=76e50525:9ae71e74:9697a0e6:6f77e132

And I cannot use this command as it leaves off the array and I need that mdadm --misc -sDY

Answer 1

Use sed twice, the second time to replace spaces with newlines:

mdadm --misc -sD | sed 's/ /=/'| sed 's/ /\n/g'

or slightly optimized, only once but with two patterns:

mdadm --misc -sD | sed 's/ /=/;s/ /\n/g'

Answer 2

I would also prefer the sed-only solution given by @SvW, but to answer your question:

The RS=" " comes too late. It is set after the first line is read. You should call it in a BEGIN block, for example, to have an effect also for the first line:

mdadm --misc -sD | sed 's/ /=/' | awk 'BEGIN{RS=" "}{print $0}'

Edit: Okay, @Thor was faster. :)

Answer 3

Your solution does not work because RS is set too late in the process. Set it before awk receives any input and it works correctly:

mdadm --misc -sD | sed 's/ /=/' | awk '{print $0}' RS=' '

Perhaps the following awk-only answer is also useful:

mdadm --misc -sD | awk '{$2=$1"="$2; $1=""} 1' OFS='\n'