make it one liner pls

[hok1man]

Hi Guys,

as title aboved, the original command is :
ps -ef | nawk '$1=="napln"{print $0}' | grep -i /data | grep -i $JOB

I'm using Solaris 8

Thanks guys.

 

[Annihilannic]

For starters, {print $0} is unnecessary, as that is the default action, so you could write:

[tt]ps -ef | nawk '$1=="napln"' | grep -i /data | grep -i $JOB[/tt]

But you don't need awk to do that part anyway, since it's built in to ps:

[tt]ps -fu napln | grep -i /data | grep -i $JOB[/tt]

Also you can combine the greps if you know the order the strings are likely to be in, e.g.

[tt]ps -fu napln | grep -i /data.*$JOB[/tt]

...where .* matches any characters between the two search strings.

If you wish to use awk, unfortunately it doesn't have a very convenient way to do case-insensitive searches. But you could do something like this:

[tt]ps -ef | awk -v job=$JOB '$1=="napln" && tolower($0) ~ /\/data/ && toupper($0) ~ tolower(job)'[/tt]

Annihilannic.

 

[Annihilannic]

Err, toupper($0) should read tolower($0) as well. :-$

Annihilannic.

 

[hok1man]

Hi Anni,

thanks for explanation..
is there any simplified command for this :
ps -aux

http://www |

nawk '$1=="napln"'

Because I need to see the full command process.
thanks,

 

[Annihilannic]

auxwwww are BSD-style ps options, you shouldn't really precede them with a -. Unfortunately with BSD options there is no built-in filter by user name, so that is probably the simplest syntax you could use.

On Linux you can actually mix BSD and UNIX style options, although it's probably not very good practice. I often use ps www -ef for example to get wide output like you describe, so you could use ps u

http://www -u

napln. Just don't tell anyone I said you could.

Annihilannic.