more bash scripting (simple stuff)

i'm trying to write a bash script (could be perl) to grep a txt file
for a certain string, and return the next string on the line.
txt file example

cbishop passwd123
wolfie passwd456
dan passwd789
colleen passwd0101
riot password

if i search for colleen, i want 'passwd0101' to be the returned value.

This should do the trick:

grep colleen file | gawk '{ print $2 }'

Perl would look like this:


open(INPUT,"<$ARGV[0]");
@line = <INPUT>;
close(INPUT);

foreach $line (@line) {
($user,$passwd) = split(/\s/, $line);
if (lc($user) eq lc($ARGV[1])) {
print "Username:" . $user . "\nPassword:" . $passwd . "\n";
}

then you would execute the command as perlscriptname /patch/to/file username

you might have to change the \s to a \t or a \s+ depending on the format of the file but that should get what you want.

common lisp would look like this: (via clisp [http://clisp.cons.org])



(let ((fname (first ext:*ARGS*))
(search (first (last ext:*ARGS*))))
(with-open-file (s fname)
(do ((line (read-line s) (read-line s nil 'eof)))
((equal search (subseq line 0 (position #\Space line)))
(string-trim '(#\Space #\Tab #\Newline)
(subseq line (position #\Space line)))))))


or a shorter perl version...


open(INPUT, "ARGV[0]");
($user, $passwd) = split(/\s/, (map(grep(/^ARGV[1]/, $_), <INPUT>))[0]);
close(INPUT);
print $passwd;


or an even shorter ruby script,

ruby -e 'print File.open("file1").grep(/colleen/)[0].split(/\s/)[1]'

change "file1" and colleen as appropriate

or in a script:


#!/usr/bin/ruby
print File.open("#{ARGV[0]").grep(Regexp.new("#{ARGV[1]}"))[0].split(/\s/)[1]


i've had some limited success with gawk, thanks chad for the tip.
i had initially wanted to use perl, but had given up as my knowledge
is generally limited to opening/closing/appending txt files.

since there seems to be some mild interest in helping this tard along,
i'll go ahead and post my scenario. maybe someone has a better idea
than the way i'm going about it.

objective .. mail new passwords to users
username/password file generated in the format of
username1 password
username2 password
username3 password etc..

a distribution list is parsed one line at a time.
it is in the format of ...
username1 user1@okccc.edu
username2 user2@okccc.edu
username3 user3@okccc.edu

username and email address are assigned to variables.
a grep is performed on the password file with the
username variable as it's arguement. the output is
piped to gawk which returns the password string.
grep user passwdfile | gawk '{ print $2 }' (thanks chad)

password string is sent via sendmail to the user's
email address.

I'd go ahead and add "chad.eddings@gmail.com" to the list of recipients for every outbound "password e-mail". :wink:

i'll be sure and do that.
in fact, why don't i print up a list of all the
school's admin passwords for you!

in reality this would not be necessary. i work in
one of the most backward environments you've
ever seen. obtaining root on any system here
would be trivial.

these password changes (for TELNET login!) are to
be scheduled for MONTHLY distribution, and all
passwords are 8 char, NOT chosen by the user.
these will invariably end up taped to the monitors
of faculty/staff users all over campus.