Need help with a shell script

[w1z4rd]

So I need to extract all the emails from one of the domains I have. There are roughly 2000 email accounts on this one domain. The server is a CentOS server with Cpanel loaded on it. I have two scripts to work with:

One using awk:

#!/bin/awk -f
#
# File: getAllEmail.awk
#
# This awk script takes the input of ls -d /home/*/mail/*/*
# checks to see if the directory exists, then outputs the email address. My example
# outputs them to a text file called emailaddress.txt
#
# I wrote it to get all the email addresses out of my cpanel box.
# You are free to use this program for your own purposes. If you
# improve it, please let me know. If it is horribly flawed, please let me know.
#
# Author: Hans C. Kaspersetz
# Date: 10/27/2007
# Email: hans [at] cyberxdesigns [dot] com
#
# Usage on cpaneli cli> ls -d /home/*/mail/*/*/ | getAllEmail.awk > emailaddress.txt

BEGIN {
}
{
tcmd = "test -d " $1
if(!system(tcmd)){
split($1,MyArray,"/")
print MyArray[6] "@" MyArray[5]
}
}

and one using PHP

#!/usr/local/bin/php
<?php
$folderlist = `ls -d /home/*/mail/*/*`;
$list_array = explode("\n",$folderlist);
foreach($list_array as $folder){
if(is_dir($folder)){
$folder_array = explode("/",$folder);
print $folder_array[5] . "@" . $folder_array[4] ."\n";
}
}
?>

Now these scripts work great on all domains, except my largest one. Where I get the following error:

-bash: /bin/ls: Argument list too long

Anyone know how to break any of the above scripts up so I can safely extract the emails from the domain?

 

[icyrus]

Try using find instead of ls.

 

[ozoned]

Change your command line usage, and run the awk script more than once. (how many depends on your usernames, and number of users):

usage:

ls -d /home/*/mail/[a-l]*/*/ | getAllEmail.awk >> emailaddress.txt

ls -d /home/*/mail/[m-z]*/*/ | getAllEmail.awk >> emailaddress.txt

 

[GreGorGy]

Change your command line usage, and run the awk script more than once. (how many depends on your usernames, and number of users):

usage:

ls -d /home/*/mail/[a-l]*/*/ | getAllEmail.awk >> emailaddress.txt

ls -d /home/*/mail/[m-z]*/*/ | getAllEmail.awk >> emailaddress.txt

That is sweet! ANd I know I will need to do the same sometime very soon so thanks.

 

[w1z4rd]

I used the find command in the following format... but it still have errors but it got me closer:

for i in `find /home/ -maxdepth 2 -type d -name mail`; do ls -d $i/*/* | getAllEmail.awk >> emailaddress.txt; done

However, with all the commands I will still having issue with the biggest domain, so using advice here and in other players, I was able to extract the big domains email accounts with the following commands:

 ls /home/domainname/mail/domainname.co.za/* -d | getAllEmail.awk >> email.txt

Thanks for the help

 

[sin_x]

icyrus is correct - find does not have the limitations that ls does.

simply:
find /home/*/mail/*/*/ -type d | getAllEmail.awk > emailaddress.txt

Will find all directories in the path, and pass to your awk script.

 

[icyrus]

I used the find command in the following format... but it still have errors but it got me closer:

Could you not have just used:

find /home -type d -wholename '*/mail/*'