Attempting to pull rows out of a file and then insert into a variable

[pmcmicha]

#!/bin/ksh -p

Example:

File: Test

Users Misc
----- ----

Data: Jack 784
Dirk 2345
Cindy 0987
Noone 3412

I use this to only pull what I want to see...

cat Test|awk '{print $1}' > /tmp/test.out

Now from here I want to be able to look at /tmp/test.out and pull each name from the file and put those into a case statement. Can this be done? I think it can be done using awks getline, but I am unable to figure this out from the man pages. Anyone have any ideas? Thanks in advance.

 

[tdatgod]

Hi,
In CSH this would be simply

set names = ( `awk '{print $1}' test` )

then
$names[1] would be "Jack"
$names[2] would be "Dirk"

and so on.

I don't know if this can be KSH.

 

[KenCunningham]

In ksh you could probably do it something like this:

for name in `cat test`
do
....case statements....
done

If you're only wanting the names from the test file, do a:

cat test | tr -s ' ' ' ' | cut -f1 -d' ' > newtest

first, to get just the names from the file in question into another file, which you can then plug into the script above. HTH.

 

[pmcmicha]

Here is the issue, this script is going to be put onto 5000 unix servers and I do not know what is contained within the test file, so I have to find a way to get those values into variables. I found something close to what I want:

sed 1q test | awk -F: '{print $1}' -- This would just pull Jack out which I could then assign a variable to, but this:

sed 2q test | awk -F: '{print $1}' -- would pull both Jack and Dirk out in this case. I only want to pull Dirk out to assign a variable to.

 

[toolkit]

Not exactly sure what you want. You could try using perl, and use a hash keyed by the user, for example if your data file looks like:

Jack:784
Dirk:2345
Cindy:0987

Then you would need to do something like:

open( DATA, "data" ) or die;
chomp( %id = map { split /:/ } <DATA> );
foreach $user ( sort keys %id ) {
    print &quot;$user has id $id{$user}\n&quot;;
}

Line 2 above simply makes a hash of 'user => id' pairs.

Hope this makes sense with what you are trying to achieve. Cheers, NEIL

 

[pmcmicha]

Maybe this will help:

#!/bin/ksh -p

# Here I am pulling the second column from test.doc

cat /all/test.doc | awk '{print $2}' > /tmp/test.out


# Now I want to look at /tmp/test.out and assign each row
# (value) a variable. I can do this for the first value
# using this... X=$(sed 1q test.out | awk -F: '{print $1}')
# The problem is I cannot do this for the second variable,
# third variable, forth variable, etc.

So my question is how can I do this? Is it possible to pull a row with a value in it, (that you do not know what it is), and assign a variable to it?

 

[MikeLacey]

The shell construct that Ken describes seems to do exactly what you want, each item (row) pulled out of the file will get assigned to a variable ($name in this case)
[tt]
#!/usr/bin/ksh

for name in $(awk '{print $1}' Test)
do
[tab]print $name
done
[/tt]
Is that what you mean? Mike
michael.j.lacey@ntlworld.com
Email welcome if you're in a hurry or something -- but post in tek-tips as well please, and I will post my reply here as well.