Compute Config file 3

[dietmarp]

Hi,

I have the need to compute a configuration file in order to perform certain tasks. The config file has different chapters.
This is - how i started - but I am kind of stock. I thought to read the config file and store all different chapters in separate array - so I can compute them after - as needed.
But how .. ? Any help appretiated.
Cheers - Dietmar

#!/usr/bin/perl
my $count="0";
$conffile="/tmp/config.cfg";
my (@chapter,@nodegroupdef,@customergroup,@linux,@windows);
open(CONF,"<$conffile") or die "Cannot open $conffile\n";
while (<CONF>) {
chomp; # no newline
s/#.*//; # no comments
s/^\s+//; # no leading white
s/\s+$//; # no trailing white
next unless length; # anything left?
if ( "$_" eq "^[NodegroupDefault" ) {
......
}



The look of the config file:

###Config File
[NodeGroupDefault]
1HOURHC
PRODUCTION
#
#
[CUSTOMER_GROUP]
customer_a=group_a
customer_b=group_b
customer_c=group_c
#
#
[Assignment_Linux]
Linux_a
Linux_b
#
[Assignment_Windows]
Windows_A
Windows_B

 

[travs69]

Why not create it in xml and read it in that way?

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Travis - Those who say it cannot be done are usually interrupted by someone else doing it; Give the wrong symptoms, get the wrong solutions;

 

[dietmarp]

Endusers shell be able to modify entries in the different chapters of the config file easy - thats why we thought to use this classic form of input config file

 

[rharsh]

This doesn't do any processing of the data in the config file, nor does it do any error checking, but it might give you somewhere to start.

my %config;
{
	my $chapter;
	while (<CONF>) {
		next if /^\s*$/;	# Blank line
		next if /^\s*#/;	# Comment
		s/^\s*//;			# Remove leading spaces
		s/\s*$//;			# Remove trailing spaces
		if (m/^\[(\w+)]/) {	# Find Chapter names
			$chapter = $1;
		} else {
			push @{$config{$chapter}}, $_;
		}
	}
}

There are probably a number of modules on CPAN that would be helpful. Take a look at Config::IniFiles or Config::Merge.

 

[dietmarp]

Thanks - that does the trick. Though I am not familiar with hashes and ask for some further help to understand it.

I know - that eg. the key Assignment_Linux contains Values Linux_a and Linux_b.

How can I assess those values for further computing?
Thanks - Dietmar
---------------------------------------------------
#!/usr/bin/perl
use strict;
use warnings;
my $conffile ="./config.cfg";
my %config;
my $config;
my $chapter;
sub get_config{
{
open(CONF,"<$conffile") or die "Cannot open $conffile\n";
while (<CONF>) {
next if /^\s*$/; # Blank line
next if /^\s*#/; # Comment
s/^\s*//; # Remove leading spaces
s/\s*$//; # Remove trailing spaces
if (m/^\[(\w+)]/) { # Find Chapter names
$chapter = $1;
} else {
push @{$config{$chapter}}, $_;
}
}
close (CONF);
}
}
&get_config();
my @KEY = keys %config;
print "@KEY\n";
-----------------------------------------------------
config.cfg

###Config File
[NodeGroupDefault]
1HOURHC
PRODUCTION
#
#
[CUSTOMER_GROUP]
customer_a=group_a
customer_b=group_b
customer_c=group_c
#
#
[Assignment_Linux]
Linux_a
Linux_b
#
[Assignment_Windows]
Windows_A
Windows_B

 

[dietmarp]

from example script and config aboth.
I added following to retrieve the keys of the hash

my @KEY = keys %config;
print "@KEY\n";

The return is:
#NodeGroupDefault CUSTOMER_GROUP Assignment_Windows Assignment_Linux

Then I try to read the value of eg "Assignment_Windows"

print "$config{'Assignment_Windows'}\n";

and the return is;
ARRAY(0x400977b0)

I'd expect this but instead
Windows_A Windows_B
to be printed.

Any idea.

 

[feherke]

Hi

print @{$config{'Assignment_Windows'}};

[gray]# or nicer[/gray]

print join ', ',@{$config{'Assignment_Windows'}};

Feherke.

http://rootshell.be/~feherke/

 

[dietmarp]

thanks

 

[dietmarp]

I run in a regex problem

the regex from example above
if (m/^\[(\w+)]/) { # Find Chapter names

does not map lines with the - character in the name.
e.g.
[Chapter1]
is mapped but
[Chapter-2]
or
[Chapter-New-3]
is not mapped.
How doese the regex to be altered to map all lines
starting with a [
and ending with a ]

Valid Characters are A-Z a-z 0-9 - _

 

[KevinADC]

if (m/^\[([\w-]+)]/) {

------------------------------------------
- Kevin, perl coder unexceptional!

 

[dietmarp]

thanks Kevin