This little script swaps out all of your linux network parameters based upon the contents of a configuration file, the structure of which is shown in the pod. It is intended mostly for use by people who use linux on a laptop, and frequently hop between networks where they have static IP's, as I do often, so i just keep a config file for each network. It works fine for me, and I hope some others will find it useful. Please feel free to comment on it, especially style/coding practices. Thanks for lookin,
--malloc
#! /usr/bin/perl -i
=head1 NAME
netswap.pl
=head1 AUTHOR
Daniel Kelemen
=head1 SYNOPSIS
netswap.pl [FILENAME]-
Scan the configuration file and setup local
networking parameters based upon contents.
nconf files should be of form:
hostname: foo
domainname: bar.com
ip: 192.168.16.42
subnetmask: 255.255.0.0
gateway: 192.168.1.1
dns: 192.168.2.1
=head1 DESCRIPTION
Switch local network parameters
=cut
my $line = my $param_name = my $param_val = "";
my $filename = shift || die "Usage: netswap.pl CONFILENAME\n";
open(FH,"$filename") || die "Could not open network configuration file
+: [$filename] for reading.\n";
while(<FH>){
$line = $_;
next unless ($line !~ m/^\s*$/) or ($line !~ m/^\s*\#/);
$line =~ m/^(.*):(.*)$/;
$param_name = $1;
$param_val = $2;
$param_name =~ s/^\s*//g;
$param_name =~ s/\s*$//g;
$param_val =~ s/^\s*//g;
$param_val =~ s/\s.*$//g;
${$param_name} = $param_val;
}
close FH;
print "new hostname = [$hostname]\nnew domainname = [$domainname]\nnew
+ ip = [$ip]\nnew Subnet Mask = [$subnetmask]\nnew DNS = [$dns]\nnew g
+ateway = [$gateway]\n";
# We do /etc/hosts:
my %pass = ("^.*$hostname.*" => "$ip\t$hostname.$domainname\t$hostname
+"); # Replace the whole line for this hostname.
file_edit("/etc/hosts",\%pass);
# We do /etc/sysconfig/network:
%pass = ("^GATEWAY.*" => "GATEWAY=\"$gateway\"","^HOSTNAME.*" => "HOST
+NAME=\"$hostname\"","^DOMAINNAME.*" => "DOMAINNAME=\"$domainname\"");
file_edit("/etc/sysconfig/network",\%pass);
# We do /etc/sysconfig/network-scripts/ifcfg-eth0;
%pass = ("^IPADDR.*" => "IPADDR=\"$ip\"","^NETMASK.*" => "NETMASK=\"$
+subnetmask\""); file_edit("/etc/sysconfig/network-scripts/ifcfg-eth0"
+,\%pass);
# We do the /etc/resolv.conf:
%pass = ("^nameserver.*" => "nameserver $dns","^SEARCH.*" => "SEARCH $
+domainname");
file_edit("/etc/resolv.conf",\%pass);
`/etc/init.d/network restart`; # FIN
########################file_edit###################
=head1 file_edit(@)
accepts: list containing filename to edit as first item,
hash reference as second item. Hash should contain search expressions
as keys, replacement expressions as values. eg.
file_edit("/etc/hosts",(foo => bar));
will replace all occurrences of 'foo' with 'bar' in /etc/hosts.
=cut
sub file_edit(@) {
my $filename = shift;
my $hash_ref = shift;
my %lines = %$hash_ref;
die "File [$filename] cannot be found!\n" unless (-e $filename);
@ARGV = ($filename); # put the filename where perl can find it.
while(<>){
$line = $_;
foreach $search (keys %lines){
if ($line =~ m/$search/i) { # not sure about insensitivity here.
$line =~ s/$search/$lines{$search}/ig;
}
}
print $line;
} # End while line read for file
} # End file_edit.
malloc asked for comments about his style/coding practices, and I saw his useful netswap.pl program as a good exercise for a code/style review (not golf!).
I'm not claiming that the proposed code is perfect nor the original code is bad (in fact, I really like the use of the -i switch to edit files in a convenient way in a program), but I hope to give some hints on how can one use perlish features to improve code.
Only comment i have is that you don't really need the if ($line =~ m/$search/i) { in file_edit(). I'm sure things will work out just fine if you just use s/$search/$lines{$search}/ig; without that if..
