/ 
Nagios-Object-0.09
River stage one • 1 direct dependent • 1 total dependent
/ Nagios::Object::Config

NAME

Nagios::Object::Config

DESCRIPTION

This is a module for parsing and processing Nagios object configuration files into perl objects.

METHODS

new()

Create a new configuration object. If Version is not specified, the already weak validation will be weakened further to allow mixing of Nagios 1.0 and 2.0 configurations. For now, the minor numbers of Version are ignored. Do not specify any letters as in '2.0a1'.

To enable regular expression matching, use either the "regexp_matching" or "true_regexp_matching" arguments to new(). See enable_regexp_matching() and enable_true_regexp_matching() below.

my $objects = Nagios::Object::Config->new();
my $objects = Nagios::Object::Config->new( Version => 1.2 );

my $objects = Nagios::Object::Config->new(
                   Version => 2.0,
                   regexp_matching => 1,
                   true_regexp_matching => 2
);
parse()

Parse a nagios object configuration file into memory. Although Nagios::Objects will be created, they are not really usable until the register() method is called.

$parser->parse( "myfile.cfg" );
COMMENTED OUT NEW PARSER WORK IN PROGRESS

# this is the new one I'm working on to improve speed and flexibility sub parse2 { my( $self, $filename ) = @_; croak "cannot read file '$filename': $!" unless ( $filename && -r $filename );

    my $object_re = qr/
        # the type gets captured into $1
        define \s* (\w+) \s*
        # capture all of the text between the brackets into $2
        {( .*?
        # match the last bracket only if followed by another definition
        )} (?= \s* define \s+ \w+ \s* {)
        # capture extra text into $3 for further processing
        (.*)$
    /sx;

    my $attr_line_re = qr/
        # attribute name => $1
        ^\s* (\w+) \s+
        # non-capture grouping
        (?:
            # capture value into $2
            # |        make sure semicolon is not escaped
            # |        |            capture comment into $3
            (.+?) \s* (?<!\\)(; \s* .+?)
            |
            # capture value into $4 when there is no comment on the same line
            (.+)
        )$ # EOL
    /x;

    $Nagios::Object::pre_link = 1;

    # $file_idx is for tracking source files so modified
    # configurations can be written out to the same files
    my $file_idx = push( @{$self->{config_files}}, $filename ) - 1;

	my $fh = gensym();
	open( $fh, "<$filename" )
	    || croak "could not open $filename for reading: $!";

    my( $entry, $type, $data ) = ( '', '', '' );
    while ( my $line = <$fh> ) {
        $entry .= $line;
        if ( $entry =~ /$entry_re/ ) {
            ($type, $data, $entry) = ($1, $2, $3);

            my( %object, %comments, $key ) = ();
            foreach ( split /[\r\n]+/, $data ) {
                if ( $line =~ /$line_re/ ) {
                    $key = $1;
                    # see the definition of $line_re above to understand this
                    $object{$1} = defined($2) ? $2 : $4;
                    $comments{$1} = $3 if defined($3);
                }
                elsif ( $line =~ /^\s*;\s*(.*)$/ ) {
                    $comments{$key} = $1;
                }
            }
        }
    }
}
find_object()

Search through the list of objects' names and return the first match. The second argument is optional. Always using it can considerably reduce the size of the list to be searched, so it is recommended.

my $object = $parser->find_object( "localhost" );
my $object = $parser->find_object( "oracle", "Nagios::Service" );
find_objects_by_regex()

Search through the list of objects' names and return a list of matches. The first argument will be evaluated as a regular expression. The second argument is required and specifies what kind of object to search for.

The regular expressions are created by translating the "*" to ".*?" and "?" to ".". For now (v0.9), this code completely ignores Nagios's use_regexp_matching and use_true_regexp_matching and does full RE matching all the time.

my @objects = $parser->find_objects_by_regex( "switch_*", "Nagios::Host" );
my @objects = $parser->find_objects_by_regex( "server0?", "Nagios::Host" );
all_objects_for_type()

Obtain a reference to all objects of the specified Nagios object type.

Usage: $objects = all_objects_for_type($object_type)

Parameters: $object_type - A specific Nagios object type, i.e. "Nagios::Contact"..

Returns: A reference to an array of references to all objects of the specified type associated with this configuration. Objects of this type added to the configuration following the call to this method _will_ be accessible through this reference after the fact.

Note that the array reference by the return value may be empty.

Example:

my $contacts = $config->all_objects_for_type("Nagios::Contact");
if (scalar(@$contacts) == 0) {
    print "No contacts have yet been defined\n";
} else {
    foreach $contact (@$contacts) {
        ...
    }
}
find_attribute()

Search through the objects parsed thus far, looking for a particular textual name. When found, return that object. If called with two arguments, it will search through all objects currently loaded until a match is found. A third argument may specify the type of object to search for, which may speed up the search considerably.

my $object = $parser->find_attribute( "command_name", "check_host_alive" );
my $object = $parser->find_attribute( "command_name", "check_host_alive", 'Nagios::Host' );
resolve()

Resolve the template for the specified object. Templates will not work until this has been done.

$parser->resolve( $object );
register()

Examine all attributes of an object and link all of it's references to other Nagios objects to their respective perl objects. If this isn't called, some methods will return the textual name instead of a perl object.

$parser->register( $host_object );
my $timeperiod_object = $host_object->notification_period;
resolve_objects()

Resolve all objects currently loaded into memory. This can be called any number of times without corruption.

$parser->resolve_objects();
register_objects()

Same deal as resolve_objects(), but as you'd guess, it registers all objects currently loaded into memory.

$parser->register_objects();
enable_regexp_matching()/disable_regexp_matching()

This correlates to the "use_regexp_matching" option in nagios.cfg. When this option is enabled, Nagios::Object::Config will translate "*" to ".*?" and "?" to "." and evaluate the result as a perl RE, anchored at both ends for any value that can point to multiple other objects (^ and $ are added to either end).

$parser->enable_regexp_matching;
$parser->disable_regexp_matching;
enable_true_regexp_matching()/disable_true_regexp_matching()

This correlates to the "use_true_regexp_matching" option in nagios.cfg. This is very similar to the enable_regexp_matching() option, but matches more data and allows more powerful RE syntax. These modules will allow you the full power of perl RE's - this is probably more than is available in Nagios, so don't blame me if something works here but not in Nagios (it's usually the other way around anyways).

The generated RE's have the same translation as above, but do not have the anchors to ^ and $.

This option always supercedes enable_regexp_matching.

$parser->enable_true_regexp_matching;
$parser->disable_true_regexp_matching;
list_hosts(), list_hostgroups(), etc.

Returns an array/arrayref of objects of the given type.

->list_hosts;
->list_hostroups;
->list_services;
->list_timeperiods;
->list_commands;
->list_contacts;
->list_contactgroups;
->list_hostdependencies;
->list_servicedependencies;
->list_hostescalation;
->list_hostgroupescalation;
->list_serviceescalation;

AUTHOR

Al Tobey <tobeya@cpan.org> Contributions From: Lynne Lawrence (API & bugs)