version 2.103622


  use Net::MAC;
  my $mac = Net::MAC->new('mac' => '08:20:00:AB:CD:EF');

  # Example: convert to a different MAC address format (dotted-decimal)
  my $dec_mac = $mac->convert(
      'base' => 10,         # convert from base 16 to base 10
      'bit_group' => 8,     # octet grouping
      'delimiter' => '.'    # dot-delimited

  print "$dec_mac\n"; # Should print

  # Example: find out whether a MAC is base 16 or base 10
  my $base = $mac->get_base();
  if ($base == 16) {
      print "$mac is in hexadecimal format\n";
  elsif ($base == 10) {
      print "$mac is in decimal format\n";
  else { die "This MAC is neither base 10 nor base 16"; }


This is a module that allows you to

- store a MAC address in a Perl object - find out information about a stored MAC address - convert a MAC address into a specified format - easily compare two MAC addresses for string or numeric equality

There are quite a few different ways that \s-1MAC\s0 addresses may be represented in textual form. The most common is arguably colon-delimited octets in hexadecimal form. When working with Cisco devices, however, you are more likely to encounter addresses that are dot-delimited 16-bit groups in hexadecimal form. In the Windows world, addresses are usually dash-delimited octets in hexadecimal form. \s-1MAC\s0 addresses in a Sun ethers file are usually non-zero-padded, colon-delimited hexadecimal octets. And sometimes, you come across dot-delimited octets in decimal form (certain Cisco \s-1SNMP\s0 \s-1MIBS\s0 actually use this). Hence the need for a common way to represent and manipulate \s-1MAC\s0 addresses in Perl.

There is a surprising amount of complexity involved in converting \s-1MAC\s0 addresses between types. This module does not attempt to understand all possible ways of representing a \s-1MAC\s0 address in a string, though most of the common ways of representing \s-1MAC\s0 addresses are supported.


\fInew()\fP method (constructor)

The new() method creates a new Net::MAC object. Possible arguments are

mac a string representing a MAC address base a number corresponding to the numeric base of the MAC possible values: 10 16 delimiter the delimiter in the MAC address string from above possible values: : - . space bit_group the number of bits between each delimiter possible values: 8 16 48 zero_padded whether bit groups have leading zero characters (Net::MAC only allows zero-padding for bit groups of 8 bits) possible values: 0 1 format the name of a MAC address format specification which takes the place of the base,delimiter,bit_group and zero_padded options above verbose write informational messages (useful for debugging) possible values: 0 1 die die() on invalid MAC address (default is to die on invalid MAC) possible values: 0 1 (default is 1)

When the new() method is called with a 'mac' argument and nothing else, the object will attempt to auto-discover metadata like bit grouping, number base, delimiter, etc. If the \s-1MAC\s0 is in an invalid or unknown format, the object will call the croak() function. If you don't want the object to croak(), you can give the new() method a die argument, such as:

my $m_obj = Net::MAC->new('mac' => '000adf012345', 'die' => 0);

There are cases where the auto-discovery will not be able to guess the numeric base of a \s-1MAC\s0. If this happens, try giving the new() method a hint, like so:

# Example: this MAC is actually in decimal-dotted notation, not hex my $mac = Net::MAC->new('mac' => '', 'base' => 10);

This is necessary for cases like the one above, where the class has no way of knowing that an address is decimal instead of hexadecimal.

If you have installed a custom \s-1MAC\s0 address format into the class (see below) then you can also pass the \*(C`format\*(C' option as a hint:

my $mac = Net::MAC->new('mac' => 'ab01~ab01~ab01', 'format' => 'My_Format');

class methods


When discovering \s-1MAC\s0 address formats, and converting between different formats (using \*(C`convert\*(C' or \*(C`as_*\*(C') the module can use predefined common formats or you can install your own for local circumstances.

For example consider a fictional device which uses \s-1MAC\s0 addresses formatted like \*(C`ab01~ab01~ab01\*(C', which would otherwise not be understood. You can install a new Format for this address style:

Net::MAC->set_format_for( 'My_Format_Name' => { base => 16, bit_group => 16, delimiter => '~', });

Now when using either the \*(C`format\*(C' option to \*(C`new()\*(C', or the \*(C`convert()\*(C' or \*(C`as_*\*(C' methods, the module will recognise this new format \*(C`My_Format_Name\*(C'. The Hashref supplied can include any of the standard options for formats as listed elsewhere in this documentation.

my $mac = Net::MAC->new('mac' => 'ab01~ab01~ab01', 'format' => 'My_Format_Name');

Custom formats sharing the same name as one shipping with the module (such as \*(C`Cisco\*(C') will override that built-in format.

accessor methods

get_mac() method

Returns the \s-1MAC\s0 address stored in the object.

get_base() method

Returns the numeric base of the \s-1MAC\s0 address. There are two possible return values:

16 hexadecimal (common) 10 decimal (uncommon)

get_delimiter() method

Returns the delimiter, if any, in the specified \s-1MAC\s0 address. A valid delimiter matches the following regular expression:


In other words, either a colon, a dash, a dot, or a space. If there is no delimiter, this method will return the undefined value (undef). If an invalid delimiter is found (like an asterisk or something), the object will call the croak() function.

get_bit_group() method

Returns the number of bits between the delimiters. A \s-1MAC\s0 address is a 48 bit address, usually delimited into 8 bit groupings (called octets), i.e.


Sometimes, \s-1MAC\s0 addresses are specified with fewer than 5 delimiters, or even no delimiters at all:

0820.00ab.cdef # get_bit_group() returns 16 082000abcdef # get_bit_group() returns 48, no delimiters at all

get_zero_padded() method

Returns a boolean value indicating whether or not the bit groups are zero-padded. A return value of 0 (false) means that the bit groups are not zero-padded, and a return value of 1 (true) means that they are zero-padded: # get_zero_padded() returns 1 0:80:2:ac:4f:ff # get zero_padded() returns 0 # get_zero_padded() returns 0

Net::MAC only allows bit groups of 8 bits to be zero-padded.

\fIconvert()\fP method

Convert an already-defined Net::MAC object into a different \s-1MAC\s0 address format. With this function you can change the delimiter, the bit grouping, or the numeric base.

# Example: convert to a different MAC address format (dotted-decimal) my $new_mac_obj = $existing_mac_obj->convert( 'base' => 16, # convert to base 16, if necessary 'bit_group' => 16, # 16 bit grouping 'delimiter' => '.' # dot-delimited );

Note that if any of the above arguments are not provided, they will be set to the following default values:

base 16 bit_group 8 (i.e. a delimiter will be used) delimiter :

Conversion to common formats

The most common formats have shortcut conversion methods that can be used instead of the convert() method with its many options.

as_Cisco() method

Cisco routers seem to usually represent \s-1MAC\s0 addresses in hexadecimal, dot-delimited, 16 bit groups.

my $mac = Net::MAC->new(mac => '00-02-03-AA-AB-FF'); my $cisco_mac = $mac->as_Cisco(); print "$cisco_mac"; # should print 0002.03aa.abff

as_IEEE() method

The \s-1IEEE\s0 802 2001 specification represents \s-1MAC\s0 addresses in hexadecimal, colon-delimited, upper case, 8 bit groups.

my $mac = Net::MAC->new(mac => '00-02-03-AA-AB-FF'); my $IEEE_mac = Net::MAC->as_IEEE(); print "$IEEE_mac"; # should print 00:02:03:AA:AB:FF

as_Microsoft() method

Microsoft usually represents \s-1MAC\s0 addresses in hexadecimal, dash delimited, upper case, 8 bit groups.

my $mac = Net::MAC->new(mac => '00:02:03:AA:AB:FF'); my $microsoft_mac = $mac->as_Microsoft(); print "$microsoft_mac"; # should print 00-02-03-AA-AB-FF

as_Sun() method

Sun represents \s-1MAC\s0 addresses in hexadecimal, colon-delimited, non-zero-padded, lower case, 8 bit groups.

my $mac = Net::MAC->new(mac => '00-02-03-AA-AB-FF'); my $sun_mac = $mac->as_Sun(); print "$sun_mac"; # should print 0:2:3:aa:ab:ff


The stringification operator "" has been overloaded to allow for the meaningful use of the instance variable in a string.

my $mac = Net::MAC->new(mac => '00:0a:23:4f:ff:ef'); print "object created for MAC address $mac"; # Should print: # object created for MAC address 00:0a:23:4f:ff:ef

\s-1MAC\s0 address comparison

The Perl operators 'eq' and 'ne' (string comparison) and '==' '!=' (numeric comparison) have been overloaded to allow simple, meaningful comparisons of two \s-1MAC\s0 addresses.

Example (two \s-1MAC\s0 addresses numerically identical but in different formats):

my $d = Net::MAC->new(mac => '', base => 10); my $h = Net::MAC->new(mac => '00:08:01:0A:10:10', base => 16); if ($d == $h) { print "$d and $h are numerically equal"; } if ($d ne $h) { print " but $d and $h are not the same string"; }


Malformed \s-1MAC\s0 addresses

Net::MAC can't handle \s-1MAC\s0 addresses where whole leading zero octets are omitted. Example: (should be

Arguably, that's their problem and not mine, but maybe someday I'll get around to supporting that case as well.

Case is not preserved

Net::MAC doesn't reliably preserve case in a \s-1MAC\s0 address. I might add a flag to the new() and convert() methods to do this. I might not.

Case is however altered when using the as_foo() formatted output methods.


Net::MacMap Net::MAC::Vendor


Oliver Gorwits <[email protected]>


Oliver Gorwits, Robin Crook, Kevin Brintnall


Karl Ward <[email protected]>


This software is Copyright (c) 2010 by Karl Ward <[email protected]>.

This is free software, licensed under:

The GNU General Public License, Version 2, June 1991