parse output from the following

[kholikt]

I have the following command output and I need to find a way that I can extract the Pseudo name and Logical device ID.

Pseudo name=emcpower259a
Logical device ID=0113
state=alive; policy=SymmOpt; priority=0; queued-IOs=0
==============================================================================
---------------- Host --------------- - Stor - -- I/O Path - -- Stats ---
### HW Path I/O Paths Interf. Mode State Q-IOs Errors
==============================================================================
3075 pci@8,600000/SUNW,qlc@1/fp@0,0 c2t5006048C53688940d132s0 FA 1cA active alive 0 0
3072 pci@8,600000/SUNW,qlc@2/fp@0,0 c3t5006048C5368894Fd132s0 FA 16cA active alive 0 0

Pseudo name=emcpower255a
Logical device ID=0117
state=alive; policy=SymmOpt; priority=0; queued-IOs=0
==============================================================================
---------------- Host --------------- - Stor - -- I/O Path - -- Stats ---
### HW Path I/O Paths Interf. Mode State Q-IOs Errors
==============================================================================
3075 pci@8,600000/SUNW,qlc@1/fp@0,0 c2t5006048C53688940d133s0 FA 1cA active alive 0 0
3072 pci@8,600000/SUNW,qlc@2/fp@0,0 c3t5006048C5368894Fd133s0 FA 16cA active alive 0 0


The output from the above will give just two column, first is the pseudo name and second is a logical device id

for example, the output from the above will be

259 0113
255 0117

notice that it also strip off the emcpowerxxxa from it too. Look like there is no simple way to do this in a shell script

[Mel Burslan · Accepted solution]

cat output|grep "Pseudo name" | cut -d"=" -f2 | sed -e "/emcpower/s/emcpower//"

This should give you 259a and 255a from the output above

cat output | grep "Logical device" |cut -d"=" -f2

output of the above should give you 0113 and 0117

Am I missing something ?

[James R. Ferguson]

Hi:

One quick way:

# cat ./myfilter
#!/bin/sh
typeset FILE=$1
awk '
/Pseudo/ {split($0,a,"=");printf "%s", a[2]}
/Logica/ {split($0,a,"=");printf " %s\n", a[2]}
' ${FILE} | sed -e 's/[^ 0-9]*//g'

...using your input:

# ./myfilter myfile
259 0113
255 0117

...or:

# | ./myfilter

Regards!

...JRF...




./myfile

[James R. Ferguson]

Hi (again):

Actually, in the spirit of conservation of needless processes (even in pipes), you could do:

# ./myfilter
#!/bin/sh
typeset FILE=$1
awk '
/Pseudo/ {split($0,a,"=");gsub(/[^ 0-9]*/,"",a[2]);printf "%s",a[2]}
/Logica/ {split($0,a,"=");gsub(/[^ 0-9]*/,"",a[2]);printf " %s\n",a[2]}
' ${FILE}

...that is, let 'awk' due all the work it is designed to do :-)

Regards!

...JRF...

[James R. Ferguson]

Hi (again):

Are you satisfied with the help that you have received?

...JRF...