=head1 NAME

Convert::ASN1 - ASN.1 Encode/Decode library

=head1 VERSION

version 0.27

=head1 SYNOPSYS

  use Convert::ASN1;

  $asn = Convert::ASN1->new;

  $asn->prepare(q<

    [APPLICATION 7] SEQUENCE {

      int INTEGER,

      str OCTET STRING

    }

  >);

  $pdu = $asn->encode( int => 7, str => "string");

  $out = $asn->decode($pdu);

  print $out->{int}," ",$out->{str},"\n";

  use Convert::ASN1 qw(:io);

  $peer   = asn_recv($sock,$buffer,0);

  $nbytes = asn_read($fh, $buffer);

  $nbytes = asn_send($sock, $buffer, $peer);

  $nbytes = asn_send($sock, $buffer);

  $nbytes = asn_write($fh, $buffer);

  $buffer = asn_get($fh);

  $yes    = asn_ready($fh)

=head1 DESCRIPTION

Convert::ASN1 encodes and decodes ASN.1 data structures using BER/DER

rules.

=head1 METHODS

=head2 new ( [OPTIONS] )

Contructor, creates a new object.

If given, B<OPTIONS> are the same ones as for L</"configure ( OPTIONS )"> below.

=head2 error ()

Returns the last error.

=head2 configure ( OPTIONS )

Configure options to control how Convert::ASN1 will perform various tasks.

Options are passed as name-value pairs.

=over 4

=item encode

Reference to a hash which contains various encode options.

=item decode

Reference to a hash which contains various decode options.

=item encoding

One of 'BER' or 'DER'. The default is 'BER'

=item tagdefault

One of 'EXPLICIT' or 'IMPLICIT'.

Default tagging conventions are normally given in the ASN.1 module definition (not supported by the parser). The ASN.1 spec states EXPLICIT tagging is the default, but this option has IMPLICIT tagging default for backward compatibility reasons.

=back

Encode options

=over 4

=item real

Which encoding to use for real's. One of 'binary', 'nr1', 'nr2', 'nr3'

=item time

This controls how UTCTime and GeneralizedTime elements are encoded. The default

is C<withzone>.

=over 4

=item utctime

The value passed will be encoded without a zone, ie a UTC value.

=item withzone

The value will be encoded with a zone. By default it will be encoded

using the local time offset. The offset may be set using the C<timezone>

configure option.

=item raw

The value passed should already be in the correct format and will be copied

into the PDU as-is.

=back

=item timezone

By default UTCTime and GeneralizedTime will be encoded using the local

time offset from UTC. This will over-ride that. It is an offset from UTC

in seconds.  This option can be overridden by passing a reference to a

list of two values as the time value. The list should contain the time

value and the offset from UTC in seconds.

=item bigint

If during encoding an value greater than 32 bits is discovered and

is not already a big integer object, then the value will first be

converted into a big integer object. This option controls the big

integer class into which the objects will be blessed. The default

is to use Math::BigInt

=back

Decode options

=over 4

=item time

This controls how a UTCTime or a GeneralizedTime element will be decoded. The default

is C<utctime>.

=over 4

=item utctime

The value returned will be a time value as returned by the C<time> function.

=item withzone

The value returned will be a reference to an array of two values. The first is the

same as with C<utctime>, the second is the timezone offset, in seconds, that was

used in the encoding.

=item raw

The value returned will be the raw encoding as extracted from the PDU.

=back

=item bigint

If during decoding any big integers are discovered (integers greater

than 32 bits), they will be decoded into big integer objects. This option

controls the big integer class into which the objects will be blessed.

The default is to use Math::BigInt.

=item null

The value to decode ASN.1 NULL types into.

If not set, it defaults to C<1>.

=back

=head2 prepare ( ASN )

Compile the given ASN.1 descripton which can be passed as a string

or as a filehandle. The syntax used is very close to ASN.1, but has

a few differences. If the ASN decribes only one macro then encode/decode can be

called on this object. If ASN describes more than one ASN.1 macro then C<find>

must be called. The method returns undef on error.

=head2 prepare_file ( ASNPATH )

Compile the ASN.1 description to be read from the specified pathname.

=head2 find ( MACRO )

Find a macro from a prepared ASN.1 description. Returns an object which can

be used for encode/decode.

=head2 encode ( VARIABLES )

Encode a PDU. Top-level variable are passed as name-value pairs, or as a reference

to a hash containing them. Returns the encoded PDU, or undef on error.

=head2 decode ( PDU )

Decode the PDU, returns a reference to a hash containg the values for the PDU. Returns

undef if there was an error.

=head2 registeroid ( OID, HANDLER )

Register a handler for all ASN.1 elements

that are C<DEFINED BY> the given OID.

B<HANDLER> must be a Convert::ASN1 object, e.g. as returned by L</"find ( MACRO )">.

=head2 registertype ( NAME, OID, HANDLER )

Register a handler for all ASN.1 elements named C<NAME>,

that are C<DEFINED BY> the given OID.

B<HANDLER> must be a Convert::ASN1 object, e.g. as returned by L</"find ( MACRO )">.

=head1 EXPORTS

As well as providing an object interface for encoding/decoding PDUs Convert::ASN1

also provides the following functions.

=head2 IO Functions

=over 4

=item asn_recv ( SOCK, BUFFER, FLAGS )

Will read a single element from the socket SOCK into BUFFER.  FLAGS may

be MSG_PEEK as exported by C<Socket>. Returns the address of the sender,

or undef if there was an error. Some systems do not support the return

of the peer address when the socket is a connected socket, in these

cases the empty string will be returned. This is the same behaviour

as the C<recv> function in perl itself.

It is recommended that if the socket is of type SOCK_DGRAM then C<recv>

be called directly instead of calling C<asn_recv>.

=item asn_read ( FH, BUFFER, OFFSET )

=item asn_read ( FH, BUFFER )

Will read a single element from the filehandle FH into BUFFER. Returns the

number of bytes read if a complete element was read, -1 if an incomplete

element was read or undef if there was an error. If OFFSET is specified

then it is assumed that BUFFER already contains an incomplete element

and new data will be appended starting at OFFSET.

If FH is a socket the asn_recv is used to read the element, so the same

restiction applies if FH is a socket of type SOCK_DGRAM.

=item asn_send ( SOCK, BUFFER, FLAGS, TO )

=item asn_send ( SOCK, BUFFER, FLAGS )

Identical to calling C<send>, see L<perlfunc>

=item asn_write ( FH, BUFFER )

Identical to calling C<syswrite> with 2 arguments, see L<perlfunc>

=item asn_get ( FH )

C<asn_get> provides buffered IO. Because it needs a buffer FH must be a GLOB

or a reference to a GLOB. C<asn_get> will use two entries in the hash element

of the GLOB to use as its buffer:

  asn_buffer - input buffer

  asn_need   - number of bytes needed for the next element, if known

Returns an element or undef if there was an error.

=item asn_ready ( FH )

C<asn_ready> works with C<asn_get>. It will return true if C<asn_get> has already

read enough data into the buffer to return a complete element.

=back

=head2 Encode/Decode Functions

=over 4

=item asn_tag ( CLASS, VALUE )

Given B<CLASS> and a B<VALUE>, calculate an integer which when encoded

will become the tag.

=item asn_decode_tag ( TAG )

Decode the given ASN.1 encoded C<TAG>.

=item asn_encode_tag ( TAG )

Encode B<TAG> value for encoding.

We assume that the tag has been correctly generated with L</"asn_tag ( CLASS, VALUE )">.

=item asn_decode_length ( LEN )

Decode the given ASN.1 decoded C<LEN>.

=item asn_encode_length ( LEN )

Encode the given C<LEN> to its ASN.1 encoding.

=back

=head2 Constants

=over 4

=item ASN_BIT_STR

=item ASN_BOOLEAN

=item ASN_ENUMERATED

=item ASN_GENERAL_TIME

=item ASN_IA5_STR

=item ASN_INTEGER

=item ASN_NULL

=item ASN_OBJECT_ID

=item ASN_OCTET_STR

=item ASN_PRINT_STR

=item ASN_REAL

=item ASN_SEQUENCE

=item ASN_SET

=item ASN_UTC_TIME

=item ASN_APPLICATION

=item ASN_CONTEXT

=item ASN_PRIVATE

=item ASN_UNIVERSAL

=item ASN_PRIMITIVE

=item ASN_CONSTRUCTOR

=item ASN_LONG_LEN

=item ASN_EXTENSION_ID

=item ASN_BIT

=back

=head2 Debug Functions

=over 4

=item asn_dump ( [FH,] BUFFER )

Try to decode the given buffer as ASN.1 structure and dump it to the

given file handle, or C<STDERR> if the handle is not given.

=item asn_hexdump ( FH, BUFFER )

=back

=head1 EXPORT TAGS

=over 4

=item :all

All exported functions

=item :const

ASN_BOOLEAN,     ASN_INTEGER,      ASN_BIT_STR,      ASN_OCTET_STR,

ASN_NULL,        ASN_OBJECT_ID,    ASN_REAL,         ASN_ENUMERATED,

ASN_SEQUENCE,    ASN_SET,          ASN_PRINT_STR,    ASN_IA5_STR,

ASN_UTC_TIME,    ASN_GENERAL_TIME,

ASN_UNIVERSAL,   ASN_APPLICATION,  ASN_CONTEXT,      ASN_PRIVATE,

ASN_PRIMITIVE,   ASN_CONSTRUCTOR,  ASN_LONG_LEN,     ASN_EXTENSION_ID, ASN_BIT

=item :debug

asn_dump, asn_hexdump

=item :io

asn_recv, asn_send, asn_read, asn_write, asn_get, asn_ready

=item :tag

asn_tag, asn_decode_tag, asn_encode_tag, asn_decode_length, asn_encode_length

=back

=head1 MAPPING ASN.1 TO PERL

Every element in the ASN.1 definition has a name, in perl a hash is used

with these names as an index and the element value as the hash value.

  # ASN.1

  int INTEGER,

  str OCTET STRING

  # Perl

  { int => 5, str => "text" }

In the case of a SEQUENCE, SET or CHOICE then the value in the namespace will

be a hash reference which will be the namespce for the elements with

that element.

  # ASN.1

  int INTEGER,

  seq SEQUENCE {

    str OCTET STRING,

    bool BOOLEAN

  }

  # Perl

  { int => 5, seq => { str => "text", bool => 1}}

If the element is a SEQUENCE OF, or SET OF, then the value in the namespace

will be an array reference. The elements in the array will be of

the type expected by the type following the OF. For example

with "SEQUENCE OF STRING" the array would contain strings. With

"SEQUENCE OF SEQUENCE { ... }" the array will contain hash references

which will be used as namespaces

  # ASN.1

  int INTEGER,

  str SEQUENCE OF OCTET STRING

  # Perl

  { int => 5, str => [ "text1", "text2"]}

  # ASN.1

  int INTEGER,

  str SEQUENCE OF SEQUENCE {

    type OCTET STRING,

    value INTEGER

  }

  # Perl

  { int => 5, str => [

    { type => "abc", value => 4 },

    { type => "def", value => -1 },

  ]}

Finally, if you wish to pre-parse ASN.1 and hold it to include

inline in your PDU, you can coerce it into the ASN.1 spec by

defining the value as ANY in the schema, and then pass the pre

encoded value inline.

  # ASN.1

  int INTEGER,

  str OCTET STRING,

  pre ANY

  # Perl

  { int => 5, str => "text", pre=>"\x03\x03\x00\x0a\x05" }

passes a pre-encoded BIT STRING instance as hex text. -But

it could be a previous run of $obj->encode() from another run

held in some variable.

=head2 Exceptions

There are some exceptions where Convert::ASN1 does not require an element to be named.

These are SEQUENCE {...}, SET {...} and CHOICE. In each case if the element is not

given a name then the elements inside the {...} will share the same namespace as

the elements outside of the {...}.

=head1 TODO

=over 4

=item *

XS implementation.

=item *

More documentation.

=item *

More tests.

=back

=head1 AUTHOR

Graham Barr <gbarr@cpan.org>

=head1 SUPPORT

Report issues via github at https://github.com/gbarr/perl-Convert-ASN1/issues

To contribute I encourage you to create a git fork of the repository at

https://github.com/gbarr/perl-Convert-ASN1 do you work on a fresh branch

created from master and submit a pull request

=head1 COPYRIGHT

Copyright (c) 2000-2012 Graham Barr <gbarr@cpan.org>. All rights reserved.

This program is free software; you can redistribute it and/or

modify it under the same terms as Perl itself.

=cut