path: root/lib/asn1/doc/src/asn1_getting_started.xml

<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE chapter SYSTEM "chapter.dtd">

<chapter>
  <header>
    <copyright>
      <year>1997</year><year>2013</year>
      <holder>Ericsson AB. All Rights Reserved.</holder>
    </copyright>
    <legalnotice>
      The contents of this file are subject to the Erlang Public License,
      Version 1.1, (the "License"); you may not use this file except in
      compliance with the License. You should have received a copy of the
      Erlang Public License along with this software. If not, it can be
      retrieved online at http://www.erlang.org/.
    
      Software distributed under the License is distributed on an "AS IS"
      basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
      the License for the specific language governing rights and limitations
      under the License.
    
    </legalnotice>

    <title>Getting Started</title>
    <prepared>Kenneth Lundin</prepared>
    <docno></docno>
    <date>1999-03-25</date>
    <rev>D</rev>
    <file>asn1_getting_started.xml</file>
  </header>

  <section>
    <title>Example</title>
      <p>The following example demonstrates the basic functionality used to
      run the Erlang ASN.1 compiler.</p>
      <p>Create a file named <c>People.asn</c> containing the following:</p>
      <pre>
People DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
  Person ::= SEQUENCE {
    name PrintableString,
    location INTEGER {home(0),field(1),roving(2)},
    age INTEGER OPTIONAL
  }
END      </pre>
      <p>This file must be compiled before it can be used.
        The ASN.1 compiler checks that the syntax is correct and that the
        text represents proper ASN.1 code before generating an abstract
        syntax tree. The code-generator then uses the abstract syntax
        tree to generate code.</p>
      <p>The generated Erlang files are placed in the current directory or
        in the directory specified with option <c>{outdir,Dir}</c>.</p>
      <p>The following shows how the compiler
        can be called from the  Erlang shell:</p>

      <pre>
1><input> asn1ct:compile("People", [ber]).</input>
ok
2>      </pre>

      <p>Option <c>verbose</c> can be added to get information
      about the generated files:</p>
      <pre>
2><input> asn1ct:compile("People", [ber,verbose]).</input>
Erlang ASN.1 compiling "People.asn" 
--{generated,"People.asn1db"}--
--{generated,"People.hrl"}--
--{generated,"People.erl"}--
ok
3>      </pre>

      <p>ASN.1 module <c>People</c> is now accepted and the
      abstract syntax tree is saved in file <c>People.asn1db</c>.
      The generated Erlang code is compiled using the Erlang compiler
      and loaded into the Erlang runtime system. There is now an API
      for <c>encode/2</c> and <c>decode/2</c> in module
      <c>People</c>, which is called like:<br></br>
      <c><![CDATA['People':encode(<Type name>, <Value>)]]></c>
      <br></br>
        or<br></br>
<c><![CDATA['People':decode(<Type name>, <Value>)]]></c></p>

      <p>Assume that there is a network
        application that receives instances of the ASN.1 defined
        type <c>Person</c>, modifies, and sends them back again:</p>

      <code type="none">
receive
   {Port,{data,Bytes}} ->
       case 'People':decode('Person',Bytes) of
           {ok,P} ->
               {ok,Answer} = 'People':encode('Person',mk_answer(P)),
               Port ! {self(),{command,Answer}};
           {error,Reason} ->
               exit({error,Reason})
       end
    end,      </code>
      <p>In this example, a series of bytes is received from an
        external source and the bytes are then decoded into a valid
        Erlang term. This was achieved with the call
        <c>'People':decode('Person',Bytes)</c>, which returned
        an Erlang value of the ASN.1 type <c>Person</c>. Then an answer was
        constructed  and encoded using
        <c>'People':encode('Person',Answer)</c>, which takes an
        instance of a defined ASN.1 type and transforms it to a
        binary according to the BER or PER encoding rules.</p>
        <p>The encoder and decoder can also be run from the shell:</p>
      <pre>
2> <input>Rockstar = {'Person',"Some Name",roving,50}.</input>
{'Person',"Some Name",roving,50}
3> <input>{ok,Bin} = 'People':encode('Person',Rockstar).</input>
{ok,&lt;&lt;243,17,19,9,83,111,109,101,32,78,97,109,101,2,1,2,
      2,1,50&gt;&gt;}
4> <input>{ok,Person} = 'People':decode('Person',Bin).</input>
{ok,{'Person',"Some Name",roving,50}}
5>      </pre>

    <section>
      <title>Module Dependencies</title>
      <p>It is common that ASN.1 modules import defined types, values, and
        other entities from another ASN.1 module.</p>
      <p>Earlier versions of the ASN.1 compiler required that modules
        that were imported from had to be compiled before the module
        that imported. This caused problems when ASN.1 modules had circular
        dependencies.</p>
      <p>Referenced modules are now parsed when the compiler finds an
        entity that is imported. No code is generated for
        the referenced module. However, the compiled modules rely on
        that the referenced modules are also compiled.</p>
    </section>
  </section>

  <section>
    <title>ASN.1 Application User Interface</title>
    <p>The <c>ASN.1</c> application provides the following two
    separate user interfaces:</p>
    <list type="bulleted">
      <item>
        <p>The module <c>asn1ct</c>, which provides the compile-time functions
          (including the compiler)</p>
      </item>
      <item>
        <p>The module <c>asn1rt_nif</c>, which provides the runtime functions
         for the ASN.1 decoder for the BER back end</p>
      </item>
    </list>
    <p>The reason for this division of the interfaces into compile-time
      and runtime
      is that only runtime modules (<c>asn1rt*</c>) need to be loaded in
      an embedded system. 
      </p>

    <section>
      <title>Compile-Time Functions</title>
      <p>The ASN.1 compiler can be started directly from the command line
        by the <c>erlc</c> program. This is convenient when compiling
        many ASN.1 files from the command line or when using Makefiles.
        Some examples of how the <c>erlc</c> command can be used to start
        the ASN.1 compiler:</p>
      <pre>
erlc Person.asn
erlc -bper Person.asn
erlc -bber ../Example.asn
erlc -o ../asnfiles -I ../asnfiles -I /usr/local/standards/asn1 Person.asn</pre>
      <p>Useful options for the ASN.1 compiler:</p>
      <taglist>
        <tag><c>-b[ber | per | uper]</c></tag>
        <item>
          <p>Choice of encoding rules. If omitted, <c>ber</c> is the
          default.</p>
        </item>
        <tag><c>-o OutDirectory</c></tag>
        <item>
          <p>Where to put the generated files. Default is the current
            directory.</p>
        </item>
        <tag><c>-I IncludeDir</c></tag>
        <item>
          <p>Where to search for <c>.asn1db</c> files and ASN.1
            source specs to resolve references to other
            modules. This option can be repeated many times if there
            are several places to search in. The compiler
            searches the current directory first.</p>
        </item>
        <tag><c>+der</c></tag>
        <item>
          <p>DER encoding rule. Only when using option <c>-ber</c>.</p>
        </item>
        <tag><c>+asn1config</c></tag>
        <item>
          <p>This functionality works together with option
            <c>ber</c>. It enables the specialized decodes, see Section
            <seealso marker="asn1_spec">Specialized Decode</seealso>.</p>
        </item>
        <tag><c>+undec_rest</c></tag>
        <item>
          <p>A buffer that holds a message being decoded can also have
          trailing bytes. If those trailing bytes are important, they
          can be returned along with the decoded value by compiling
          the ASN.1 specification with option <c>+undec_rest</c>.
          The return value from the decoder is
          <c>{ok,Value,Rest}</c> where <c>Rest</c> is a binary
          containing the trailing bytes.</p>
        </item>
        <tag><c>+'Any Erlc Option'</c></tag>
        <item>
          <p>Any option can be added to the Erlang compiler when
            compiling the generated Erlang files. Any option
            unrecognized by the ASN.1 compiler is passed to the
            Erlang compiler.</p>
        </item>
      </taglist>
      <p>For a complete description of <c>erlc</c>, see
        ERTS Reference Manual.</p>
      <p>The compiler and other compile-time functions can also be started
        from the Erlang shell. Here follows a brief
        description of the primary functions. For a
        complete description of each function, see module <c>asn1ct</c> in
        the <seealso marker="asn1ct">ASN.1 Reference Manual</seealso>.</p>
      <p>The compiler is started by <c>asn1ct:compile/1</c> with
        default options, or <c>asn1ct:compile/2</c> if explicit options
        are given.</p>
      <p>Example:</p>
      <pre>
asn1ct:compile("H323-MESSAGES.asn1").      </pre>
      <p>This equals:</p>
      <pre>
asn1ct:compile("H323-MESSAGES.asn1",[ber]).      </pre>
      <p>If PER encoding is wanted:</p>
      <pre>
asn1ct:compile("H323-MESSAGES.asn1",[per]).      </pre>
      <p>The generic encode and decode functions can be called
      as follows:</p>
      <pre>
'H323-MESSAGES':encode('SomeChoiceType',{call,&lt;&lt;"octetstring"&gt;&gt;}).
'H323-MESSAGES':decode('SomeChoiceType',Bytes).      </pre>
    </section>

    <section>
      <title>Runtime Functions</title>
      <p>When an ASN.1 specification is compiled with option <c>ber</c>,
      the <c>asn1rt_nif</c> module and the NIF library in
      <c>asn1/priv_dir</c> are needed at runtime.</p>
      <p>By calling function <c>info/0</c> in a generated module, you
      get information about which compiler options were used.</p>
    </section>

    <section>
      <title>Errors</title>
      <p>Errors detected at
        compile-time are displayed on the screen together with line
        numbers indicating where in the source file the respective error
        was detected. If no errors are found, an Erlang ASN.1 module is
        created.</p>
      <p>The runtime encoders and decoders execute within a catch and
      return <c>{ok, Data}</c> or
        <c>{error, {asn1, Description}}</c> where
        <c>Description</c> is
        an Erlang term describing the error.</p>
    </section>
  </section>

  <section>
    <marker id="inlineExamples"></marker>
    <title>Multi-File Compilation</title>
    <p>There are various reasons for using multi-file compilation:</p>
    <list type="bulleted">
      <item>To choose the name for the generated module, for
      example, because you need to compile the same specs for
      different encoding rules.</item>
      <item>You want only one resulting module.</item>
    </list>
    <p>Specify which ASN.1 specs to compile in a module with extension
      <c>.set.asn</c>. Choose a module name and provide the
      names of the ASN.1 specs. For example, if you have the specs
      <c>File1.asn</c>, <c>File2.asn</c>, and <c>File3.asn</c>, your
      module <c>MyModule.set.asn</c> looks as follows:</p>
    <pre>
File1.asn
File2.asn
File3.asn    </pre>
    <p>If you compile with the following, the result is one merged
    module <c>MyModule.erl</c> with the generated code from the three
    ASN.1 specs:</p>
    <code type="none">
~> erlc MyModule.set.asn    </code>
  </section>

  <section>
    <title>Remark about Tags</title>

    <p>Tags used to be important for all users of ASN.1, because it
    was necessary to add tags manually to certain constructs in order
    for the ASN.1 specification to be valid. Example of
    an old-style specification:</p>

    <pre>
Tags DEFINITIONS ::=
BEGIN
  Afters ::= CHOICE { cheese [0] IA5String,
                      dessert [1] IA5String }
END </pre>

    <p>Without the tags (the numbers in square brackets) the ASN.1
    compiler refused to compile the file.</p>

    <p>In 1994 the global tagging mode <c>AUTOMATIC TAGS</c> was introduced.
    By putting <c>AUTOMATIC TAGS</c> in the module header, the ASN.1
    compiler automatically adds tags when needed. The following is the
    same specification in <c>AUTOMATIC TAGS</c> mode:</p>

    <pre>
Tags DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
  Afters ::= CHOICE { cheese IA5String,
                      dessert IA5String }
END </pre>

    <p>Tags are not mentioned any more in this User's Guide.</p>
  </section>

  <section>
    <marker id="ASN1Types"></marker>
    <title>ASN.1 Types</title>
    <p>This section describes the ASN.1 types including their
      functionality, purpose, and how values are assigned in Erlang.
      </p>
    <p>ASN.1 has both primitive and constructed types:</p>
    <p></p>
    <table>
      <row>
        <cell align="left" valign="middle"><em>Primitive Types</em></cell>
        <cell align="left" valign="middle"><em>Constructed Types</em></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#BOOLEAN">BOOLEAN</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#SEQUENCE">SEQUENCE</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#INTEGER">INTEGER</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#SET">SET</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#REAL">REAL</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#CHOICE">CHOICE</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#NULL">NULL</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#SOF">SET OF and SEQUENCE OF</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#ENUMERATED">ENUMERATED</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#ANY">ANY</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#BIT STRING">BIT STRING</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#ANY">ANY DEFINED BY</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#OCTET STRING">OCTET STRING</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#NegotiationTypes">EXTERNAL</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#Character Strings">Character Strings</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#NegotiationTypes">EMBEDDED PDV</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#OBJECT IDENTIFIER">OBJECT IDENTIFIER</seealso></cell>
        <cell align="left" valign="middle"><seealso marker="#NegotiationTypes">CHARACTER STRING</seealso></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#Object Descriptor">Object Descriptor</seealso></cell>
        <cell align="left" valign="middle"></cell>
      </row>
      <row>
        <cell align="left" valign="middle"><seealso marker="#The TIME types">TIME Types</seealso></cell>
        <cell align="left" valign="middle"></cell>
      </row>
      <tcaption>Supported ASN.1 Types</tcaption>
    </table>
    <marker id="TypeNameValue"></marker>
    <note>
      <p>The values of each ASN.1 type have their own representation in Erlang, as
        described in the following sections. Users must provide
        these values for encoding according to the representation, as shown in the
        following example:</p>
    </note>
    <pre>
Operational ::= BOOLEAN --ASN.1 definition    </pre>
    <p>In Erlang code it can look as follows:</p>
    <pre>
Val = true,
{ok,Bytes} = MyModule:encode('Operational', Val),    </pre>

    <section>
      <marker id="BOOLEAN"></marker>
      <title>BOOLEAN</title>
      <p>Booleans in ASN.1 express values that can be either
        <c>TRUE</c> or <c>FALSE</c>.
        The meanings assigned to <c>TRUE</c> and <c>FALSE</c> are outside the scope
        of this text.</p>
      <p>In ASN.1 it is possible to have:</p>
      <pre>
Operational ::= BOOLEAN</pre>
      <p>Assigning a value to type <c>Operational</c> in Erlang is possible by
        using the following Erlang code:</p>
      <code type="erl">
Myvar1 = true,</code>
      <p>Thus, in Erlang the atoms <c>true</c> and <c>false</c> are used
        to encode a boolean value.</p>
    </section>

    <section>
      <marker id="INTEGER"></marker>
      <title>INTEGER</title>
      <p>ASN.1 itself specifies indefinitely large integers. Erlang
        systems with version 4.3 and higher support very large
        integers, in practice indefinitely large integers.</p>
      <p>The concept of subtyping can be applied to integers and
        to other ASN.1 types. The details of subtyping are not
        explained here; for more information, see X.680. Various
        syntaxes are allowed when defining a type as an integer:</p>
      <pre>
T1 ::= INTEGER
T2 ::= INTEGER (-2..7)
T3 ::= INTEGER (0..MAX)
T4 ::= INTEGER (0&lt;..MAX)
T5 ::= INTEGER (MIN&lt;..-99)
T6 ::= INTEGER {red(0),blue(1),white(2)}</pre>
      <p>The Erlang representation of an ASN.1 <c>INTEGER</c> is an integer or
        an atom if a <c>Named Number List</c> (see <c>T6</c> in the previous
        list) is specified.</p>
      <p>The following is an example of Erlang code that assigns values for the
        types in the previous list:</p>
      <pre>
T1value = 0,
T2value = 6,
T6value1 = blue,
T6value2 = 0,
T6value3 = white</pre>
      <p>These Erlang variables are now bound to valid instances of
        ASN.1 defined types. This style of value can be passed directly
        to the encoder for transformation into a series of bytes.</p>
      <p>The decoder returns an atom if the value corresponds to a
        symbol in the <c>Named Number List</c>.</p>
    </section>

    <section>
      <marker id="REAL"></marker>
      <title>REAL</title>
      <p>The following ASN.1 type is used for real numbers:</p>
      <pre>
R1 ::= REAL</pre>
      <p>It is assigned a value in Erlang as follows:</p>
      <pre>
R1value1 = "2.14",
R1value2 = {256,10,-2},</pre>
      <p>In the last line, notice that the tuple {256,10,-2} is the real number
        2.56 in a special notation, which encodes faster than simply
        stating the number as <c>"2.56"</c>. The arity three tuple is
        <c>{Mantissa,Base,Exponent}</c>, that is, Mantissa * Base^Exponent.</p>
    </section>

    <section>
      <marker id="NULL"></marker>
      <title>NULL</title>
      <p>The type <c>NULL</c> is suitable where supply and recognition of a value
        is important but the actual value is not.</p>
      <pre>
Notype ::= NULL</pre>
      <p>This type is assigned in Erlang as follows:</p>
      <pre>
N1 = 'NULL',</pre>
      <p>The actual value is the quoted atom <c>'NULL'</c>.</p>
    </section>

    <section>
      <marker id="ENUMERATED"></marker>
      <title>ENUMERATED</title>
      <p>The type <c>ENUMERATED</c> can be used when the value you want to
        describe can only take one of a set of predefined values. Example:</p>
      <pre>
DaysOfTheWeek ::= ENUMERATED { 
    sunday(1),monday(2),tuesday(3),
    wednesday(4),thursday(5),friday(6),saturday(7) }</pre>
      <p>For example, to assign a weekday value in Erlang, use the same atom
        as in the <c>Enumerations</c> of the type definition:</p>
      <pre>
Day1 = saturday,</pre>
      <p>The enumerated type is similar to an integer type, when
        defined with a set of predefined values. The difference is that
        an enumerated type can only have specified
        values, whereas an integer can have any value.</p>
    </section>

    <section>
      <marker id="BIT STRING"></marker>
      <title>BIT STRING</title>
      <p>The type <c>BIT STRING</c> can be used to model information that
        is made up of arbitrary length series of bits. It is intended
        to be used for selection of flags, not for binary files.</p>
      <p>In ASN.1, <c>BIT STRING</c> definitions can look as follows:</p>
      <pre>
Bits1 ::= BIT STRING
Bits2 ::= BIT STRING {foo(0),bar(1),gnu(2),gnome(3),punk(14)}</pre>
      <p>The following two notations are available for representation of <c>BIT
        STRING</c> values in Erlang and as input to the encode functions:</p>
      <list type="ordered">
        <item>A bitstring. By default, a <c>BIT STRING</c> with no
         symbolic names is decoded to an Erlang bitstring.</item>
        <item>A list of atoms corresponding to atoms in the <c>NamedBitList</c>
         in the <c>BIT STRING</c> definition. A <c>BIT STRING</c> with symbolic
         names is always decoded to the format shown in the following
         example:</item>
      </list>
      <pre>
Bits1Val1 = &lt;&lt;0:1,1:1,0:1,1:1,1:1&gt;&gt;,
Bits2Val1 = [gnu,punk],
Bits2Val2 = &lt;&lt;2#1110:4&gt;&gt;,
Bits2Val3 = [bar,gnu,gnome],</pre>
      <p><c>Bits2Val2</c> and <c>Bits2Val3</c> denote the same value.</p>
      <p><c>Bits2Val1</c> is assigned symbolic values. The assignment means
        that the bits corresponding to <c>gnu</c> and <c>punk</c>, that is, bits
        2 and 14 are set to 1, and the rest are set to 0. The symbolic values
        are shown as a list of values. If a named value, which is not
        specified in the type definition, is shown, a runtime error occurs.</p>
      <p><c>BIT STRING</c>s can also be subtyped with, for example, a <c>SIZE</c>
        specification:</p>
      <pre>
Bits3 ::= BIT STRING (SIZE(0..31))      </pre>
      <p>This means that no bit higher than 31 can be set.</p>

      <section>
	<title>Deprecated Representations for BIT STRING</title>
	<p>In addition to the representations described earlier, the
	following deprecated representations are available if the
	specification has been compiled with option
	<c>legacy_erlang_types</c>:</p>
	<list type="ordered">
	  <item>Aa a list of binary digits (0 or 1). This format is
	  accepted as input to the encode functions, and a <c>BIT STRING</c>
	  is decoded to this format if option
	  <em>legacy_bit_string</em> is given.
	  </item>
	  <item>As <c>{Unused,Binary}</c> where <c>Unused</c> denotes
	  how many trailing zero-bits 0-7 that are unused in the
	  least significant byte in <c>Binary</c>. This format is
	  accepted as input to the encode functions, and a <c>BIT
	  STRING</c> is decoded to this format if
	  <c>compact_bit_string</c> has been given.
	  </item>
	  <item>As a hexadecimal number (or an integer). Avoid this
	  as it is easy to misinterpret a <c>BIT
	  STRING</c> value in this format.
	  </item>
	</list>
      </section>
    </section>

    <section>
      <marker id="OCTET STRING"></marker>
      <title>OCTET STRING</title>
      <p><c>OCTET STRING</c> is the simplest of all ASN.1 types. <c>OCTET
      STRING</c> only moves or transfers, for example, binary files or other
      unstructured information complying with two rules: the
      bytes consist of octets and encoding is not required.</p>
      <p>It is possible to have the following ASN.1 type definitions:</p>
      <pre>
O1 ::= OCTET STRING
O2 ::= OCTET STRING (SIZE(28))      </pre>
      <p>With the following example assignments in Erlang:</p>
      <pre>
O1Val = &lt;&lt;17,13,19,20,0,0,255,254&gt;&gt;,
O2Val = &lt;&lt;"must be exactly 28 chars...."&gt;&gt;,</pre>
       <p>By default, an <c>OCTET STRING</c> is always represented as
       an Erlang binary. If the specification has been compiled with
       option <c>legacy_erlang_types</c>, the encode functions
       accept both lists and binaries, and the decode functions
       decode an <c>OCTET STRING</c> to a list.</p>
    </section>

    <section>
      <marker id="Character Strings"></marker>
      <title>Character Strings</title>
      <p>ASN.1 supports a wide variety of character sets. The main difference
        between an <c>OCTET STRING</c> and a character string is that the
        <c>OCTET STRING</c> has no imposed semantics on the bytes delivered.</p>
      <p>However, when using, for example, IA5String (which closely
        resembles ASCII), byte 65 (in decimal
        notation) <em>means</em> character 'A'.
        </p>
      <p>For example, if a defined type is to be a VideotexString and
        an octet is received with the unsigned integer value <c>X</c>,
        the octet is to be interpreted as specified in standard
        ITU-T T.100, T.101. 
        </p>
      <p>The  ASN.1 to Erlang compiler
        does not determine the correct interpretation of each BER
        string octet value with different character strings. The
        application is responsible for interpretation
        of octets. Therefore, from the BER
        string point of view, octets are very similar to
        character strings and are compiled in the same way.
        </p>
      <p>When PER is
        used, there is a significant difference in the encoding scheme
        between <c>OCTET STRING</c>s and other strings. The constraints
        specified for a type are especially important for PER, where
        they affect the encoding.
        </p>
      <p>Examples:</p>
      <pre>
Digs ::= NumericString (SIZE(1..3))
TextFile ::= IA5String (SIZE(0..64000))      </pre>
      <p>The corresponding Erlang assignments:</p>
      <pre>
DigsVal1 = "456",
DigsVal2 = "123",
TextFileVal1 = "abc...xyz...",
TextFileVal2 = [88,76,55,44,99,121 .......... a lot of characters here ....]</pre>
      <p>The Erlang representation for "BMPString" and
        "UniversalString" is either a list of ASCII values or a list
        of quadruples. The quadruple representation associates to the
        Unicode standard representation of characters. The ASCII
        characters are all represented by quadruples beginning with
        three zeros like {0,0,0,65} for character 'A'. When
        decoding a value for these strings, the result is a list of
        quadruples, or integers when the value is an ASCII character.</p>

      <p>The following example shows how it works. Assume the following
        specification is in file <c>PrimStrings.asn1</c>:</p>
       <pre>
PrimStrings DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
   BMP ::= BMPString
END    </pre>

       <p>Encoding and decoding some strings:</p>

      <pre>
1> <input>asn1ct:compile('PrimStrings', [ber]).</input>
ok
2> <input>{ok,Bytes1} = 'PrimStrings':encode('BMP', [{0,0,53,53},{0,0,45,56}]).</input>
{ok,&lt;&lt;30,4,53,54,45,56>>}
3> <input>'PrimStrings':decode('BMP', Bytes1).</input>
{ok,[{0,0,53,53},{0,0,45,56}]}
4> <input>{ok,Bytes2} = 'PrimStrings':encode('BMP', [{0,0,53,53},{0,0,0,65}]).</input>
{ok,&lt;&lt;30,4,53,53,0,65>>}
5> <input>'PrimStrings':decode('BMP', Bytes2).</input>
{ok,[{0,0,53,53},65]}
6> <input>{ok,Bytes3} = 'PrimStrings':encode('BMP', "BMP string").</input>
{ok,&lt;&lt;30,20,0,66,0,77,0,80,0,32,0,115,0,116,0,114,0,105,0,110,0,103>>}
7> <input>'PrimStrings':decode('BMP', Bytes3).</input>
{ok,"BMP string"}      </pre>

      <p>Type UTF8String is represented as a UTF-8 encoded binary in
      Erlang. Such binaries can be created directly using the binary syntax
      or by converting from a list of Unicode code points using function
      <c>unicode:characters_to_binary/1</c>.</p>

      <p>The following shows examples of how UTF-8 encoded binaries can
      be created and manipulated:</p>
      <pre>
1> <input>Gs = "Мой маленький Гном".</input>
[1052,1086,1081,32,1084,1072,1083,1077,1085,1100,1082,1080,
 1081,32,1043,1085,1086,1084]
2> <input>Gbin = unicode:characters_to_binary(Gs).</input>
&lt;&lt;208,156,208,190,208,185,32,208,188,208,176,208,187,208,
  181,208,189,209,140,208,186,208,184,208,185,32,208,147,
  208,...>>
3> <input>Gbin = &lt;&lt;"Мой маленький Гном"/utf8>>.</input>
&lt;&lt;208,156,208,190,208,185,32,208,188,208,176,208,187,208,
  181,208,189,209,140,208,186,208,184,208,185,32,208,147,
  208,...>>
4> <input>Gs = unicode:characters_to_list(Gbin).</input>
[1052,1086,1081,32,1084,1072,1083,1077,1085,1100,1082,1080,
 1081,32,1043,1085,1086,1084]</pre>

      <p>For details, see the <seealso marker="stdlib:unicode">unicode</seealso>
      module in <c>stdlib</c>.</p>

      <p>In the following example, this ASN.1 specification is used:</p>
      <pre>
UTF DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
   UTF ::= UTF8String
END   </pre>

      <p>Encoding and decoding a string with Unicode characters:</p>

      <pre>
5> <input>asn1ct:compile('UTF', [ber]).</input>
ok
6> <input>{ok,Bytes1} = 'UTF':encode('UTF', &lt;&lt;"Гном"/utf8>>).</input>
{ok,&lt;&lt;12,8,208,147,208,189,208,190,208,188>>}
7> <input>{ok,Bin1} = 'UTF':decode('UTF', Bytes1).</input>
{ok,&lt;&lt;208,147,208,189,208,190,208,188>>}
8> <input>io:format("~ts\n", [Bin1]).</input>
Гном
ok
9> <input>unicode:characters_to_list(Bin1).</input>
[1043,1085,1086,1084]   </pre>
    </section>

    <section>
      <marker id="OBJECT IDENTIFIER"></marker>
      <title>OBJECT IDENTIFIER</title>
      <p>The type <c>OBJECT IDENTIFIER</c> is used whenever a unique identity is
        required. An ASN.1 module, a transfer syntax, and so on, is identified
        with an <c>OBJECT IDENTIFIER</c>. Assume the following example:</p>
      <pre>
Oid ::= OBJECT IDENTIFIER</pre>
      <p>Therefore, the following example is a valid Erlang instance of
        type 'Oid':</p>
      <pre>
OidVal1 = {1,2,55},</pre>
      <p>The <c>OBJECT IDENTIFIER</c> value is simply a tuple with the
        consecutive values, which must be integers.
        </p>
      <p>The first value is limited to the values 0, 1, or 2. The
        second value must be in the range 0..39 when the first value
        is 0 or 1.
        </p>
      <p>The <c>OBJECT IDENTIFIER</c> is an important type and it is
        widely used within different standards to identify various
        objects uniquely. Dubuisson: ASN.1 - Communication Between
        Heterogeneous Systems includes an
        easy-to-understand description of the use of
        <c>OBJECT IDENTIFIER</c>.</p>
    </section>

    <section>
      <marker id="Object Descriptor"></marker>
      <title>Object Descriptor</title>
      <p>Values of this type can be assigned a value as an ordinary string
      as follows:</p>

      <pre>
      "This is the value of an Object descriptor"</pre>
    </section>

    <section>
      <marker id="The TIME types"></marker>
      <title>TIME Types</title>
      <p>Two time types are defined within ASN.1: Generalized
        Time and Universal Time Coordinated (UTC). Both are assigned a
        value as an ordinary string within double quotes, for example,
        "19820102070533.8".</p>
      <p>For DER encoding, the compiler does not check the validity
        of the time values. The DER requirements upon those strings are
        regarded as a matter for the application to fulfill.</p>
    </section>

    <section>
      <marker id="SEQUENCE"></marker>
      <title>SEQUENCE</title>
      <p>The structured types of ASN.1 are constructed from other types
        in a manner similar to the concepts of array and struct in  C.</p>
      <p>A <c>SEQUENCE</c> in ASN.1 is
        comparable with a struct in C and a record in Erlang.
        A <c>SEQUENCE</c> can be defined as follows:</p>
      <pre>
Pdu ::= SEQUENCE {
   a INTEGER,
   b REAL,
   c OBJECT IDENTIFIER,
   d NULL }      </pre>
      <p>This is a 4-component structure called <c>Pdu</c>. The record format
        is the major format for representation of <c>SEQUENCE</c> in Erlang.
        For each <c>SEQUENCE</c> and <c>SET</c> in an ASN.1 module an Erlang
        record declaration is generated. For <c>Pdu</c>, a record
        like the following is defined:</p>
      <pre>
-record('Pdu',{a, b, c, d}).      </pre>
      <p>The record declarations for a module <c>M</c> are placed in a
        separate <c>M.hrl</c> file.</p>
      <p>Values can be assigned in Erlang as follows:</p>
      <pre>
MyPdu = #'Pdu'{a=22,b=77.99,c={0,1,2,3,4},d='NULL'}.      </pre>
      <p>The decode functions return a record as result when decoding
      a <c>SEQUENCE</c> or a <c>SET</c>.</p>

      <p>A <c>SEQUENCE</c> and a <c>SET</c> can contain a component
      with a <c>DEFAULT</c> keyword followed by the actual value, which
      is the default value. The <c>DEFAULT</c> keyword means that the
      application doing the encoding can omit encoding of the value, which
      results in fewer bytes to send to the receiving application.</p>

      <p>An application can use the atom <c>asn1_DEFAULT</c> to indicate
      that the encoding is to be omitted for that position in
      the <c>SEQUENCE</c>.</p>

      <p>Depending on the encoding rules, the encoder can also compare
      the given value to the default value and automatically omit the
      encoding if the values are equal. How much effort the encoder makes
      to compare the values depends on the encoding rules. The DER
      encoding rules forbid encoding a value equal to the default value,
      so it has a more thorough and time-consuming comparison than the
      encoders for the other encoding rules.</p>

      <p>In the following example, this ASN.1 specification is used:</p>
      <pre>
File DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
Seq1 ::= SEQUENCE {
    a INTEGER DEFAULT 1,
    b Seq2 DEFAULT {aa TRUE, bb 15}
}

Seq2 ::= SEQUENCE {
    aa BOOLEAN,
    bb INTEGER
}

Seq3 ::= SEQUENCE {
    bs BIT STRING {a(0), b(1), c(2)} DEFAULT {a, c}
}
END </pre>
      <p>Example where the BER encoder is able to omit encoding
      of the default values:</p>
      <pre>
1> <input>asn1ct:compile('File', [ber]).</input>
ok
2> <input>'File':encode('Seq1', {'Seq1',asn1_DEFAULT,asn1_DEFAULT}).</input>
{ok,&lt;&lt;48,0>>}
3> <input>'File':encode('Seq1', {'Seq1',1,{'Seq2',true,15}}).</input>
{ok,&lt;&lt;48,0>>}   </pre>

     <p>Example with a named <c>BIT STRING</c> where the BER
     encoder does not omit the encoding:</p>
     <pre>
4> <input>'File':encode('Seq3', {'Seq3',asn1_DEFAULT).</input>
{ok,&lt;&lt;48,0>>}
5> <input>'File':encode('Seq3', {'Seq3',&lt;&lt;16#101:3>>).</input>
{ok,&lt;&lt;48,4,128,2,5,160>>}     </pre>

     <p>The DER encoder omits the encoding for the same <c>BIT STRING</c>:</p>
     <pre>
6> <input>asn1ct:compile('File', [ber,der]).</input>
ok
7> <input>'File':encode('Seq3', {'Seq3',asn1_DEFAULT).</input>
{ok,&lt;&lt;48,0>>}
8> <input>'File':encode('Seq3', {'Seq3',&lt;&lt;16#101:3>>).</input>
{ok,&lt;&lt;48,0>>}     </pre>
    </section>

    <section>
      <marker id="SET"></marker>
      <title>SET</title>
      <p>In Erlang, the <c>SET</c> type is used exactly as <c>SEQUENCE</c>.
      Notice that if BER or DER encoding rules are used, decoding a
      <c>SET</c> is slower than decoding a <c>SEQUENCE</c> because the
      components must be sorted.</p>
    </section>

    <section>
      <title>Extensibility for SEQUENCE and SET</title>
      <p>When a <c>SEQUENCE</c> or <c>SET</c> contains an extension marker
        and extension components as the following, the type can get more
        components in newer versions of the ASN.1 spec:</p>
      <pre>
SExt ::= SEQUENCE {
           a INTEGER,
           ...,
           b BOOLEAN }</pre>
      <p>In this case it has got a new
        component <c>b</c>. Thus, incoming messages that are decoded
        can have more or fever components than this one.
        </p>
      <p>The component <c>b</c> is treated as
        an original component when encoding a message. In this case, as
        it is not an optional element, it must be encoded.
        </p>
      <p>During decoding, the <c>b</c> field of the record gets the decoded
        value of the <c>b</c>
        component, if present, otherwise the value <c>asn1_NOVALUE</c>.</p>
    </section>

    <section>
      <marker id="CHOICE"></marker>
      <title>CHOICE</title>
      <p>The type <c>CHOICE</c> is a space saver and is similar to the
      concept of a 'union' in C.</p>
      <p>Assume the following:</p>
      <pre>
SomeModuleName DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
T ::= CHOICE {
        x REAL,
        y INTEGER,
        z OBJECT IDENTIFIER }
END </pre>
      <p>It is then possible to assign values as follows:</p>
      <pre>
TVal1 = {y,17},
TVal2 = {z,{0,1,2}},</pre>
      <p>A <c>CHOICE</c> value is always represented as the tuple
        <c>{ChoiceAlternative, Val}</c> where <c>ChoiceAlternative</c>
        is an atom denoting the selected choice alternative.
      </p>

      <section>
        <title>Extensible CHOICE</title>
        <p>When a <c>CHOICE</c> contains an extension marker and the
         decoder detects an unknown alternative of the <c>CHOICE</c>,
         the value is represented as follows:</p>
        <pre>
{asn1_ExtAlt, BytesForOpenType}</pre>
        <p>Here <c>BytesForOpenType</c> is a list of bytes constituting the
          encoding of the "unknown" <c>CHOICE</c> alternative.</p>
      </section>
    </section>

    <section>
      <marker id="SOF"></marker>
      <title>SET OF and SEQUENCE OF</title>
      <p>The types <c>SET OF</c> and <c>SEQUENCE OF</c> correspond
        to the concept of an array
        in several programming languages. The Erlang syntax for
        both types is straightforward, for example:</p>
      <pre>
Arr1 ::= SET SIZE (5) OF INTEGER (4..9) 
Arr2 ::= SEQUENCE OF OCTET STRING      </pre>
      <p>In Erlang the following can apply:</p>
      <pre>
Arr1Val = [4,5,6,7,8],
Arr2Val = ["abc",[14,34,54],"Octets"],      </pre>
      <p>Notice that the definition of type <c>SET OF</c> implies that
        the order of the components is undefined, but in practice there is 
        no difference between <c>SET OF</c> and <c>SEQUENCE OF</c>.
        The ASN.1 compiler for Erlang does not randomize the order of the
        <c>SET OF</c> components before encoding.</p>
      <p>However, for a value of type <c>SET OF</c>, the DER
        encoding format requires the elements to be sent in ascending
        order of their encoding, which implies an expensive sorting
        procedure in runtime. Therefore it is recommended to
        use <c>SEQUENCE OF</c> instead of <c>SET OF</c> if possible.</p>
    </section>

    <section>
      <marker id="ANY"></marker>
      <title>ANY and ANY DEFINED BY</title>
      <p>The types <c>ANY</c> and <c>ANY DEFINED BY</c> have been removed
        from the standard since 1994. It is recommended not to use
        these types any more. They can, however, exist in some old ASN.1
        modules. The idea with this type was to leave a "hole" in a 
        definition where it was possible to
        put unspecified data of any kind, even non-ASN.1 data.</p>
      <p>A value of this type is encoded as an <c>open type</c>.</p>
      <p>Instead of <c>ANY</c> and <c>ANY DEFINED BY</c>, it is
        recommended to use
        <c>information object class</c>, <c>table constraints</c>, and
        <c>parameterization</c>. In particular the construct
        <c>TYPE-IDENTIFIER.@Type</c> accomplish the same as the
        deprecated <c>ANY</c>.</p>
      <p>See also
      <seealso marker="#Information Object">Information object</seealso>.</p>
    </section>

    <section>
      <marker id="NegotiationTypes"></marker>
      <title>EXTERNAL, EMBEDDED PDV, and CHARACTER STRING</title>
      <p>The types <c>EXTERNAL</c>, <c>EMBEDDED PDV</c>, and
      <c>CHARACTER STRING</c> are used in presentation layer negotiation.
        They are encoded according to their associated type, see X.680.</p>
      <p>The type <c>EXTERNAL</c> had a slightly different associated type
        before 1994. X.691 states that encoding must follow
        the older associated type. So, generated encode/decode
        functions convert values of the newer format to the older format
        before encoding. This implies that it is allowed to use
        <c>EXTERNAL</c> type values of either format for encoding. Decoded
        values are always returned in the newer format.</p>
    </section>

    <section>
      <title>Embedded Named Types</title>
      <p>The structured types previously described can have other named
        types as their components. The general syntax to assign a value
        to component <c>C</c> of a named ASN.1 type <c>T</c> in Erlang
        is the record syntax <c>#'T'{'C'=Value}</c>.
        Here <c>Value</c> can be a value of yet another type <c>T2</c>,
        for example:</p>
      <pre>
EmbeddedExample DEFINITIONS AUTOMATIC TAGS ::=
BEGIN
B ::= SEQUENCE {
        a Arr1,
        b T }

Arr1 ::= SET SIZE (5) OF INTEGER (4..9) 

T ::= CHOICE {
        x REAL,
        y INTEGER,
        z OBJECT IDENTIFIER }
        END      </pre>
      <p><c>SEQUENCE</c> <c>b</c> can be encoded as follows in Erlang:</p>
      <pre>
1> 'EmbeddedExample':encode('B', {'B',[4,5,6,7,8],{x,"7.77"}}).
{ok,&lt;&lt;5,56,0,8,3,55,55,55,46,69,45,50>>} </pre>
    </section>
  </section>

  <section>
    <title>Naming of Records in .hrl Files</title>
    <p>When an ASN.1 specification is compiled, all defined types of type
      <c>SET</c> or <c>SEQUENCE</c> result in a corresponding record in the
      generated <c>.hrl</c> file. This is because the values for
      <c>SET</c> and <c>SEQUENCE</c> are represented as records as
      mentioned earlier.</p>
    <p>Some special cases of this functionality are presented in the
      next section.</p>

    <section>
      <title>Embedded Structured Types</title>
      <p>In ASN.1 it is also possible to have components that are themselves
        structured types.
        For example, it is possible to have the following:</p>
      <pre>
Emb ::= SEQUENCE {
    a SEQUENCE OF OCTET STRING,
    b SET {
       a INTEGER,
       b INTEGER DEFAULT 66},
    c CHOICE {
       a INTEGER,
       b FooType } }

FooType ::= [3] VisibleString      </pre>
      <p>The following records are generated because of type <c>Emb</c>:</p>
      <pre>
-record('Emb,{a, b, c}).
-record('Emb_b',{a, b = asn1_DEFAULT}). % the embedded SET type </pre>
      <p>Values of type <c>Emb</c> can be assigned as follows:</p>
      <code type="none">
V = #'Emb'{a=["qqqq",[1,2,255]], 
           b = #'Emb_b'{a=99}, 
           c ={b,"Can you see this"}}.</code>
      <p>For an embedded type of type <c>SEQUENCE</c>/<c>SET</c> in a
        <c>SEQUENCE</c>/<c>SET</c>, the record name is extended with an
        underscore and the component name. If the embedded structure is
        deeper with the <c>SEQUENCE</c>, <c>SET</c>, or <c>CHOICE</c>
        types in the line, each component name/alternative name is
        added to the record name.</p>
      <p>Example:</p>
      <pre>
Seq ::= SEQUENCE{
    a CHOICE{
        b SEQUENCE {
           c  INTEGER
        }
    }
}      </pre>
      <p>This results in the following record:</p>
      <pre>
-record('Seq_a_b',{c}).      </pre>
      <p>If the structured type has a component with an embedded
        <c>SEQUENCE OF</c>/<c>SET OF</c> which embedded type in turn
        is a <c>SEQUENCE</c>/<c>SET</c>, it gives a record with the
        <c>SEQUENCE OF</c>/<c>SET OF</c>
        addition as in the following example:</p>
      <pre>
Seq ::= SEQUENCE {
    a SEQUENCE OF SEQUENCE {
           b
               }
    c SET OF SEQUENCE {
           d
               }
}      </pre>
      <p>This results in the following records:</p>
      <pre>
-record('Seq_a_SEQOF'{b}).
-record('Seq_c_SETOF'{d}).      </pre>
      <p>A parameterized type is to be considered as an embedded
        type. Each time such a type is referenced, an instance of it is
        defined. Thus, in the following example a record with name
        <c>'Seq_b'</c> is generated in the <c>.hrl</c> file and is used
        to hold values:</p>
      <pre>
Seq ::= SEQUENCE {
    b PType{INTEGER}
}

PType{T} ::= SEQUENCE{
    id T
}      </pre>
    </section>

    <section>
      <title>Recursive Types</title>
      <p>Types that refer to themselves are called recursive types.
      Example:</p>
      <pre>
Rec ::= CHOICE {
     nothing NULL,
     something SEQUENCE {
          a INTEGER,
          b OCTET STRING,
          c Rec }}      </pre>
      <p>This is allowed in ASN.1 and the ASN.1-to-Erlang compiler
      supports this recursive type.
      A value for this type is assigned in Erlang as follows:</p>
      <pre>
V = {something,#'Rec_something'{a = 77, 
                                b = "some octets here", 
                                c = {nothing,'NULL'}}}.      </pre>
    </section>
  </section>

  <section>
    <title>ASN.1 Values</title>
    <p>Values can be assigned to an ASN.1 type within the ASN.1 code
      itself, as opposed to the actions in the previous section where
      a value was assigned to an ASN.1 type in Erlang. The full value
      syntax of ASN.1 is supported and X.680 describes in detail how
      to assign values in ASN.1. A short example:</p>
    <pre>
TT ::= SEQUENCE {
   a INTEGER,
   b SET OF OCTET STRING }

tt TT ::= {a 77,b {"kalle","kula"}}    </pre>
    <p>The value defined here can be used in several ways. It can, for
      example, be used as the value in some <c>DEFAULT</c> component:</p>
    <pre>
SS ::= SET {
    s OBJECT IDENTIFIER,
    val TT DEFAULT tt }    </pre>
    <p>It can also be used from inside an Erlang program. If this ASN.1
      code is defined in ASN.1 module <c>Values</c>, the ASN.1 value
      <c>tt</c> can be reached from Erlang as a function call to
      <c>'Values':tt()</c> as in the following example:</p>
    <pre>
1> <input>Val = 'Values':tt().</input>
{'TT',77,["kalle","kula"]}
2> <input>{ok,Bytes} = 'Values':encode('TT',Val).</input>
{ok,&lt;&lt;48,18,128,1,77,161,13,4,5,107,97,108,108,101,4,4,
      107,117,108,97&gt;&gt;}
4> <input>'Values':decode('TT',Bytes).</input>
{ok,{'TT',77,["kalle","kula"]}}
5>  </pre>
    <p>This example shows that a function is generated by the compiler
      that returns a valid Erlang representation of the value, although
      the value is of a complex type.</p>
    <p>Furthermore, a macro is generated for each value in the <c>.hrl</c>
      file. So, the defined value <c>tt</c> can also be extracted by
      <c>?tt</c> in application code.</p>
  </section>

  <section>
    <title>Macros</title>
    <p>The type <c>MACRO</c> is not supported. It is no longer part of
    the ASN.1 standard.</p>
  </section>

  <section>
    <marker id="Information Object"></marker>
    <title>ASN.1 Information Objects (X.681)</title>
    <p>Information Object Classes, Information Objects, and Information
      Object Sets (in the following called classes, objects, and
      object sets, respectively) are defined in the standard
      definition X.681. Only a brief explanation is given here.</p>
    <p>These constructs makes it possible to define open types, that
      is, values of that type can be of any ASN.1 type. Also,
      relationships can be defined between different types and
      values, as classes can hold types, values, objects, object
      sets, and other classes in their fields. A class can be
      defined in ASN.1 as follows:</p>
    <pre>
GENERAL-PROCEDURE ::= CLASS {
      &amp;Message,
      &amp;Reply               OPTIONAL,
      &amp;Error               OPTIONAL,
      &amp;id          PrintableString UNIQUE
}
WITH SYNTAX {
      NEW MESSAGE     &amp;Message
      [REPLY           &amp;Reply]
      [ERROR           &amp;Error]
      ADDRESS          &amp;id
}    </pre>
    <p>An object is an instance of a class. An object set is a set
      containing objects of a specified class. A definition can look
      as follows:</p>
    <pre>
object1 GENERAL-PROCEDURE ::= {
    NEW MESSAGE      PrintableString
    ADDRESS          "home"
}

object2 GENERAL-PROCEDURE ::= {
    NEW MESSAGE INTEGER
    ERROR INTEGER
    ADDRESS "remote"
}</pre>
    <p>The object <c>object1</c> is an instance of the class
      <c>GENERAL-PROCEDURE</c> and has one type field and one
      fixed type value field. The object <c>object2</c> has also an
      optional field <c>ERROR</c>, which is a type field. The field
      <c>ADDRESS</c> is a <c>UNIQUE</c> field. Objects in an object set
      must have unique values in their <c>UNIQUE</c> field, as in
      <c>GENERAL-PROCEDURES</c>:</p>
    <pre>
GENERAL-PROCEDURES GENERAL-PROCEDURE ::= {
    object1 | object2}    </pre>
    <p>You cannot encode a class, object, or object set, only refer to
      it when defining other ASN.1 entities. Typically you refer to a
      class as well as to object sets by table constraints and component
      relation constraints (X.682) in ASN.1 types, as in the following:</p>
    <pre>
StartMessage  ::= SEQUENCE {
    msgId  GENERAL-PROCEDURE.&amp;id  ({GENERAL-PROCEDURES}),
    content GENERAL-PROCEDURE.&amp;Message ({GENERAL-PROCEDURES}{@msgId}),
    }    </pre>
    <p>In type <c>StartMessage</c>, the constraint following field
      <c>content</c> tells that in a value of type
      <c>StartMessage</c> the value in field <c>content</c> must
      come from the same object that is chosen by field <c>msgId</c>.</p>
    <p>So, the value
      <c>#'StartMessage'{msgId="home",content="Any Printable String"}</c>
      is legal to encode as a <c>StartMessage</c> value. However, the value
      <c>#'StartMessage'{msgId="remote", content="Some String"}</c>
      is illegal as the constraint in <c>StartMessage</c> tells that
      when you have chosen a value from a specific object in object
      set <c>GENERAL-PROCEDURES</c> in field 
      <c>msgId</c>, you must choose a value from that same object in
      the content field too. In this second case, it is to be
      any <c>INTEGER</c> value.</p>
    <p><c>StartMessage</c> can in field <c>content</c> be
      encoded with a value of any type that an object in object set
      <c>GENERAL-PROCEDURES</c> has in its <c>NEW MESSAGE</c> field.
      This field refers to a type field
      <c>&amp;Message</c> in the class. Field <c>msgId</c> is always
      encoded as a <c>PrintableString</c>, as the field refers to a
      fixed type in the class.</p>
    <p>In practice, object sets are usually declared to be extensible so
      that more objects can be added to the set later. Extensibility is
      indicated as follows:</p>
    <pre>
GENERAL-PROCEDURES GENERAL-PROCEDURE ::= {
    object1 | object2, ...}    </pre>
     <p>When decoding a type that uses an extensible set constraint,
     it is always possible that the value in field <c>UNIQUE</c>
     is unknown (that is, the type has been encoded with a later
     version of the ASN.1 specification). The unencoded data is then
     returned wrapped in a tuple as follows:</p>

     <pre>
{asn1_OPENTYPE,Binary}</pre>

     <p>Here <c>Binary</c> is an Erlang binary that contains the encoded
     data. (If option <c>legacy_erlang_types</c> has been given,
     only the binary is returned.)</p>
  </section>

  <section>
    <title>Parameterization (X.683)</title>
    <p>Parameterization, which is defined in X.683, can be used when
      defining types, values, value sets, classes, objects, or object sets.
      A part of a definition can be supplied as a parameter. For
      example, if a <c>Type</c> is used in a definition with a certain
      purpose, you want the type name to express the intention. This
      can be done with parameterization.</p>
    <p>When many types (or another ASN.1 entity) only differ in some
      minor cases, but the structure of the types is similar, only
      one general type can be defined and the differences can be supplied
      through parameters.</p>
    <p>Example of use of parameterization:</p>
    <pre>
General{Type} ::= SEQUENCE
{
     number     INTEGER,
     string     Type
}
      
T1 ::= General{PrintableString}

T2 ::= General{BIT STRING}</pre>
    <p>An example of a value that can be encoded as type <c>T1</c> is
      <c>{12,"hello"}</c>.</p>
    <p>Notice that the compiler does not generate encode/decode functions
      for parameterized types, only for the instances of the parameterized
      types. Therefore, if a file contains the types <c>General{}</c>,
      <c>T1</c>, and <c>T2</c> as in the previous example, encode/decode
      functions are only generated for <c>T1</c> and <c>T2</c>.
      </p>
  </section>
</chapter>