create_table.sgml

<!--
$Header: /cvsroot/pgsql/doc/src/sgml/ref/create_table.sgml,v 1.39 2001/01/05 06:34:16 tgl Exp $
Postgres documentation
-->

<refentry id="SQL-CREATETABLE">
 <refmeta>
  <refentrytitle id="sql-createtable-title">
   CREATE TABLE
  </refentrytitle>
  <refmiscinfo>SQL - Language Statements</refmiscinfo>
 </refmeta>
 <refnamediv>
  <refname>
   CREATE TABLE
  </refname>
  <refpurpose>
   Creates a new table
  </refpurpose>
 </refnamediv>
 <refsynopsisdiv>
  <refsynopsisdivinfo>
   <date>2001-01-04</date>
  </refsynopsisdivinfo>
  <synopsis>
CREATE [ TEMPORARY | TEMP ] TABLE <replaceable class="PARAMETER">table</replaceable> (
    <replaceable class="PARAMETER">column</replaceable> <replaceable class="PARAMETER">type</replaceable>
    [ NULL | NOT NULL ] [ UNIQUE ] [ DEFAULT <replaceable class="PARAMETER">value</replaceable> ]
    [<replaceable>column_constraint_clause</replaceable> | PRIMARY KEY } [ ... ] ]
    [, ... ]
    [, PRIMARY KEY ( <replaceable class="PARAMETER">column</replaceable> [, ...] ) ]
    [, CHECK ( <replaceable class="PARAMETER">condition</replaceable> ) ]
    [, <replaceable>table_constraint_clause</replaceable> ]
    ) [ INHERITS ( <replaceable>inherited_table</replaceable> [, ...] ) ]
  </synopsis>
  
  <refsect2 id="R2-SQL-CREATETABLE-1">
   <title>
    Inputs
   </title>

   <para>

    <variablelist>
     <varlistentry>
      <term>TEMPORARY</term>
      <listitem>
       <para>
	The table is created only for this session, and is
	automatically dropped on session exit.
	Existing permanent tables with the same name are not visible
	while the temporary table exists.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term><replaceable class="PARAMETER">table</replaceable></term>
      <listitem>
       <para>
	The name of the new table to be created.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term><replaceable class="PARAMETER">column</replaceable></term>
      <listitem>
       <para>
	The name of a column.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term><replaceable class="PARAMETER">type</replaceable></term>
      <listitem>
       <para>
	The type of the column. This may include array specifiers.
	Refer to the <citetitle>PostgreSQL User's Guide</citetitle> for
	further information about data types and arrays.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term>DEFAULT <replaceable class="PARAMETER">value</replaceable></term>
      <listitem>
       <para>
	A default value for a column.
	See the DEFAULT clause for more information.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term><replaceable>column_constraint_clause</replaceable></term>
      <listitem>
       <para>
	The optional column constraint clauses specify a list of
	integrity constraints or tests which new or updated entries must
	satisfy for an insert or update operation to succeed. Each
	constraint must evaluate to a boolean expression. Although
	<acronym>SQL92</acronym> requires the <replaceable
	class="PARAMETER">column_constraint_clause</replaceable> to
	refer to that column only, <productname>Postgres</productname>
	allows multiple columns to be referenced within a single column
	constraint. See the column constraint clause for more
	information.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term><replaceable>table_constraint_clause</replaceable></term>
      <listitem>
       <para>
	The optional table CONSTRAINT clause specifies a
	list of integrity constraints which new or updated entries must
	satisfy for an insert or update operation to succeed. Each
	constraint must evaluate to a boolean expression. Multiple
	columns may be referenced within a single constraint. Only one
	PRIMARY KEY clause may be specified for a table;
	PRIMARY KEY <replaceable>column</replaceable> (a table
	constraint) and PRIMARY KEY (a column constraint)
	are mutually exclusive. See the table constraint clause for
	more information.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term>INHERITS <replaceable class="PARAMETER">inherited_table</replaceable></term>
      <listitem>
       <para>
	The optional INHERITS clause specifies a list of table
	names from which this table automatically inherits all fields.
	If any inherited field name appears more than once, 
	<productname>Postgres</productname>
	reports an error.
	<productname>Postgres</productname> automatically allows the created
	table to inherit functions on tables above it in the inheritance
	hierarchy.
       </para>
      </listitem>
     </varlistentry>
    </variablelist>
   </para>    
  </refsect2>

  <refsect2 id="R2-SQL-CREATETABLE-2">
   <title>
    Outputs
   </title>
   <para>

    <variablelist>
     <varlistentry>
      <term><computeroutput>
CREATE
       </computeroutput></term>
      <listitem>
       <para>
	Message returned if table is successfully created.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term><computeroutput>
ERROR
       </computeroutput></term>
      <listitem>
       <para>
	Message returned if table creation failed.
	This is usually accompanied by some descriptive text, such as:
	<computeroutput>
ERROR:  Relation '<replaceable class="parameter">table</replaceable>' already exists
	</computeroutput>
	, which occurs at runtime if the table specified already exists
	in the database.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term><computeroutput>
ERROR:  DEFAULT: type mismatched
       </computeroutput></term>
      <listitem>
       <para>
	If data type of default value doesn't match the
	column definition's data type.
       </para>
      </listitem>
     </varlistentry>
    </variablelist>
   </para>
  </refsect2>
 </refsynopsisdiv>

 <refsect1 id="R1-SQL-CREATETABLE-1">
  <title>
   Description
  </title>

  <para>
   <command>CREATE TABLE</command> will enter a new table
   into the current data base. The table will be "owned" by the user issuing the
   command.
  </para>

  <para>
   Each <replaceable class="PARAMETER">type</replaceable>
   may be a simple type, a complex type (set) or an array type.
   Each attribute may be specified to be non-null and
   each may have a default value, specified by the
   <xref linkend="R1-SQL-DEFAULTCLAUSE-1-TITLE" endterm="R1-SQL-DEFAULTCLAUSE-1-TITLE">.
  </para>

  <para>
   <note>
    <para>
     Consistent array dimensions within an
     attribute are not enforced. This will likely change in a future
     release.
    </para>
   </note>
  </para>

  <para>
   The optional INHERITS
   clause specifies a collection of table names from which this table
   automatically inherits all fields.  If any inherited field name
   appears more than once, Postgres reports an error.  Postgres automatically
   allows the created table to inherit functions on tables above it in
   the inheritance hierarchy.  Inheritance of functions is done according
   to the conventions of the Common Lisp Object System (CLOS).
  </para>

  <para>
   Each new table <replaceable class="PARAMETER">table</replaceable>
   is automatically created as a type.  Therefore, one or more rows
   from the table are automatically a type and can be used in 
   <xref linkend="sql-altertable" endterm="sql-altertable-title">
   or other <command>CREATE TABLE</command> statements.
  </para>

  <para>
   The new table is created as a heap with no initial data.
   A table can have no more than 1600 columns (in practice, the
   effective limit is lower because of tuple-length constraints).
   A table cannot have the same name as a system catalog table.
  </para>
 </refsect1>

 <refsect1 id="R1-SQL-DEFAULTCLAUSE-1">
  <title id="R1-SQL-DEFAULTCLAUSE-1-TITLE">
   DEFAULT Clause
  </title>
  <para>
   <synopsis>
DEFAULT <replaceable class="PARAMETER">value</replaceable>
   </synopsis>
  </para>
  <refsect2 id="R2-SQL-DEFAULTCLAUSE-1">
   <title>
    Inputs
   </title>
   <para>

    <variablelist>
     <varlistentry>
      <term><replaceable class="parameter">value</replaceable></term>
      <listitem>
       <para>
	The possible values for the default value expression are:
	<itemizedlist>
	 <listitem>
	  <simpara>
	   a literal value
	  </simpara>
	 </listitem>
	 <listitem>
	  <simpara>
	   a user function
	  </simpara>
	 </listitem>
	 <listitem>
	  <simpara>
	   a niladic function
	  </simpara>
	 </listitem>
	</itemizedlist>
       </para>
      </listitem>
     </varlistentry>	
    </variablelist>
   </para>
  </refsect2>

  <refsect2 id="R2-SQL-DEFAULTCLAUSE-2">
   <title>
    Outputs
   </title>
   <para>
    None.
   </para>
  </refsect2>
  
  <refsect2 id="R2-SQL-DEFAULTCLAUSE-3">
   <title>
    Description
   </title>
   <para>
    The DEFAULT clause assigns a default data value to a column
    (via a column definition in the CREATE TABLE statement). 
    The data type of a default value must match the column definition's
    data type.
   </para>
   <para>
    An INSERT operation that includes a column without a specified
    default value will assign the NULL value to the column
    if no explicit data value is provided for it.
    Default <replaceable class="parameter">literal</replaceable> means
    that the default is the specified constant value.
    Default <replaceable class="parameter">niladic-function</replaceable>
    or <replaceable class="parameter">user-function</replaceable> means
    that the default
    is the value of the specified function at the time of the INSERT.
   </para>
   <para>
    There are two types of niladic functions:
    <variablelist>
     <varlistentry>
      <term>niladic USER</term>
      <listitem>
       <variablelist>
	<varlistentry>
	 <term>CURRENT_USER / USER</term>
	 <listitem>
	  <simpara>See CURRENT_USER function</simpara>
	 </listitem>
	</varlistentry>
	<varlistentry>
	 <term>SESSION_USER</term>
	 <listitem>
	  <simpara>See CURRENT_USER function</simpara>
	 </listitem>
	</varlistentry>
	<varlistentry>
	 <term>SYSTEM_USER</term>
	 <listitem>
	  <simpara>Not implemented</simpara>
	 </listitem>
	</varlistentry>
       </variablelist>
      </listitem>
     </varlistentry>
     <varlistentry>
      <term>niladic datetime</term>
      <listitem>
       <variablelist>
	<varlistentry>
	 <term>CURRENT_DATE</term>
	 <listitem>
	  <simpara>See CURRENT_DATE function</simpara>
	 </listitem>
	</varlistentry>
	<varlistentry>
	 <term>CURRENT_TIME</term>
	 <listitem>
	  <simpara>See CURRENT_TIME function</simpara>
	 </listitem>
	</varlistentry>
	<varlistentry>
	 <term>CURRENT_TIMESTAMP</term>
	 <listitem>
	  <simpara>See CURRENT_TIMESTAMP function</simpara>
	 </listitem>
	</varlistentry>
       </variablelist>
      </listitem>
     </varlistentry>
    </variablelist>
   </para>

  </refsect2>
  <refsect2 id="R2-SQL-DEFAULTCLAUSE-4">
   <title>
    Usage
   </title>

   <para>
    To assign a constant value as the default for the
    columns <literal>did</literal> and <literal>number</literal>,
    and a string literal to the column <literal>did</literal>:

    <programlisting>
CREATE TABLE video_sales (
    did      VARCHAR(40) DEFAULT 'luso films',
    number   INTEGER DEFAULT 0,
    total    CASH DEFAULT '$0.0'
);
    </programlisting>
   </para>
   <para>
    To assign an existing sequence
    as the default for the column <literal>did</literal>,
    and a literal to the column <literal>name</literal>:

    <programlisting>
CREATE TABLE distributors (
    did      DECIMAL(3)  DEFAULT NEXTVAL('serial'),
    name     VARCHAR(40) DEFAULT 'luso films'
);
    </programlisting>
   </para>
  </refsect2>
 </refsect1>

 <refsect1 id="R1-SQL-COLUMNCONSTRAINT-1">
  <title id="R1-SQL-COLUMNCONSTRAINT-1-TITLE">
   Column CONSTRAINT Clause
  </title>
  <para>
   <synopsis>
[ CONSTRAINT <replaceable class="parameter">name</replaceable> ] { [
    NULL | NOT NULL ] | UNIQUE | PRIMARY KEY | CHECK <replaceable
     class="parameter">constraint</replaceable> | REFERENCES 
     <replaceable class="parameter">reftable</replaceable>
     (<replaceable class="parameter">refcolumn</replaceable>)
     [ MATCH <replaceable class="parameter">matchtype</replaceable> ]
     [ ON DELETE <replaceable class="parameter">action</replaceable> ] 
     [ ON UPDATE <replaceable class="parameter">action</replaceable> ]
     [ [ NOT ] DEFERRABLE ] 
     [ INITIALLY <replaceable class="parameter">checktime</replaceable> ] }
    [, ...]
   </synopsis>
  </para>

  <refsect2 id="R2-SQL-COLUMNCONSTRAINT-1">
   <title>
    Inputs
   </title>
   <para>

    <variablelist>
     <varlistentry>
      <term><replaceable class="parameter">name</replaceable></term>
      <listitem>
       <para>
	An arbitrary name given to the integrity constraint. 
	If <replaceable class="parameter">name</replaceable> is not specified,
	it is generated from the table and column names,
	which should ensure uniqueness for
	<replaceable class="parameter">name</replaceable>.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term>NULL</term>
      <listitem>
       <para>
	The column is allowed to contain NULL values. This is the default.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term>NOT NULL</term>
      <listitem>
       <para>
	The column is not allowed to contain NULL values.
	This is equivalent to the column constraint
	CHECK (<replaceable class="PARAMETER">column</replaceable> NOT NULL).
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term>UNIQUE</term>
      <listitem>
       <para>
	The column must have unique values. In <productname>Postgres</productname>
	this is enforced by an implicit creation of a unique index on the table.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term>PRIMARY KEY</term>
      <listitem>
       <para>
	This column is a primary key, which implies that uniqueness is
	enforced by the system and that other tables may rely on this
	column as a unique identifier for rows. See PRIMARY KEY for more
	information.
       </para>
      </listitem>
     </varlistentry>

     <varlistentry>
      <term>
<replaceable class="parameter">constraint</replaceable>
      </term>
      <listitem>
       <para>
	The definition of the constraint.
       </para>
      </listitem>
     </varlistentry>
    </variablelist>
   </para>
  </refsect2>

  <refsect2 id="R2-SQL-COLUMNCONSTRAINT-2">
   <title>
    Description
   </title>

   <para>
    The optional constraint clauses specify constraints or tests which
    new or updated entries must satisfy for an insert or update
    operation to succeed. Each constraint must evaluate to a boolean
    expression. Multiple attributes may be referenced within a single
    constraint. The use of PRIMARY KEY as a table constraint is mutually
    incompatible with PRIMARY KEY as a column constraint.
   </para>

   <para>
    A constraint is a named rule: an SQL object which helps define
    valid sets of values by putting limits on the results of INSERT,
    UPDATE or DELETE operations performed on a Base Table. 
   </para>

   <para>
    There are two ways to define integrity constraints:
    table constraints, covered later, and column constraints, covered here.
   </para>

   <para>
    A column constraint is an integrity constraint defined as part of a
    column definition, and logically becomes a table constraint as soon
    as it is created. The column constraints available are:

    <simplelist columns="1">
     <member>PRIMARY KEY</member>
     <member>REFERENCES</member>
     <member>UNIQUE</member>
     <member>CHECK</member>
     <member>NOT NULL</member>
    </simplelist>
   </para>
  </refsect2>
   
  <refsect2 id="R2-SQL-NOTNULL-1">
   <title>
    NOT NULL Constraint
   </title>
   <synopsis>
[ CONSTRAINT <replaceable class="parameter">name</replaceable> ] NOT NULL 
   </synopsis>
   <para>
    The NOT NULL constraint specifies a rule that a column may
    contain only non-null values. 
    This is a column constraint only, and not allowed
    as a table constraint.
   </para>

   <refsect3 id="R3-SQL-NOTNULL-1">
    <title>
     Outputs
    </title>
    <para>
     <variablelist>
      <varlistentry>
       <term><replaceable>status</replaceable></term>
       <listitem>
	<para>
	 <variablelist>
	  <varlistentry>
	   <term><computeroutput>
ERROR:  ExecAppend: Fail to add null value in not null attribute "<replaceable class="parameter">column</replaceable>".
	    </computeroutput></term>
	   <listitem>
	    <para>
	     This error occurs at runtime if one tries to insert a null value
	     into a column which has a NOT NULL constraint.
	    </para>
	   </listitem>
	  </varlistentry>
	 </variablelist>
	</para>
       </listitem>
      </varlistentry>
     </variablelist> 
    </para>
   </refsect3>

   <refsect3 id="R3-SQL-NOTNULL-2">
    <title>
     Description
    </title>
    <para>
    </para>
   </refsect3>

   <refsect3 id="R3-SQL-NOTNULL-3">
    <title>
     Usage
    </title>

    <para>
     Define two NOT NULL column constraints on the table
     <classname>distributors</classname>,
     one of which being a named constraint:

     <programlisting>
CREATE TABLE distributors (
    did      DECIMAL(3) CONSTRAINT no_null NOT NULL,
    name     VARCHAR(40) NOT NULL
);
     </programlisting>
    </para>
   </refsect3>
  </refsect2>

  <refsect2 id="R2-SQL-UNIQUECLAUSE-1">
   <title>
    UNIQUE Constraint
   </title>
   <synopsis>
[ CONSTRAINT <replaceable class="parameter">name</replaceable> ] UNIQUE
   </synopsis>

   <refsect3>
    <title>Inputs</title>
    <para>

     <variablelist>
      <varlistentry>
       <term>CONSTRAINT <replaceable class="parameter">name</replaceable></term>
       <listitem>
	<para>
	 An arbitrary label given to a constraint.
	</para>
       </listitem>
      </varlistentry>
     </variablelist>
    </para>
   </refsect3>
   
   <refsect3>
    <title>Outputs</title>
    <para>
     <variablelist>
      <varlistentry>
       <term><replaceable>status</replaceable></term>
       <listitem>
	<para>
	 <variablelist>
	  <varlistentry>
	   <term><computeroutput>
ERROR: Cannot insert a duplicate key into a unique index.
	    </computeroutput></term>
	   <listitem>
	    <para>
	     This error occurs at runtime if one tries to insert a
	     duplicate value into a column.
	    </para>
	   </listitem>
	  </varlistentry>
	 </variablelist>
	</para>
       </listitem>
      </varlistentry>
     </variablelist>
    </para>
   </refsect3>

   <refsect3>
    <title>
     Description
    </title>

    <para>
     The UNIQUE constraint specifies a rule that a group of one or
     more distinct columns of a table may contain only unique values.
    </para>
    <para>
     The column definitions of the specified columns do not have to
     include a NOT NULL constraint to be included in a UNIQUE
     constraint. Having more than one null value in a column without a
     NOT NULL constraint, does not violate a UNIQUE constraint. (This
     deviates from the <acronym>SQL92</acronym> definition, but is a
     more sensible convention. See the section on compatibility for more
     details.)
    </para>
    <para>
     Each UNIQUE column constraint must name a column that is
     different from the set of columns named by any other UNIQUE or
     PRIMARY KEY constraint defined for the table.
    </para>
    <note>
     <para>
      <productname>Postgres</productname> automatically creates a unique
      index for each UNIQUE constraint, to assure
      data integrity. See CREATE INDEX for more information.
     </para>
    </note>
   </refsect3>

   <refsect3 id="R3-SQL-UNIQUECLAUSE-3">
    <title>
     Usage
    </title>

    <para>
     Defines a UNIQUE column constraint for the table distributors.
     UNIQUE column constraints can only be defined on one column
     of the table:
     <programlisting>
CREATE TABLE distributors (
    did      DECIMAL(3),
    name     VARCHAR(40) UNIQUE
);
  </programlisting>

     which is equivalent to the following specified as a table constraint:
     <programlisting>
CREATE TABLE distributors (
    did      DECIMAL(3),
    name     VARCHAR(40),
    UNIQUE(name)
);
     </programlisting>
    </para>
   </refsect3>
  </refsect2>

  <refsect2 id="R2-SQL-CHECK-1">
   <title>
    The CHECK Constraint
   </title>
   <synopsis>
[ CONSTRAINT <replaceable class="parameter">name</replaceable> ] CHECK
    ( <replaceable>condition</replaceable> [, ...] ) 
   </synopsis>
   <refsect3 id="R3-SQL-CHECK-1">
    <title>Inputs</title>
    <para>

     <variablelist>
      <varlistentry>
       <term><replaceable class="parameter">name</replaceable></term>
       <listitem>
	<para>
	 An arbitrary name given to a constraint.
	</para>
       </listitem>
      </varlistentry>
      <varlistentry>
       <term><replaceable>condition</replaceable></term>
       <listitem>
	<para>
	 Any valid conditional expression evaluating to a boolean result.
	</para>
       </listitem>
      </varlistentry>
     </variablelist>
    </para>
   </refsect3>

   <refsect3 id="R3-SQL-CHECK-2">
    <title>
     Outputs
    </title>
    <para>

     <variablelist>
      <varlistentry>
       <term><replaceable>status</replaceable></term>
       <listitem>
	<para>

	 <variablelist>
	  <varlistentry>
	   <term><computeroutput>
ERROR:  ExecAppend: rejected due to CHECK constraint "<replaceable class="parameter">table_column</replaceable>".
	    </computeroutput></term>
	   <listitem>
	    <para>
	     This error occurs at runtime if one tries to insert an illegal
	     value into a column subject to a CHECK constraint.
	    </para>
	   </listitem>
	  </varlistentry>
	 </variablelist>
	</para>
       </listitem>
      </varlistentry>
     </variablelist>
    </para>
   </refsect3>

   <refsect3>
    <title>Description</title>
    <para>
     The CHECK constraint specifies a restriction on allowed values
     within a column. The CHECK constraint is also allowed as a table
     constraint.
    </para>
    <para>
     The SQL92 CHECK column constraints can only be defined on, and
     refer to, one column of the table.
     <productname>Postgres</productname> does not have this restriction.
    </para>
   </refsect3>
  </refsect2>
  
  <refsect2 id="R2-SQL-PRIMARYKEY-1">
   <title>
    PRIMARY KEY Constraint
   </title>
   <synopsis>
[ CONSTRAINT <replaceable class="PARAMETER">name</replaceable> ] PRIMARY KEY 
   </synopsis>

   <refsect3>
    <title>Inputs</title>
    <para>
     <variablelist>
      <varlistentry>
       <term>CONSTRAINT <replaceable class="PARAMETER">name</replaceable></term>
       <listitem>
	<para>
	 An arbitrary name for the constraint.
	</para>
       </listitem>
      </varlistentry>
     </variablelist>
    </para>
   </refsect3>

   <refsect3>
    <title>Outputs</title>
    <variablelist>
     <varlistentry>
      <term><computeroutput>
ERROR: Cannot insert a duplicate key into a unique index.
       </computeroutput></term>
      <listitem>
       <para>
	This occurs at runtime if one tries to insert a duplicate value into
	a column subject to a PRIMARY KEY constraint.
       </para>
      </listitem>
     </varlistentry>
    </variablelist>
   </refsect3>

   <refsect3>
    <title>Description</title>
    <para>
     The PRIMARY KEY column constraint specifies that a column of a
     table may contain only unique (non-duplicate), non-NULL values. The
     definition of the specified column does not have to include an
     explicit NOT NULL constraint to be included in a PRIMARY KEY
     constraint.
    </para>
    <para>
     Only one PRIMARY KEY can be specified for a table.
    </para>
   </refsect3>

   <refsect3 id="R3-SQL-PRIMARYKEY-3">
    <title>
     Notes
    </title>
    <para>
     <productname>Postgres</productname> automatically creates
     a unique index to assure
     data integrity (see CREATE INDEX statement).
    </para>
    <para>
     The PRIMARY KEY constraint should name a set of columns that is
     different from other sets of columns named by any UNIQUE constraint
     defined for the same table, since it will result in duplication
     of equivalent indexes and unproductive additional runtime overhead.
     However, <productname>Postgres</productname> does not specifically
     disallow this.
    </para>
   </refsect3>
  </refsect2>

  <refsect2 id="R2-SQL-REFERENCES-1">
   <refsect2info>
    <date>2000-02-04</date>
   </refsect2info>
   <title>
    REFERENCES Constraint
   </title>
   <synopsis>
[ CONSTRAINT <replaceable class="parameter">name</replaceable> ] REFERENCES <replaceable class="parameter">reftable</replaceable> [ ( <replaceable class="parameter">refcolumn</replaceable> ) ] 
    [ MATCH <replaceable class="parameter">matchtype</replaceable> ]
    [ ON DELETE <replaceable class="parameter">action</replaceable> ] 
    [ ON UPDATE <replaceable class="parameter">action</replaceable> ]
    [ [ NOT ] DEFERRABLE ] 
    [ INITIALLY <replaceable class="parameter">checktime</replaceable> ]
   </synopsis>
   <para>
    The REFERENCES constraint specifies a rule that a column
    value is checked against the values of another column.
    REFERENCES can also be specified as part of
    a FOREIGN KEY table constraint.
   </para>

   <refsect3 id="R3-SQL-REFERENCES-1">
    <title>Inputs</title>

    <para>
     <variablelist>
      <varlistentry>
       <term>CONSTRAINT <replaceable class="PARAMETER">name</replaceable></term>
       <listitem>
	<para>
	 An arbitrary name for the constraint.
	</para>
       </listitem>
      </varlistentry>

      <varlistentry>
       <term><replaceable class="parameter">reftable</replaceable></term>
       <listitem>
        <para>
         The table that contains the data to check against.
        </para>
       </listitem>
      </varlistentry>

      <varlistentry>
       <term><replaceable class="parameter">refcolumn</replaceable></term>
       <listitem>
        <para>
         The column in <replaceable class="parameter">reftable</replaceable>
         to check the data against.  If this is not specified, the PRIMARY KEY of the
         <replaceable class="parameter">reftable</replaceable> is used.
        </para>
       </listitem>
      </varlistentry>
      <varlistentry>
       <term>MATCH <replaceable class="parameter">matchtype</replaceable></term>
       <listitem>
        <para>
         There are three match types: MATCH FULL, MATCH PARTIAL, and a
         default match type if none is specified. MATCH FULL will not
         allow one column of a multi-column foreign key to be NULL
         unless all foreign key columns are NULL. The default MATCH type
         allows a some foreign key columns to be NULL while other parts
         of the foreign key are not NULL. MATCH PARTIAL is currently not
	 supported.
        </para>
       </listitem>
      </varlistentry>
      <varlistentry>
       <term>ON DELETE <replaceable class="parameter">action</replaceable></term>
       <listitem>
        <para>
	 The action to do when a referenced row in the referenced table is being
         deleted.  There are the following actions.
         <variablelist>
          <varlistentry>
           <term>NO ACTION</term>
           <listitem>
            <para>
             Produce error if foreign key violated.  This is the default.
            </para>
           </listitem>
          </varlistentry>
          <varlistentry>
           <term>RESTRICT</term>
           <listitem>
            <para>
             Same as NO ACTION.
            </para>
           </listitem>
          </varlistentry>
          <varlistentry>
           <term>CASCADE</term>
           <listitem>
            <para>
             Delete any rows referencing the deleted row.
            </para>
           </listitem>
          </varlistentry>
          <varlistentry>
           <term>SET NULL</term>
           <listitem>