doc/src/sgml/func/func-conditional.sgml - master - githubmirror/postgres

doc/src/sgml/func/func-conditional.sgml
283 строки9 KB
Andrew Dunstan
Split func.sgml into more manageable pieces
04 авг 2025, 16:04
04 авг 2025, 16:044e23c9e
 <sect1 id="functions-conditional">
  <title>Conditional Expressions</title>

  <indexterm>
   <primary>CASE</primary>
  </indexterm>

  <indexterm>
   <primary>conditional expression</primary>
  </indexterm>

  <para>
   This section describes the <acronym>SQL</acronym>-compliant conditional expressions
   available in <productname>PostgreSQL</productname>.
  </para>

  <tip>
   <para>
    If your needs go beyond the capabilities of these conditional
    expressions, you might want to consider writing a server-side function
    in a more expressive programming language.
   </para>
  </tip>

   <note>
    <para>
     Although <token>COALESCE</token>, <token>GREATEST</token>, and
     <token>LEAST</token> are syntactically similar to functions, they are
     not ordinary functions, and thus cannot be used with explicit
     <token>VARIADIC</token> array arguments.
    </para>
   </note>

  <sect2 id="functions-case">
   <title><literal>CASE</literal></title>

  <para>
   The <acronym>SQL</acronym> <token>CASE</token> expression is a
   generic conditional expression, similar to if/else statements in
   other programming languages:

<synopsis>
CASE WHEN <replaceable>condition</replaceable> THEN <replaceable>result</replaceable>
     <optional>WHEN ...</optional>
     <optional>ELSE <replaceable>result</replaceable></optional>
END
</synopsis>

   <token>CASE</token> clauses can be used wherever
   an expression is valid.  Each <replaceable>condition</replaceable> is an
   expression that returns a <type>boolean</type> result.  If the condition's
   result is true, the value of the <token>CASE</token> expression is the
   <replaceable>result</replaceable> that follows the condition, and the
   remainder of the <token>CASE</token> expression is not processed.  If the
   condition's result is not true, any subsequent <token>WHEN</token> clauses
   are examined in the same manner.  If no <token>WHEN</token>
   <replaceable>condition</replaceable> yields true, the value of the
   <token>CASE</token> expression is the <replaceable>result</replaceable> of the
   <token>ELSE</token> clause.  If the <token>ELSE</token> clause is
   omitted and no condition is true, the result is null.
  </para>

   <para>
    An example:
<screen>
SELECT * FROM test;

 a
---
 1
 2
 3


SELECT a,
       CASE WHEN a=1 THEN 'one'
            WHEN a=2 THEN 'two'
            ELSE 'other'
       END
    FROM test;

 a | case
---+-------
 1 | one
 2 | two
 3 | other
</screen>
   </para>

  <para>
   The data types of all the <replaceable>result</replaceable>
   expressions must be convertible to a single output type.
   See <xref linkend="typeconv-union-case"/> for more details.
  </para>

  <para>
   There is a <quote>simple</quote> form of <token>CASE</token> expression
   that is a variant of the general form above:

<synopsis>
CASE <replaceable>expression</replaceable>
    WHEN <replaceable>value</replaceable> THEN <replaceable>result</replaceable>
    <optional>WHEN ...</optional>
    <optional>ELSE <replaceable>result</replaceable></optional>
END
</synopsis>

   The first
   <replaceable>expression</replaceable> is computed, then compared to
   each of the <replaceable>value</replaceable> expressions in the
   <token>WHEN</token> clauses until one is found that is equal to it.  If
   no match is found, the <replaceable>result</replaceable> of the
   <token>ELSE</token> clause (or a null value) is returned.  This is similar
   to the <function>switch</function> statement in C.
  </para>

   <para>
    The example above can be written using the simple
    <token>CASE</token> syntax:
<screen>
SELECT a,
       CASE a WHEN 1 THEN 'one'
              WHEN 2 THEN 'two'
              ELSE 'other'
       END
    FROM test;

 a | case
---+-------
 1 | one
 2 | two
 3 | other
</screen>
   </para>

   <para>
    A <token>CASE</token> expression does not evaluate any subexpressions
    that are not needed to determine the result.  For example, this is a
    possible way of avoiding a division-by-zero failure:
<programlisting>
SELECT ... WHERE CASE WHEN x &lt;&gt; 0 THEN y/x &gt; 1.5 ELSE false END;
</programlisting>
   </para>

   <note>
    <para>
     As described in <xref linkend="syntax-express-eval"/>, there are various
     situations in which subexpressions of an expression are evaluated at
     different times, so that the principle that <quote><token>CASE</token>
     evaluates only necessary subexpressions</quote> is not ironclad.  For
     example a constant <literal>1/0</literal> subexpression will usually result in
     a division-by-zero failure at planning time, even if it's within
     a <token>CASE</token> arm that would never be entered at run time.
    </para>
   </note>
  </sect2>

  <sect2 id="functions-coalesce-nvl-ifnull">
   <title><literal>COALESCE</literal></title>

  <indexterm>
   <primary>COALESCE</primary>
  </indexterm>

  <indexterm>
   <primary>NVL</primary>
  </indexterm>

  <indexterm>
   <primary>IFNULL</primary>
  </indexterm>

<synopsis>
<function>COALESCE</function>(<replaceable>value</replaceable> <optional>, ...</optional>)
</synopsis>

  <para>
   The <function>COALESCE</function> function returns the first of its
   arguments that is not null.  Null is returned only if all arguments
   are null.  It is often used to substitute a default value for
   null values when data is retrieved for display, for example:
<programlisting>
SELECT COALESCE(description, short_description, '(none)') ...
</programlisting>
   This returns <varname>description</varname> if it is not null, otherwise
   <varname>short_description</varname> if it is not null, otherwise <literal>(none)</literal>.
  </para>

   <para>
    The arguments must all be convertible to a common data type, which
    will be the type of the result (see
    <xref linkend="typeconv-union-case"/> for details).
   </para>

   <para>
    Like a <token>CASE</token> expression, <function>COALESCE</function> only
    evaluates the arguments that are needed to determine the result;
    that is, arguments to the right of the first non-null argument are
    not evaluated.  This SQL-standard function provides capabilities similar
    to <function>NVL</function> and <function>IFNULL</function>, which are used in some other
    database systems.
   </para>
  </sect2>

  <sect2 id="functions-nullif">
   <title><literal>NULLIF</literal></title>

  <indexterm>
   <primary>NULLIF</primary>
  </indexterm>

<synopsis>
<function>NULLIF</function>(<replaceable>value1</replaceable>, <replaceable>value2</replaceable>)
</synopsis>

  <para>
   The <function>NULLIF</function> function returns a null value if
   <replaceable>value1</replaceable> equals <replaceable>value2</replaceable>;
   otherwise it returns <replaceable>value1</replaceable>.
   This can be used to perform the inverse operation of the
   <function>COALESCE</function> example given above:
<programlisting>
SELECT NULLIF(value, '(none)') ...
</programlisting>
   In this example, if <literal>value</literal> is <literal>(none)</literal>,
   null is returned, otherwise the value of <literal>value</literal>
   is returned.
  </para>

  <para>
   The two arguments must be of comparable types.
   To be specific, they are compared exactly as if you had
   written <literal><replaceable>value1</replaceable>
   = <replaceable>value2</replaceable></literal>, so there must be a
   suitable <literal>=</literal> operator available.
  </para>

  <para>
   The result has the same type as the first argument &mdash; but there is
   a subtlety.  What is actually returned is the first argument of the
   implied <literal>=</literal> operator, and in some cases that will have
   been promoted to match the second argument's type.  For
   example, <literal>NULLIF(1, 2.2)</literal> yields <type>numeric</type>,
   because there is no <type>integer</type> <literal>=</literal>
   <type>numeric</type> operator,
   only <type>numeric</type> <literal>=</literal> <type>numeric</type>.
  </para>

  </sect2>

  <sect2 id="functions-greatest-least">
   <title><literal>GREATEST</literal> and <literal>LEAST</literal></title>

  <indexterm>
   <primary>GREATEST</primary>
  </indexterm>
  <indexterm>
   <primary>LEAST</primary>
  </indexterm>

<synopsis>
<function>GREATEST</function>(<replaceable>value</replaceable> <optional>, ...</optional>)
</synopsis>
<synopsis>
<function>LEAST</function>(<replaceable>value</replaceable> <optional>, ...</optional>)
</synopsis>

   <para>
    The <function>GREATEST</function> and <function>LEAST</function> functions select the
    largest or smallest value from a list of any number of expressions.
    The expressions must all be convertible to a common data type, which
    will be the type of the result
    (see <xref linkend="typeconv-union-case"/> for details).
   </para>

   <para>
    NULL values in the argument list are ignored.  The result will be NULL
    only if all the expressions evaluate to NULL.  (This is a deviation from
    the SQL standard.  According to the standard, the return value is NULL if
    any argument is NULL.  Some other databases behave this way.)
   </para>
  </sect2>
 </sect1>