Shell Programming and Scripting

I am not sure if this can be done easily. I am trying to clean up our /etc/shadow files to be standard for password expiration, warn dates, etc. I have a script called aging_set.bash that takes two arguments. First argument is the user's name the second argument is how many days til their password expires (30 days, 60 days, 90 days etc). The script uses usermod to make changes to the /etc/shadow file for expiration and works fine if I do each name manually. I have 300 servers to do this on and I really dont want to do it manually so I have generated a text file called users.txt that has the user name and max password days.

user1 30
user2 30
user3 60 etc....

I am trying to do a for loop by doing:

for x in `cat users.txt`;do aging_set.bash $x;done

I thought it would read the whole line and the aging_set.bash script would recognize this as two seperate arguements. However it is reading each column as a seperate entry. I did the following:

for x in `cat users.txt`;do echo $x;done
and I got

user1
30
user2
30
user3
60
...


Any help is appreciated...

Thanks
Robert

There are probably a gazillion ways of doing this. I'd personally use arrays, but a simple solution could be substituting the space between user and number with a character in users.txt, then pass $x while sed'ing the character into a space, say...

Code:

# cat users.txt
user1@60
user2@60
user3@30
user4@30

#cat script
USER=$1
DAYS=$2
echo $USER $DAYS

# for x in $(cat users.txt); do
> script $(echo $x | sed 's/@/ /')
>done

user1 60
user2 60
user3 30
user4 30

Code:

#!/bin/ksh

while read user age junk
do
   echo "user->[${user}] age->[${age}]"
#   aging_set.bash "${user}" "${age}"
done <users.txt

When satisfied, uncomment the 'aging_set.bash' line.