=head1 NAME

Bytes::Random::Secure::Tiny - A tiny Perl extension to generate
cryptographically-secure random bytes.


    use Bytes::Random::Secure::Tiny;

    my $rng = Bytes::Random::Secure::Tiny->new; # Seed with 256 bits.

    my $bytes  = $rng->bytes(32);              # A string of 32 random bytes.
    my $long   = $rng->irand;                  # 32-bit random unsigned int.
    my $hex    = $rng->bytes_hex(10);          # 10 random bytes as hex digits.
    my $string = $rng->string_from('abc', 10); # Random string from a, b, & c.


L<Bytes::Random::Secure::Tiny> provides random bytes from a cryptographically
secure random number generator (ISAAC), seeded from strong entropy sources on
a wide variety of platforms. It does so without external dependencies (except
on Windows), and has a minimal but useful user interface patterned after the
module L<Bytes::Random::Secure>.

L<Bytes::Random::Secure> has a handful of dependencies. And its UI may be
bigger than a typical user needs. L<Bytes::Random::Secure::Tiny> is designed
to provide what 90% of Bytes::Random::Secure's users need, but with a simpler
user interface, and in a single module with no dependencies beyond core Perl.

In most cases this module may be used as a light-weight drop-in replacement
for L<Bytes::Random::Secure>.


This module aims to provide a generalized tool for generating cryptographically
secure randomness in a way that can fit into many applications while providing
a zero dependency toolchain, and a user interface that is both minimal and
simple. Common use-cases may include:

=over 4

=item * Creating temporary passphrases.

=item * Generating random salts.

=item * Generating a secret that can be hashed along with session cookies.

=item * Nonces. 

=item * Feeding secure key-gen utilities.


L<Bytes::Random::Secure::Tiny> employs several well-designed algorithms
adapted from established CPAN tools to generate a strong random seed, and then
to instantiate a high quality cryptographically secure pseudo-random number
generator based on the seed. It has taken significant research to come up with
a strong and sensible choice of established and published algorithms. The
interface is designed with minimalism and simplicity in mind.

In particular, the CSPRNG is based on the same algorithm used by 
L<Math::Random::ISAAC>, and the seeding is based on algorithms from 

Furthermore, this module runs its randomness through both statistical tests
and NIST L<FIPS-140|https://en.wikipedia.org/wiki/FIPS_140> tests to verify

As a C<::Tiny> module, the additional goals of low (or no) dependencies and a
light-weight code base make this an ideal choice for environments where heavier
dependency chains are problematic.

=head1 EXPORTS

Nothing is exported.

=head1 METHODS

=head2 new

    my $rng = Bytes::Random::Secure::Tiny->new;

Instantiate the pseudo-random number generator object. The seeding of the ISAAC
CSPRING defaults to 256 bits from a non-blocking entropy source. The CSPRNG
object should be instantiated as infrequently as practical; there is no benefit
to re-seeding, with the single cavaet that the CSPRNG object should not be
shared by threads or forked processes.

=head3 Constructor Parameters

Parameters described below are optional and case-insensitive.

=over 4

=item bits

    my $rng = Bytes::Random::Secure::Tiny->new(bits => 512);

Number of bits to use in seeding. Must be a value between 64 and 8192
inclusive, and must satisfy C<bits==2**n>. The default value is 256.

=item nonblocking

    my $nb_rng = Bytes::Random::Secure::Tiny->new(nonblocking=>1);
    my $bl_rng = Bytes::Random::Secure::Tiny->new(nonblocking=>0);

If set to a false value, a blocking entropy source may be used in seeding. This
is generally not necessary, as the non-blocking sources used are considered by 
most to be strong enough for cryptographic purposes.

Instantiating with a blocking source can exhaust system entropy (this has been
seen in testing), and in such cases C<new> will block until sufficient entropy
is generated.

The default is to use a non-blocking source, and you should probably accept
that default.


=head2 bytes

    my $random_bytes = $rng->bytes($n);

Returns a string of C<$n> random bytes. C<$n> must be a positive integer.

=head2 bytes_hex

    my $random_hex = $rng->bytes_hex(6); # E.g. f35dde7c02a4

Returns a string of hex digits. Each byte is represented by two lower-cased hex
digits. Therefore, C<< $rng->bytes_hex(1) >> will return a string of length 2,
such as C<7F>. There is no C<0x> prepended to the hex digits.

=head2 string_from

    my $random_string = $rng->string_from('abcdefg', 10);

Returns a string of random octets selected from the "Bag" string (in this case
ten octets from 'abcdefg'). Repeated bag characters are weighted according to
their frequency. For example, given the bag 'aabc', the character 'a' will be
selected approximately 50% of the time, though being I<random>, there are no
guarantees it will be selected at all.  For the bag 'abc', each character has
the same weight. The output may contain duplicate characters.  For example:

    say $rng->string_from('a', 10); # Must always be 'aaaaaaaaaa'

=head2 irand

    my $unsigned_long = $random->irand;

Returns a pseudo-random 32-bit unsigned integer. The value will satisfy
C<< 0 <= x <= 2**32-1 >>.

=head2 shuffle

    my $aref_shuffled = $random->shuffle($aref);

Shuffles the contents of a reference to an array in situ, and returns
the same reference.

L<List::Util>, which ships with Perl, includes C<shuffle> function. But that
function is flawed in two ways. First, from a cryptographic standpoint,
it uses Perl's C<rand>, which is not a CSPRNG, and therefore is inadequate.

Second, because Perl's rand has an internal state of just 32 bits, it cannot
possibly generate all permutations of arrays containing 13 or more elements.

This module's C<shuffle> uses a CSPRNG, and also benefits from large seeds
and a huge internal state. ISAAC can be seeded with up to 8192 bits, yielding
2^8192 possible initial states, and 2^8288 possible internal states. A seed of
8192 bits will assure that for arrays of up to 966 elements every permutation
is accessible.


Nothing to configure.


This module requires Perl 5.8 or newer. Unicode support in C<string_from> is
best with Perl 5.8.9 or newer. See the INSTALLATION section in this document
for details.


L<Bytes::Random::Secure::Tiny> uses an embedded version of the ISAAC
algorithm adapted from L<Math::Random::ISAAC> as its CSPRNG, but will
silently upgrade to using L<Math::Random::ISAAC> proper if it is available
on the target system.

C<Bytes::Random::Secure::Tiny> seeds using an embedded adaptation of
L<Crypt::Random::Seed>, but it will silently upgrade to using
L<Crypt::Random::Seed> proper if it is available on the target system.

If performance is a consideration and you are able to install
L<Math::Random::ISAAC::XS>, do so; L<Bytes::Random::Secure::Tiny> will
silently upgrade to using C<Math::Random::ISAAC::XS> instead of the embedded
ISAAC CSPRING. L<Math::Random::ISAAC::XS> implements the same ISAAC CSPRNG
algorithm in C and XS for speed.


When programming for parallel computation, create a unique
C<Bytes::Random::Secure::Tiny> object within each process or thread.
L<Bytes::Random::Secure::Tiny> uses a CSPRNG, and sharing the same RNG between
threads or processes will share the same seed and the same starting point. By
instantiating the B::R::S::T object after forking or creating threads, a unique
randomness stream will be created per thread or process.

Always share the same RNG object between all non-concurrent consumers within
a process, but never share the same RNG between threads or forked processes.



It's easy to generate weak pseudo-random bytes. It's also easy to think you're
generating strong pseudo-random bytes when really you're not. And it's hard to
test for pseudo-random cryptographic acceptable quality. There are many high
quality random number generators that are suitable for statistical purposes,
but not necessarily up to the rigors of cryptographic use.

Assuring strong (ie, secure) random bytes in a way that works across a wide
variety of platforms is also challenging. A primary goal for this module is to
provide cryptographically secure pseudo-random bytes while still meeting the
secondary goals of simplicity, minimalism, and no dependencies. If more
fine-grained control over seeding methods is needed, use
L<Bytes::Random::Secure> instead.

=head2 ISAAC

The L<ISAAC algorithm|https://en.wikipedia.org/wiki/ISAAC_(cipher)> is
considered a cryptographically strong pseudo-random number generator. It has
1.0e2466 possible initial states. The best known attack for discovering initial
state would theoretically take a complexity of approximately 4.67e1240, which
is of no practical consequence to ISAAC's security. Cycles are guaranteed to
have a minimum length of 2**40, with an average cycle of 2**8295. Because there
is no practical attack capable of discovering initial state, and because the
average cycle is so long, it's generally unnecessary to re-seed a running
application. The results are uniformly distributed, unbiased, and unpredictable
unless the seed is known.

To confirm the quality of the CSPRNG, this module's test suite implements the
L<FIPS-140-1|http://csrc.nist.gov/publications/fips/fips1401.htm> tests for
strong random number generators. See the comments in C<t/27-fips140-1.t> for


The C<string_from> method permits the user to pass a "bag" (or source) string
containing Unicode characters. For any modern Perl version, this will work
just as you would hope. But some versions of Perl older than 5.8.9 exhibited
varying degrees of bugginess in their handling of Unicode. If you're depending
on the Unicode features of this module while using Perl versions older than
5.8.9 be sure to test thoroughly, and don't be surprised when the outcome isn't
as expected. ...this is to be expected. Upgrade. This module works at the 
octet level, not grapheme cluster.


Care is taken so that there is no modulo bias in the randomness returned. This
is exactly I<why> the C<string_from> method is preferable to a home-grown
random string solution. However, the algorithm to eliminate modulo bias can
impact the performance of the C<string_from> method. Any time the length of the
bag string is significantly less than the nearest greater or equal factor of
2**32, performance will degrade. Unfortunately there is no known algorithm that
improves upon this situation. Fortunately, for sanely sized strings, it's a
minor issue. To put it in perspective, even in the case of passing a "bag"
string of length 2**31 (which is huge), the expected time to return random
bytes will only double.


No special requirements.

=head1 SEE ALSO

If support for hardware entropy generators is needed, use
L<Bytes::Random::Secure>. Other good CSPRNG's include L<Crypt::Random> and

=head1 AUTHOR

David Oswald I<< <davido@cpan.org> >>

=head1 BUGS

Please report any bugs or feature requests to
C<bug-bytes-random-secure at rt.cpan.org>, or through the web interface at
L<http://rt.cpan.org/NoAuth/ReportBug.html?Queue=Bytes-Random-Secure-Tiny>.  I
will be notified, and then you'll automatically be notified of progress on
your bug as I make changes.

=head1 SUPPORT

You can find documentation for this module with the perldoc command.

    perldoc Bytes::Random::Secure

You can also look for information at:

=over 4

=item * Github Repo: L<https://github.com/daoswald/Bytes-Random-Secure-Tiny>

=item * RT: CPAN's request tracker (report bugs here)


=item * AnnoCPAN: Annotated CPAN documentation


=item * CPAN Ratings


=item * Search CPAN




Dana Jacobsen ( I<< <dana@acm.org> >> ) for his work that led to
L<Crypt::Random::Seed>, and for ideas and code reviews.


Copyright 2015 David Oswald.

This program is free software; you can redistribute it and/or modify it
under the terms of either: the GNU General Public License as published
by the Free Software Foundation; or the Artistic License.

See http://dev.perl.org/licenses/ for more information.