NAME

       Data::UUID::LibUUID - uuid.h based UUID generation (versions 2 and 4 depending on
       platform)

SYNOPSIS

           use Data::UUID::LibUUID;

           my $uuid = new_uuid_string();

DESCRIPTION

       This module provides bindings for libuuid shipped with e2fsprogs or uuid-dev on debian,
       and also works with the system uuid.h on darwin.

EXPORTS

       new_uuid_string $version
       new_uuid_binary $version
           Returns a new UUID in string (dash separated hex) or binary (16 octets) format.

           $version can be either 2, or 4 and defaults to whatever the underlying implementation
           prefers.

           Version 1 is timestamp/MAC based UUIDs, like Data::UUID provides. They reveal time and
           host information, so they may be considered a security risk.

           Version 2 is described here
           <http://www.opengroup.org/onlinepubs/9696989899/chap5.htm#tagcjh_08_02_01_01>.  It is
           similar to version 1 but considered more secure.

           Version 4 is based just on random data. This is not guaranteed to be high quality
           random data, but usually is supposed to be.

           On MacOS X "getpid" is called before UUID generation, to ensure UUIDs are unique
           accross forks. Behavior on other platforms may vary.

       uuid_to_binary $str_or_bin
           Converts a UUID from string or binary format to binary format.

           Returns undef on a non UUID argument.

       uuid_to_string $str_or_bin
           Converts a UUID from string or binary format to string format.

           Returns undef on a non UUID argument.

       uuid_eq $str_or_bin, $str_or_bin
           Checks if two UUIDs are equivalent. Returns true if they are, or false if they aren't.

           Returns undef on non UUID arguments.

       uuid_compare $str_or_bin, $str_or_bin
           Returns -1, 0 or 1 depending on the lexicographical order of the UUID. This works like
           the "cmp" builtin.

           Returns undef on non UUID arguments.

       new_dce_uuid_string
       new_dce_uuid_binary
           These two subroutines are a little hackish in that they take no arguments but also do
           not validate the arguments, so they can be abused as methods:

               package MyFoo;

               use Data::UUID::LibUUID (
                   new_dce_uuid_string => { -as "generate_uuid" },
               );

               sub yadda {
                   my $self = shift;
                   my $id = $self->generate_uuid;
               }

           This allows the ID generation code to be subclassed, but still keeps the hassle down
           to a minimum. DCE is UUID version two specification.

       ascending_ident
           Creates a lexically ascending identifier containing a UUID, high resolution timestamp,
           and a counter.

           This is not a UUID (it's longer), but if you can store variable length identifier (and
           exposing the system clock is not an issue) they can be used to create an identifier
           that is both universally unique, and lexically increasing.

           Note that while the identifiers are universally unique, there is no universal ordering
           (that would require synchronization), so identifiers generated on different machines
           or even different process/thread could have IDs which interleave.

TODO

       •   Consider bundling libuuid for when no system "uuid.h" exists.

SEE ALSO

       Data::GUID, Data::UUID, UUID, <http://e2fsprogs.sourceforge.net/>

VERSION CONTROL

       This module is maintained using Darcs. You can get the latest version from
       <http://nothingmuch.woobling.org/code>, and use "darcs send" to commit changes.

AUTHOR

       Yuval Kogman <nothingmuch@woobling.org>

COPYRIGHT

           Copyright (c) 2008 Yuval Kogman. All rights reserved
           This program is free software; you can redistribute
           it and/or modify it under the same terms as Perl itself.