| Current File : //usr/share/doc/postgresql-9.2.24/html/plpgsql-control-structures.html |
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<HTML
><HEAD
><TITLE
>Control Structures</TITLE
><META
NAME="GENERATOR"
CONTENT="Modular DocBook HTML Stylesheet Version 1.79"><LINK
REV="MADE"
HREF="mailto:pgsql-docs@postgresql.org"><LINK
REL="HOME"
TITLE="PostgreSQL 9.2.24 Documentation"
HREF="index.html"><LINK
REL="UP"
TITLE="PL/pgSQL - SQL Procedural Language"
HREF="plpgsql.html"><LINK
REL="PREVIOUS"
TITLE="Basic Statements"
HREF="plpgsql-statements.html"><LINK
REL="NEXT"
TITLE="Cursors"
HREF="plpgsql-cursors.html"><LINK
REL="STYLESHEET"
TYPE="text/css"
HREF="stylesheet.css"><META
HTTP-EQUIV="Content-Type"
CONTENT="text/html; charset=ISO-8859-1"><META
NAME="creation"
CONTENT="2017-11-06T22:43:11"></HEAD
><BODY
CLASS="SECT1"
><DIV
CLASS="NAVHEADER"
><TABLE
SUMMARY="Header navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TH
COLSPAN="5"
ALIGN="center"
VALIGN="bottom"
><A
HREF="index.html"
>PostgreSQL 9.2.24 Documentation</A
></TH
></TR
><TR
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
TITLE="Basic Statements"
HREF="plpgsql-statements.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
HREF="plpgsql.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="60%"
ALIGN="center"
VALIGN="bottom"
>Chapter 39. <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> - <ACRONYM
CLASS="ACRONYM"
>SQL</ACRONYM
> Procedural Language</TD
><TD
WIDTH="20%"
ALIGN="right"
VALIGN="top"
><A
TITLE="Cursors"
HREF="plpgsql-cursors.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
></TABLE
><HR
ALIGN="LEFT"
WIDTH="100%"></DIV
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="PLPGSQL-CONTROL-STRUCTURES"
>39.6. Control Structures</A
></H1
><P
> Control structures are probably the most useful (and
important) part of <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
>. With
<SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
>'s control structures,
you can manipulate <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> data in a very
flexible and powerful way.
</P
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="PLPGSQL-STATEMENTS-RETURNING"
>39.6.1. Returning From a Function</A
></H2
><P
> There are two commands available that allow you to return data
from a function: <TT
CLASS="COMMAND"
>RETURN</TT
> and <TT
CLASS="COMMAND"
>RETURN
NEXT</TT
>.
</P
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN56796"
>39.6.1.1. <TT
CLASS="COMMAND"
>RETURN</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>RETURN <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
>;</PRE
><P
> <TT
CLASS="COMMAND"
>RETURN</TT
> with an expression terminates the
function and returns the value of
<TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> to the caller. This form
is used for <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> functions that do
not return a set.
</P
><P
> When returning a scalar type, any expression can be used. The
expression's result will be automatically cast into the
function's return type as described for assignments. To return a
composite (row) value, you must write a record or row variable
as the <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
>.
</P
><P
> If you declared the function with output parameters, write just
<TT
CLASS="COMMAND"
>RETURN</TT
> with no expression. The current values
of the output parameter variables will be returned.
</P
><P
> If you declared the function to return <TT
CLASS="TYPE"
>void</TT
>, a
<TT
CLASS="COMMAND"
>RETURN</TT
> statement can be used to exit the function
early; but do not write an expression following
<TT
CLASS="COMMAND"
>RETURN</TT
>.
</P
><P
> The return value of a function cannot be left undefined. If
control reaches the end of the top-level block of the function
without hitting a <TT
CLASS="COMMAND"
>RETURN</TT
> statement, a run-time
error will occur. This restriction does not apply to functions
with output parameters and functions returning <TT
CLASS="TYPE"
>void</TT
>,
however. In those cases a <TT
CLASS="COMMAND"
>RETURN</TT
> statement is
automatically executed if the top-level block finishes.
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN56817"
>39.6.1.2. <TT
CLASS="COMMAND"
>RETURN NEXT</TT
> and <TT
CLASS="COMMAND"
>RETURN QUERY</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>RETURN NEXT <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
>;
RETURN QUERY <TT
CLASS="REPLACEABLE"
><I
>query</I
></TT
>;
RETURN QUERY EXECUTE <TT
CLASS="REPLACEABLE"
><I
>command-string</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> USING <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
>, ... </SPAN
>] </SPAN
>];</PRE
><P
> When a <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> function is declared to return
<TT
CLASS="LITERAL"
>SETOF <TT
CLASS="REPLACEABLE"
><I
>sometype</I
></TT
></TT
>, the procedure
to follow is slightly different. In that case, the individual
items to return are specified by a sequence of <TT
CLASS="COMMAND"
>RETURN
NEXT</TT
> or <TT
CLASS="COMMAND"
>RETURN QUERY</TT
> commands, and
then a final <TT
CLASS="COMMAND"
>RETURN</TT
> command with no argument
is used to indicate that the function has finished executing.
<TT
CLASS="COMMAND"
>RETURN NEXT</TT
> can be used with both scalar and
composite data types; with a composite result type, an entire
<SPAN
CLASS="QUOTE"
>"table"</SPAN
> of results will be returned.
<TT
CLASS="COMMAND"
>RETURN QUERY</TT
> appends the results of executing
a query to the function's result set. <TT
CLASS="COMMAND"
>RETURN
NEXT</TT
> and <TT
CLASS="COMMAND"
>RETURN QUERY</TT
> can be freely
intermixed in a single set-returning function, in which case
their results will be concatenated.
</P
><P
> <TT
CLASS="COMMAND"
>RETURN NEXT</TT
> and <TT
CLASS="COMMAND"
>RETURN
QUERY</TT
> do not actually return from the function —
they simply append zero or more rows to the function's result
set. Execution then continues with the next statement in the
<SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> function. As successive
<TT
CLASS="COMMAND"
>RETURN NEXT</TT
> or <TT
CLASS="COMMAND"
>RETURN
QUERY</TT
> commands are executed, the result set is built
up. A final <TT
CLASS="COMMAND"
>RETURN</TT
>, which should have no
argument, causes control to exit the function (or you can just
let control reach the end of the function).
</P
><P
> <TT
CLASS="COMMAND"
>RETURN QUERY</TT
> has a variant
<TT
CLASS="COMMAND"
>RETURN QUERY EXECUTE</TT
>, which specifies the
query to be executed dynamically. Parameter expressions can
be inserted into the computed query string via <TT
CLASS="LITERAL"
>USING</TT
>,
in just the same way as in the <TT
CLASS="COMMAND"
>EXECUTE</TT
> command.
</P
><P
> If you declared the function with output parameters, write just
<TT
CLASS="COMMAND"
>RETURN NEXT</TT
> with no expression. On each
execution, the current values of the output parameter
variable(s) will be saved for eventual return as a row of the
result. Note that you must declare the function as returning
<TT
CLASS="LITERAL"
>SETOF record</TT
> when there are multiple output
parameters, or <TT
CLASS="LITERAL"
>SETOF <TT
CLASS="REPLACEABLE"
><I
>sometype</I
></TT
></TT
>
when there is just one output parameter of type
<TT
CLASS="REPLACEABLE"
><I
>sometype</I
></TT
>, in order to create a set-returning
function with output parameters.
</P
><P
> Here is an example of a function using <TT
CLASS="COMMAND"
>RETURN
NEXT</TT
>:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE TABLE foo (fooid INT, foosubid INT, fooname TEXT);
INSERT INTO foo VALUES (1, 2, 'three');
INSERT INTO foo VALUES (4, 5, 'six');
CREATE OR REPLACE FUNCTION getAllFoo() RETURNS SETOF foo AS
$BODY$
DECLARE
r foo%rowtype;
BEGIN
FOR r IN SELECT * FROM foo
WHERE fooid > 0
LOOP
-- can do some processing here
RETURN NEXT r; -- return current row of SELECT
END LOOP;
RETURN;
END
$BODY$
LANGUAGE plpgsql;
SELECT * FROM getallfoo();</PRE
><P>
</P
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> The current implementation of <TT
CLASS="COMMAND"
>RETURN NEXT</TT
>
and <TT
CLASS="COMMAND"
>RETURN QUERY</TT
> stores the entire result set
before returning from the function, as discussed above. That
means that if a <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> function produces a
very large result set, performance might be poor: data will be
written to disk to avoid memory exhaustion, but the function
itself will not return until the entire result set has been
generated. A future version of <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> might
allow users to define set-returning functions
that do not have this limitation. Currently, the point at
which data begins being written to disk is controlled by the
<A
HREF="runtime-config-resource.html#GUC-WORK-MEM"
>work_mem</A
>
configuration variable. Administrators who have sufficient
memory to store larger result sets in memory should consider
increasing this parameter.
</P
></BLOCKQUOTE
></DIV
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="PLPGSQL-CONDITIONALS"
>39.6.2. Conditionals</A
></H2
><P
> <TT
CLASS="COMMAND"
>IF</TT
> and <TT
CLASS="COMMAND"
>CASE</TT
> statements let you execute
alternative commands based on certain conditions.
<SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> has three forms of <TT
CLASS="COMMAND"
>IF</TT
>:
<P
></P
></P><UL
><LI
><P
><TT
CLASS="LITERAL"
>IF ... THEN</TT
></P
></LI
><LI
><P
><TT
CLASS="LITERAL"
>IF ... THEN ... ELSE</TT
></P
></LI
><LI
><P
><TT
CLASS="LITERAL"
>IF ... THEN ... ELSIF ... THEN ... ELSE</TT
></P
></LI
></UL
><P>
and two forms of <TT
CLASS="COMMAND"
>CASE</TT
>:
<P
></P
></P><UL
><LI
><P
><TT
CLASS="LITERAL"
>CASE ... WHEN ... THEN ... ELSE ... END CASE</TT
></P
></LI
><LI
><P
><TT
CLASS="LITERAL"
>CASE WHEN ... THEN ... ELSE ... END CASE</TT
></P
></LI
></UL
><P>
</P
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN56899"
>39.6.2.1. <TT
CLASS="LITERAL"
>IF-THEN</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>IF <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END IF;</PRE
><P
> <TT
CLASS="LITERAL"
>IF-THEN</TT
> statements are the simplest form of
<TT
CLASS="LITERAL"
>IF</TT
>. The statements between
<TT
CLASS="LITERAL"
>THEN</TT
> and <TT
CLASS="LITERAL"
>END IF</TT
> will be
executed if the condition is true. Otherwise, they are
skipped.
</P
><P
> Example:
</P><PRE
CLASS="PROGRAMLISTING"
>IF v_user_id <> 0 THEN
UPDATE users SET email = v_email WHERE user_id = v_user_id;
END IF;</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN56912"
>39.6.2.2. <TT
CLASS="LITERAL"
>IF-THEN-ELSE</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>IF <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
ELSE
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END IF;</PRE
><P
> <TT
CLASS="LITERAL"
>IF-THEN-ELSE</TT
> statements add to
<TT
CLASS="LITERAL"
>IF-THEN</TT
> by letting you specify an
alternative set of statements that should be executed if the
condition is not true. (Note this includes the case where the
condition evaluates to NULL.)
</P
><P
> Examples:
</P><PRE
CLASS="PROGRAMLISTING"
>IF parentid IS NULL OR parentid = ''
THEN
RETURN fullname;
ELSE
RETURN hp_true_filename(parentid) || '/' || fullname;
END IF;</PRE
><P>
</P><PRE
CLASS="PROGRAMLISTING"
>IF v_count > 0 THEN
INSERT INTO users_count (count) VALUES (v_count);
RETURN 't';
ELSE
RETURN 'f';
END IF;</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN56925"
>39.6.2.3. <TT
CLASS="LITERAL"
>IF-THEN-ELSIF</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>IF <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
[<SPAN
CLASS="OPTIONAL"
> ELSIF <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
[<SPAN
CLASS="OPTIONAL"
> ELSIF <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
...</SPAN
>]</SPAN
>]
[<SPAN
CLASS="OPTIONAL"
> ELSE
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> </SPAN
>]
END IF;</PRE
><P
> Sometimes there are more than just two alternatives.
<TT
CLASS="LITERAL"
>IF-THEN-ELSIF</TT
> provides a convenient
method of checking several alternatives in turn.
The <TT
CLASS="LITERAL"
>IF</TT
> conditions are tested successively
until the first one that is true is found. Then the
associated statement(s) are executed, after which control
passes to the next statement after <TT
CLASS="LITERAL"
>END IF</TT
>.
(Any subsequent <TT
CLASS="LITERAL"
>IF</TT
> conditions are <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>not</I
></SPAN
>
tested.) If none of the <TT
CLASS="LITERAL"
>IF</TT
> conditions is true,
then the <TT
CLASS="LITERAL"
>ELSE</TT
> block (if any) is executed.
</P
><P
> Here is an example:
</P><PRE
CLASS="PROGRAMLISTING"
>IF number = 0 THEN
result := 'zero';
ELSIF number > 0 THEN
result := 'positive';
ELSIF number < 0 THEN
result := 'negative';
ELSE
-- hmm, the only other possibility is that number is null
result := 'NULL';
END IF;</PRE
><P>
</P
><P
> The key word <TT
CLASS="LITERAL"
>ELSIF</TT
> can also be spelled
<TT
CLASS="LITERAL"
>ELSEIF</TT
>.
</P
><P
> An alternative way of accomplishing the same task is to nest
<TT
CLASS="LITERAL"
>IF-THEN-ELSE</TT
> statements, as in the
following example:
</P><PRE
CLASS="PROGRAMLISTING"
>IF demo_row.sex = 'm' THEN
pretty_sex := 'man';
ELSE
IF demo_row.sex = 'f' THEN
pretty_sex := 'woman';
END IF;
END IF;</PRE
><P>
</P
><P
> However, this method requires writing a matching <TT
CLASS="LITERAL"
>END IF</TT
>
for each <TT
CLASS="LITERAL"
>IF</TT
>, so it is much more cumbersome than
using <TT
CLASS="LITERAL"
>ELSIF</TT
> when there are many alternatives.
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN56959"
>39.6.2.4. Simple <TT
CLASS="LITERAL"
>CASE</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>CASE <TT
CLASS="REPLACEABLE"
><I
>search-expression</I
></TT
>
WHEN <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
>, <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> ... </SPAN
>]</SPAN
>] THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
[<SPAN
CLASS="OPTIONAL"
> WHEN <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
>, <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> ... </SPAN
>]</SPAN
>] THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
... </SPAN
>]
[<SPAN
CLASS="OPTIONAL"
> ELSE
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> </SPAN
>]
END CASE;</PRE
><P
> The simple form of <TT
CLASS="COMMAND"
>CASE</TT
> provides conditional execution
based on equality of operands. The <TT
CLASS="REPLACEABLE"
><I
>search-expression</I
></TT
>
is evaluated (once) and successively compared to each
<TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> in the <TT
CLASS="LITERAL"
>WHEN</TT
> clauses.
If a match is found, then the corresponding
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> are executed, and then control
passes to the next statement after <TT
CLASS="LITERAL"
>END CASE</TT
>. (Subsequent
<TT
CLASS="LITERAL"
>WHEN</TT
> expressions are not evaluated.) If no match is
found, the <TT
CLASS="LITERAL"
>ELSE</TT
> <TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> are
executed; but if <TT
CLASS="LITERAL"
>ELSE</TT
> is not present, then a
<TT
CLASS="LITERAL"
>CASE_NOT_FOUND</TT
> exception is raised.
</P
><P
> Here is a simple example:
</P><PRE
CLASS="PROGRAMLISTING"
>CASE x
WHEN 1, 2 THEN
msg := 'one or two';
ELSE
msg := 'other value than one or two';
END CASE;</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN56991"
>39.6.2.5. Searched <TT
CLASS="LITERAL"
>CASE</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>CASE
WHEN <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
[<SPAN
CLASS="OPTIONAL"
> WHEN <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> THEN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
... </SPAN
>]
[<SPAN
CLASS="OPTIONAL"
> ELSE
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> </SPAN
>]
END CASE;</PRE
><P
> The searched form of <TT
CLASS="COMMAND"
>CASE</TT
> provides conditional execution
based on truth of Boolean expressions. Each <TT
CLASS="LITERAL"
>WHEN</TT
> clause's
<TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> is evaluated in turn,
until one is found that yields <TT
CLASS="LITERAL"
>true</TT
>. Then the
corresponding <TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> are executed, and
then control passes to the next statement after <TT
CLASS="LITERAL"
>END CASE</TT
>.
(Subsequent <TT
CLASS="LITERAL"
>WHEN</TT
> expressions are not evaluated.)
If no true result is found, the <TT
CLASS="LITERAL"
>ELSE</TT
>
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> are executed;
but if <TT
CLASS="LITERAL"
>ELSE</TT
> is not present, then a
<TT
CLASS="LITERAL"
>CASE_NOT_FOUND</TT
> exception is raised.
</P
><P
> Here is an example:
</P><PRE
CLASS="PROGRAMLISTING"
>CASE
WHEN x BETWEEN 0 AND 10 THEN
msg := 'value is between zero and ten';
WHEN x BETWEEN 11 AND 20 THEN
msg := 'value is between eleven and twenty';
END CASE;</PRE
><P>
</P
><P
> This form of <TT
CLASS="COMMAND"
>CASE</TT
> is entirely equivalent to
<TT
CLASS="LITERAL"
>IF-THEN-ELSIF</TT
>, except for the rule that reaching
an omitted <TT
CLASS="LITERAL"
>ELSE</TT
> clause results in an error rather
than doing nothing.
</P
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="PLPGSQL-CONTROL-STRUCTURES-LOOPS"
>39.6.3. Simple Loops</A
></H2
><P
> With the <TT
CLASS="LITERAL"
>LOOP</TT
>, <TT
CLASS="LITERAL"
>EXIT</TT
>,
<TT
CLASS="LITERAL"
>CONTINUE</TT
>, <TT
CLASS="LITERAL"
>WHILE</TT
>, <TT
CLASS="LITERAL"
>FOR</TT
>,
and <TT
CLASS="LITERAL"
>FOREACH</TT
> statements, you can arrange for your
<SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> function to repeat a series of commands.
</P
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN57033"
>39.6.3.1. <TT
CLASS="LITERAL"
>LOOP</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>[<SPAN
CLASS="OPTIONAL"
> <<<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>>> </SPAN
>]
LOOP
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END LOOP [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>];</PRE
><P
> <TT
CLASS="LITERAL"
>LOOP</TT
> defines an unconditional loop that is repeated
indefinitely until terminated by an <TT
CLASS="LITERAL"
>EXIT</TT
> or
<TT
CLASS="COMMAND"
>RETURN</TT
> statement. The optional
<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> can be used by <TT
CLASS="LITERAL"
>EXIT</TT
>
and <TT
CLASS="LITERAL"
>CONTINUE</TT
> statements within nested loops to
specify which loop those statements refer to.
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN57049"
>39.6.3.2. <TT
CLASS="LITERAL"
>EXIT</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>EXIT [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>] [<SPAN
CLASS="OPTIONAL"
> WHEN <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> </SPAN
>];</PRE
><P
> If no <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> is given, the innermost
loop is terminated and the statement following <TT
CLASS="LITERAL"
>END
LOOP</TT
> is executed next. If <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>
is given, it must be the label of the current or some outer
level of nested loop or block. Then the named loop or block is
terminated and control continues with the statement after the
loop's/block's corresponding <TT
CLASS="LITERAL"
>END</TT
>.
</P
><P
> If <TT
CLASS="LITERAL"
>WHEN</TT
> is specified, the loop exit occurs only if
<TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> is true. Otherwise, control passes
to the statement after <TT
CLASS="LITERAL"
>EXIT</TT
>.
</P
><P
> <TT
CLASS="LITERAL"
>EXIT</TT
> can be used with all types of loops; it is
not limited to use with unconditional loops.
</P
><P
> When used with a
<TT
CLASS="LITERAL"
>BEGIN</TT
> block, <TT
CLASS="LITERAL"
>EXIT</TT
> passes
control to the next statement after the end of the block.
Note that a label must be used for this purpose; an unlabelled
<TT
CLASS="LITERAL"
>EXIT</TT
> is never considered to match a
<TT
CLASS="LITERAL"
>BEGIN</TT
> block. (This is a change from
pre-8.4 releases of <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>, which
would allow an unlabelled <TT
CLASS="LITERAL"
>EXIT</TT
> to match
a <TT
CLASS="LITERAL"
>BEGIN</TT
> block.)
</P
><P
> Examples:
</P><PRE
CLASS="PROGRAMLISTING"
>LOOP
-- some computations
IF count > 0 THEN
EXIT; -- exit loop
END IF;
END LOOP;
LOOP
-- some computations
EXIT WHEN count > 0; -- same result as previous example
END LOOP;
<<ablock>>
BEGIN
-- some computations
IF stocks > 100000 THEN
EXIT ablock; -- causes exit from the BEGIN block
END IF;
-- computations here will be skipped when stocks > 100000
END;</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN57081"
>39.6.3.3. <TT
CLASS="LITERAL"
>CONTINUE</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>CONTINUE [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>] [<SPAN
CLASS="OPTIONAL"
> WHEN <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> </SPAN
>];</PRE
><P
> If no <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> is given, the next iteration of
the innermost loop is begun. That is, all statements remaining
in the loop body are skipped, and control returns
to the loop control expression (if any) to determine whether
another loop iteration is needed.
If <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> is present, it
specifies the label of the loop whose execution will be
continued.
</P
><P
> If <TT
CLASS="LITERAL"
>WHEN</TT
> is specified, the next iteration of the
loop is begun only if <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> is
true. Otherwise, control passes to the statement after
<TT
CLASS="LITERAL"
>CONTINUE</TT
>.
</P
><P
> <TT
CLASS="LITERAL"
>CONTINUE</TT
> can be used with all types of loops; it
is not limited to use with unconditional loops.
</P
><P
> Examples:
</P><PRE
CLASS="PROGRAMLISTING"
>LOOP
-- some computations
EXIT WHEN count > 100;
CONTINUE WHEN count < 50;
-- some computations for count IN [50 .. 100]
END LOOP;</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN57103"
>39.6.3.4. <TT
CLASS="LITERAL"
>WHILE</TT
></A
></H3
><PRE
CLASS="SYNOPSIS"
>[<SPAN
CLASS="OPTIONAL"
> <<<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>>> </SPAN
>]
WHILE <TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
> LOOP
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END LOOP [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>];</PRE
><P
> The <TT
CLASS="LITERAL"
>WHILE</TT
> statement repeats a
sequence of statements so long as the
<TT
CLASS="REPLACEABLE"
><I
>boolean-expression</I
></TT
>
evaluates to true. The expression is checked just before
each entry to the loop body.
</P
><P
> For example:
</P><PRE
CLASS="PROGRAMLISTING"
>WHILE amount_owed > 0 AND gift_certificate_balance > 0 LOOP
-- some computations here
END LOOP;
WHILE NOT done LOOP
-- some computations here
END LOOP;</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="PLPGSQL-INTEGER-FOR"
>39.6.3.5. <TT
CLASS="LITERAL"
>FOR</TT
> (Integer Variant)</A
></H3
><PRE
CLASS="SYNOPSIS"
>[<SPAN
CLASS="OPTIONAL"
> <<<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>>> </SPAN
>]
FOR <TT
CLASS="REPLACEABLE"
><I
>name</I
></TT
> IN [<SPAN
CLASS="OPTIONAL"
> REVERSE </SPAN
>] <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> .. <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> BY <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> </SPAN
>] LOOP
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END LOOP [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>];</PRE
><P
> This form of <TT
CLASS="LITERAL"
>FOR</TT
> creates a loop that iterates over a range
of integer values. The variable
<TT
CLASS="REPLACEABLE"
><I
>name</I
></TT
> is automatically defined as type
<TT
CLASS="TYPE"
>integer</TT
> and exists only inside the loop (any existing
definition of the variable name is ignored within the loop).
The two expressions giving
the lower and upper bound of the range are evaluated once when entering
the loop. If the <TT
CLASS="LITERAL"
>BY</TT
> clause isn't specified the iteration
step is 1, otherwise it's the value specified in the <TT
CLASS="LITERAL"
>BY</TT
>
clause, which again is evaluated once on loop entry.
If <TT
CLASS="LITERAL"
>REVERSE</TT
> is specified then the step value is
subtracted, rather than added, after each iteration.
</P
><P
> Some examples of integer <TT
CLASS="LITERAL"
>FOR</TT
> loops:
</P><PRE
CLASS="PROGRAMLISTING"
>FOR i IN 1..10 LOOP
-- i will take on the values 1,2,3,4,5,6,7,8,9,10 within the loop
END LOOP;
FOR i IN REVERSE 10..1 LOOP
-- i will take on the values 10,9,8,7,6,5,4,3,2,1 within the loop
END LOOP;
FOR i IN REVERSE 10..1 BY 2 LOOP
-- i will take on the values 10,8,6,4,2 within the loop
END LOOP;</PRE
><P>
</P
><P
> If the lower bound is greater than the upper bound (or less than,
in the <TT
CLASS="LITERAL"
>REVERSE</TT
> case), the loop body is not
executed at all. No error is raised.
</P
><P
> If a <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> is attached to the
<TT
CLASS="LITERAL"
>FOR</TT
> loop then the integer loop variable can be
referenced with a qualified name, using that
<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>.
</P
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="PLPGSQL-RECORDS-ITERATING"
>39.6.4. Looping Through Query Results</A
></H2
><P
> Using a different type of <TT
CLASS="LITERAL"
>FOR</TT
> loop, you can iterate through
the results of a query and manipulate that data
accordingly. The syntax is:
</P><PRE
CLASS="SYNOPSIS"
>[<SPAN
CLASS="OPTIONAL"
> <<<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>>> </SPAN
>]
FOR <TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> IN <TT
CLASS="REPLACEABLE"
><I
>query</I
></TT
> LOOP
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END LOOP [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>];</PRE
><P>
The <TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> is a record variable, row variable,
or comma-separated list of scalar variables.
The <TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> is successively assigned each row
resulting from the <TT
CLASS="REPLACEABLE"
><I
>query</I
></TT
> and the loop body is
executed for each row. Here is an example:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE FUNCTION cs_refresh_mviews() RETURNS integer AS $$
DECLARE
mviews RECORD;
BEGIN
RAISE NOTICE 'Refreshing materialized views...';
FOR mviews IN SELECT * FROM cs_materialized_views ORDER BY sort_key LOOP
-- Now "mviews" has one record from cs_materialized_views
RAISE NOTICE 'Refreshing materialized view %s ...', quote_ident(mviews.mv_name);
EXECUTE 'TRUNCATE TABLE ' || quote_ident(mviews.mv_name);
EXECUTE 'INSERT INTO '
|| quote_ident(mviews.mv_name) || ' '
|| mviews.mv_query;
END LOOP;
RAISE NOTICE 'Done refreshing materialized views.';
RETURN 1;
END;
$$ LANGUAGE plpgsql;</PRE
><P>
If the loop is terminated by an <TT
CLASS="LITERAL"
>EXIT</TT
> statement, the last
assigned row value is still accessible after the loop.
</P
><P
> The <TT
CLASS="REPLACEABLE"
><I
>query</I
></TT
> used in this type of <TT
CLASS="LITERAL"
>FOR</TT
>
statement can be any SQL command that returns rows to the caller:
<TT
CLASS="COMMAND"
>SELECT</TT
> is the most common case,
but you can also use <TT
CLASS="COMMAND"
>INSERT</TT
>, <TT
CLASS="COMMAND"
>UPDATE</TT
>, or
<TT
CLASS="COMMAND"
>DELETE</TT
> with a <TT
CLASS="LITERAL"
>RETURNING</TT
> clause. Some utility
commands such as <TT
CLASS="COMMAND"
>EXPLAIN</TT
> will work too.
</P
><P
> <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> variables are substituted into the query text,
and the query plan is cached for possible re-use, as discussed in
detail in <A
HREF="plpgsql-implementation.html#PLPGSQL-VAR-SUBST"
>Section 39.10.1</A
> and
<A
HREF="plpgsql-implementation.html#PLPGSQL-PLAN-CACHING"
>Section 39.10.2</A
>.
</P
><P
> The <TT
CLASS="LITERAL"
>FOR-IN-EXECUTE</TT
> statement is another way to iterate over
rows:
</P><PRE
CLASS="SYNOPSIS"
>[<SPAN
CLASS="OPTIONAL"
> <<<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>>> </SPAN
>]
FOR <TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> IN EXECUTE <TT
CLASS="REPLACEABLE"
><I
>text_expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> USING <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> [<SPAN
CLASS="OPTIONAL"
>, ... </SPAN
>] </SPAN
>] LOOP
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END LOOP [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>];</PRE
><P>
This is like the previous form, except that the source query
is specified as a string expression, which is evaluated and replanned
on each entry to the <TT
CLASS="LITERAL"
>FOR</TT
> loop. This allows the programmer to
choose the speed of a preplanned query or the flexibility of a dynamic
query, just as with a plain <TT
CLASS="COMMAND"
>EXECUTE</TT
> statement.
As with <TT
CLASS="COMMAND"
>EXECUTE</TT
>, parameter values can be inserted
into the dynamic command via <TT
CLASS="LITERAL"
>USING</TT
>.
</P
><P
> Another way to specify the query whose results should be iterated
through is to declare it as a cursor. This is described in
<A
HREF="plpgsql-cursors.html#PLPGSQL-CURSOR-FOR-LOOP"
>Section 39.7.4</A
>.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="PLPGSQL-FOREACH-ARRAY"
>39.6.5. Looping Through Arrays</A
></H2
><P
> The <TT
CLASS="LITERAL"
>FOREACH</TT
> loop is much like a <TT
CLASS="LITERAL"
>FOR</TT
> loop,
but instead of iterating through the rows returned by a SQL query,
it iterates through the elements of an array value.
(In general, <TT
CLASS="LITERAL"
>FOREACH</TT
> is meant for looping through
components of a composite-valued expression; variants for looping
through composites besides arrays may be added in future.)
The <TT
CLASS="LITERAL"
>FOREACH</TT
> statement to loop over an array is:
</P><PRE
CLASS="SYNOPSIS"
>[<SPAN
CLASS="OPTIONAL"
> <<<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>>> </SPAN
>]
FOREACH <TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> SLICE <TT
CLASS="REPLACEABLE"
><I
>number</I
></TT
> </SPAN
>] IN ARRAY <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
> LOOP
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
END LOOP [<SPAN
CLASS="OPTIONAL"
> <TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
> </SPAN
>];</PRE
><P>
</P
><P
> Without <TT
CLASS="LITERAL"
>SLICE</TT
>, or if <TT
CLASS="LITERAL"
>SLICE 0</TT
> is specified,
the loop iterates through individual elements of the array produced
by evaluating the <TT
CLASS="REPLACEABLE"
><I
>expression</I
></TT
>.
The <TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> variable is assigned each
element value in sequence, and the loop body is executed for each element.
Here is an example of looping through the elements of an integer
array:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE FUNCTION sum(int[]) RETURNS int8 AS $$
DECLARE
s int8 := 0;
x int;
BEGIN
FOREACH x IN ARRAY $1
LOOP
s := s + x;
END LOOP;
RETURN s;
END;
$$ LANGUAGE plpgsql;</PRE
><P>
The elements are visited in storage order, regardless of the number of
array dimensions. Although the <TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> is
usually just a single variable, it can be a list of variables when
looping through an array of composite values (records). In that case,
for each array element, the variables are assigned from successive
columns of the composite value.
</P
><P
> With a positive <TT
CLASS="LITERAL"
>SLICE</TT
> value, <TT
CLASS="LITERAL"
>FOREACH</TT
>
iterates through slices of the array rather than single elements.
The <TT
CLASS="LITERAL"
>SLICE</TT
> value must be an integer constant not larger
than the number of dimensions of the array. The
<TT
CLASS="REPLACEABLE"
><I
>target</I
></TT
> variable must be an array,
and it receives successive slices of the array value, where each slice
is of the number of dimensions specified by <TT
CLASS="LITERAL"
>SLICE</TT
>.
Here is an example of iterating through one-dimensional slices:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE FUNCTION scan_rows(int[]) RETURNS void AS $$
DECLARE
x int[];
BEGIN
FOREACH x SLICE 1 IN ARRAY $1
LOOP
RAISE NOTICE 'row = %', x;
END LOOP;
END;
$$ LANGUAGE plpgsql;
SELECT scan_rows(ARRAY[[1,2,3],[4,5,6],[7,8,9],[10,11,12]]);
NOTICE: row = {1,2,3}
NOTICE: row = {4,5,6}
NOTICE: row = {7,8,9}
NOTICE: row = {10,11,12}</PRE
><P>
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="PLPGSQL-ERROR-TRAPPING"
>39.6.6. Trapping Errors</A
></H2
><P
> By default, any error occurring in a <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
>
function aborts execution of the function, and indeed of the
surrounding transaction as well. You can trap errors and recover
from them by using a <TT
CLASS="COMMAND"
>BEGIN</TT
> block with an
<TT
CLASS="LITERAL"
>EXCEPTION</TT
> clause. The syntax is an extension of the
normal syntax for a <TT
CLASS="COMMAND"
>BEGIN</TT
> block:
</P><PRE
CLASS="SYNOPSIS"
>[<SPAN
CLASS="OPTIONAL"
> <<<TT
CLASS="REPLACEABLE"
><I
>label</I
></TT
>>> </SPAN
>]
[<SPAN
CLASS="OPTIONAL"
> DECLARE
<TT
CLASS="REPLACEABLE"
><I
>declarations</I
></TT
> </SPAN
>]
BEGIN
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>
EXCEPTION
WHEN <TT
CLASS="REPLACEABLE"
><I
>condition</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> OR <TT
CLASS="REPLACEABLE"
><I
>condition</I
></TT
> ... </SPAN
>] THEN
<TT
CLASS="REPLACEABLE"
><I
>handler_statements</I
></TT
>
[<SPAN
CLASS="OPTIONAL"
> WHEN <TT
CLASS="REPLACEABLE"
><I
>condition</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> OR <TT
CLASS="REPLACEABLE"
><I
>condition</I
></TT
> ... </SPAN
>] THEN
<TT
CLASS="REPLACEABLE"
><I
>handler_statements</I
></TT
>
... </SPAN
>]
END;</PRE
><P>
</P
><P
> If no error occurs, this form of block simply executes all the
<TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>, and then control passes
to the next statement after <TT
CLASS="LITERAL"
>END</TT
>. But if an error
occurs within the <TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
>, further
processing of the <TT
CLASS="REPLACEABLE"
><I
>statements</I
></TT
> is
abandoned, and control passes to the <TT
CLASS="LITERAL"
>EXCEPTION</TT
> list.
The list is searched for the first <TT
CLASS="REPLACEABLE"
><I
>condition</I
></TT
>
matching the error that occurred. If a match is found, the
corresponding <TT
CLASS="REPLACEABLE"
><I
>handler_statements</I
></TT
> are
executed, and then control passes to the next statement after
<TT
CLASS="LITERAL"
>END</TT
>. If no match is found, the error propagates out
as though the <TT
CLASS="LITERAL"
>EXCEPTION</TT
> clause were not there at all:
the error can be caught by an enclosing block with
<TT
CLASS="LITERAL"
>EXCEPTION</TT
>, or if there is none it aborts processing
of the function.
</P
><P
> The <TT
CLASS="REPLACEABLE"
><I
>condition</I
></TT
> names can be any of
those shown in <A
HREF="errcodes-appendix.html"
>Appendix A</A
>. A category
name matches any error within its category. The special
condition name <TT
CLASS="LITERAL"
>OTHERS</TT
> matches every error type except
<TT
CLASS="LITERAL"
>QUERY_CANCELED</TT
>. (It is possible, but often unwise,
to trap <TT
CLASS="LITERAL"
>QUERY_CANCELED</TT
> by name.) Condition names are
not case-sensitive. Also, an error condition can be specified
by <TT
CLASS="LITERAL"
>SQLSTATE</TT
> code; for example these are equivalent:
</P><PRE
CLASS="PROGRAMLISTING"
>WHEN division_by_zero THEN ...
WHEN SQLSTATE '22012' THEN ...</PRE
><P>
</P
><P
> If a new error occurs within the selected
<TT
CLASS="REPLACEABLE"
><I
>handler_statements</I
></TT
>, it cannot be caught
by this <TT
CLASS="LITERAL"
>EXCEPTION</TT
> clause, but is propagated out.
A surrounding <TT
CLASS="LITERAL"
>EXCEPTION</TT
> clause could catch it.
</P
><P
> When an error is caught by an <TT
CLASS="LITERAL"
>EXCEPTION</TT
> clause,
the local variables of the <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
> function
remain as they were when the error occurred, but all changes
to persistent database state within the block are rolled back.
As an example, consider this fragment:
</P><PRE
CLASS="PROGRAMLISTING"
>INSERT INTO mytab(firstname, lastname) VALUES('Tom', 'Jones');
BEGIN
UPDATE mytab SET firstname = 'Joe' WHERE lastname = 'Jones';
x := x + 1;
y := x / 0;
EXCEPTION
WHEN division_by_zero THEN
RAISE NOTICE 'caught division_by_zero';
RETURN x;
END;</PRE
><P>
When control reaches the assignment to <TT
CLASS="LITERAL"
>y</TT
>, it will
fail with a <TT
CLASS="LITERAL"
>division_by_zero</TT
> error. This will be caught by
the <TT
CLASS="LITERAL"
>EXCEPTION</TT
> clause. The value returned in the
<TT
CLASS="COMMAND"
>RETURN</TT
> statement will be the incremented value of
<TT
CLASS="LITERAL"
>x</TT
>, but the effects of the <TT
CLASS="COMMAND"
>UPDATE</TT
> command will
have been rolled back. The <TT
CLASS="COMMAND"
>INSERT</TT
> command preceding the
block is not rolled back, however, so the end result is that the database
contains <TT
CLASS="LITERAL"
>Tom Jones</TT
> not <TT
CLASS="LITERAL"
>Joe Jones</TT
>.
</P
><DIV
CLASS="TIP"
><BLOCKQUOTE
CLASS="TIP"
><P
><B
>Tip: </B
> A block containing an <TT
CLASS="LITERAL"
>EXCEPTION</TT
> clause is significantly
more expensive to enter and exit than a block without one. Therefore,
don't use <TT
CLASS="LITERAL"
>EXCEPTION</TT
> without need.
</P
></BLOCKQUOTE
></DIV
><DIV
CLASS="EXAMPLE"
><A
NAME="PLPGSQL-UPSERT-EXAMPLE"
></A
><P
><B
>Example 39-2. Exceptions with <TT
CLASS="COMMAND"
>UPDATE</TT
>/<TT
CLASS="COMMAND"
>INSERT</TT
></B
></P
><P
> This example uses exception handling to perform either
<TT
CLASS="COMMAND"
>UPDATE</TT
> or <TT
CLASS="COMMAND"
>INSERT</TT
>, as appropriate:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE TABLE db (a INT PRIMARY KEY, b TEXT);
CREATE FUNCTION merge_db(key INT, data TEXT) RETURNS VOID AS
$$
BEGIN
LOOP
-- first try to update the key
UPDATE db SET b = data WHERE a = key;
IF found THEN
RETURN;
END IF;
-- not there, so try to insert the key
-- if someone else inserts the same key concurrently,
-- we could get a unique-key failure
BEGIN
INSERT INTO db(a,b) VALUES (key, data);
RETURN;
EXCEPTION WHEN unique_violation THEN
-- Do nothing, and loop to try the UPDATE again.
END;
END LOOP;
END;
$$
LANGUAGE plpgsql;
SELECT merge_db(1, 'david');
SELECT merge_db(1, 'dennis');</PRE
><P>
This coding assumes the <TT
CLASS="LITERAL"
>unique_violation</TT
> error is caused by
the <TT
CLASS="COMMAND"
>INSERT</TT
>, and not by, say, an <TT
CLASS="COMMAND"
>INSERT</TT
> in a
trigger function on the table. It might also misbehave if there is
more than one unique index on the table, since it will retry the
operation regardless of which index caused the error.
More safety could be had by using the
features discussed next to check that the trapped error was the one
expected.
</P
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="PLPGSQL-EXCEPTION-DIAGNOSTICS"
>39.6.6.1. Obtaining information about an error</A
></H3
><P
> Exception handlers frequently need to identify the specific error that
occurred. There are two ways to get information about the current
exception in <SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
>: special variables and the
<TT
CLASS="COMMAND"
>GET STACKED DIAGNOSTICS</TT
> command.
</P
><P
> Within an exception handler, the special variable
<TT
CLASS="VARNAME"
>SQLSTATE</TT
> contains the error code that corresponds to
the exception that was raised (refer to <A
HREF="errcodes-appendix.html#ERRCODES-TABLE"
>Table A-1</A
>
for a list of possible error codes). The special variable
<TT
CLASS="VARNAME"
>SQLERRM</TT
> contains the error message associated with the
exception. These variables are undefined outside exception handlers.
</P
><P
> Within an exception handler, one may also retrieve
information about the current exception by using the
<TT
CLASS="COMMAND"
>GET STACKED DIAGNOSTICS</TT
> command, which has the form:
</P><PRE
CLASS="SYNOPSIS"
>GET STACKED DIAGNOSTICS <TT
CLASS="REPLACEABLE"
><I
>variable</I
></TT
> = <TT
CLASS="REPLACEABLE"
><I
>item</I
></TT
> [<SPAN
CLASS="OPTIONAL"
> , ... </SPAN
>];</PRE
><P>
Each <TT
CLASS="REPLACEABLE"
><I
>item</I
></TT
> is a key word identifying a status
value to be assigned to the specified variable (which should be
of the right data type to receive it). The currently available
status items are shown in <A
HREF="plpgsql-control-structures.html#PLPGSQL-EXCEPTION-DIAGNOSTICS-VALUES"
>Table 39-1</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="PLPGSQL-EXCEPTION-DIAGNOSTICS-VALUES"
></A
><P
><B
>Table 39-1. Error diagnostics values</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><COL><THEAD
><TR
><TH
>Name</TH
><TH
>Type</TH
><TH
>Description</TH
></TR
></THEAD
><TBODY
><TR
><TD
><TT
CLASS="LITERAL"
>RETURNED_SQLSTATE</TT
></TD
><TD
>text</TD
><TD
>the SQLSTATE error code of the exception</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>MESSAGE_TEXT</TT
></TD
><TD
>text</TD
><TD
>the text of the exception's primary message</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>PG_EXCEPTION_DETAIL</TT
></TD
><TD
>text</TD
><TD
>the text of the exception's detail message, if any</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>PG_EXCEPTION_HINT</TT
></TD
><TD
>text</TD
><TD
>the text of the exception's hint message, if any</TD
></TR
><TR
><TD
><TT
CLASS="LITERAL"
>PG_EXCEPTION_CONTEXT</TT
></TD
><TD
>text</TD
><TD
>line(s) of text describing the call stack</TD
></TR
></TBODY
></TABLE
></DIV
><P
> If the exception did not set a value for an item, an empty string
will be returned.
</P
><P
> Here is an example:
</P><PRE
CLASS="PROGRAMLISTING"
>DECLARE
text_var1 text;
text_var2 text;
text_var3 text;
BEGIN
-- some processing which might cause an exception
...
EXCEPTION WHEN OTHERS THEN
GET STACKED DIAGNOSTICS text_var1 = MESSAGE_TEXT,
text_var2 = PG_EXCEPTION_DETAIL,
text_var3 = PG_EXCEPTION_HINT;
END;</PRE
><P>
</P
></DIV
></DIV
></DIV
><DIV
CLASS="NAVFOOTER"
><HR
ALIGN="LEFT"
WIDTH="100%"><TABLE
SUMMARY="Footer navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
><A
HREF="plpgsql-statements.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="index.html"
ACCESSKEY="H"
>Home</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
><A
HREF="plpgsql-cursors.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
>Basic Statements</TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="plpgsql.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
>Cursors</TD
></TR
></TABLE
></DIV
></BODY
></HTML
>