Current File : //opt/alt/postgresql11/usr/share/doc/alt-postgresql11-9.2.24/html/xfunc-volatility.html
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<HTML
><HEAD
><TITLE
>Function Volatility Categories</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="Extending SQL"
HREF="extend.html"><LINK
REL="PREVIOUS"
TITLE="Function Overloading"
HREF="xfunc-overload.html"><LINK
REL="NEXT"
TITLE="Procedural Language Functions"
HREF="xfunc-pl.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="Function Overloading"
HREF="xfunc-overload.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
HREF="extend.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="60%"
ALIGN="center"
VALIGN="bottom"
>Chapter 35. Extending <ACRONYM
CLASS="ACRONYM"
>SQL</ACRONYM
></TD
><TD
WIDTH="20%"
ALIGN="right"
VALIGN="top"
><A
TITLE="Procedural Language Functions"
HREF="xfunc-pl.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
></TABLE
><HR
ALIGN="LEFT"
WIDTH="100%"></DIV
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="XFUNC-VOLATILITY"
>35.6. Function Volatility Categories</A
></H1
><P
> Every function has a <I
CLASS="FIRSTTERM"
>volatility</I
> classification, with
the possibilities being <TT
CLASS="LITERAL"
>VOLATILE</TT
>, <TT
CLASS="LITERAL"
>STABLE</TT
>, or
<TT
CLASS="LITERAL"
>IMMUTABLE</TT
>. <TT
CLASS="LITERAL"
>VOLATILE</TT
> is the default if the
<A
HREF="sql-createfunction.html"
>CREATE FUNCTION</A
>
command does not specify a category. The volatility category is a
promise to the optimizer about the behavior of the function:
<P
></P
></P><UL
><LI
><P
> A <TT
CLASS="LITERAL"
>VOLATILE</TT
> function can do anything, including modifying
the database. It can return different results on successive calls with
the same arguments. The optimizer makes no assumptions about the
behavior of such functions. A query using a volatile function will
re-evaluate the function at every row where its value is needed.
</P
></LI
><LI
><P
> A <TT
CLASS="LITERAL"
>STABLE</TT
> function cannot modify the database and is
guaranteed to return the same results given the same arguments
for all rows within a single statement. This category allows the
optimizer to optimize multiple calls of the function to a single
call. In particular, it is safe to use an expression containing
such a function in an index scan condition. (Since an index scan
will evaluate the comparison value only once, not once at each
row, it is not valid to use a <TT
CLASS="LITERAL"
>VOLATILE</TT
> function in an
index scan condition.)
</P
></LI
><LI
><P
> An <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> function cannot modify the database and is
guaranteed to return the same results given the same arguments forever.
This category allows the optimizer to pre-evaluate the function when
a query calls it with constant arguments. For example, a query like
<TT
CLASS="LITERAL"
>SELECT ... WHERE x = 2 + 2</TT
> can be simplified on sight to
<TT
CLASS="LITERAL"
>SELECT ... WHERE x = 4</TT
>, because the function underlying
the integer addition operator is marked <TT
CLASS="LITERAL"
>IMMUTABLE</TT
>.
</P
></LI
></UL
><P>
</P
><P
> For best optimization results, you should label your functions with the
strictest volatility category that is valid for them.
</P
><P
> Any function with side-effects <SPAN
CLASS="emphasis"
><I
CLASS="EMPHASIS"
>must</I
></SPAN
> be labeled
<TT
CLASS="LITERAL"
>VOLATILE</TT
>, so that calls to it cannot be optimized away.
Even a function with no side-effects needs to be labeled
<TT
CLASS="LITERAL"
>VOLATILE</TT
> if its value can change within a single query;
some examples are <TT
CLASS="LITERAL"
>random()</TT
>, <TT
CLASS="LITERAL"
>currval()</TT
>,
<TT
CLASS="LITERAL"
>timeofday()</TT
>.
</P
><P
> Another important example is that the <CODE
CLASS="FUNCTION"
>current_timestamp</CODE
>
family of functions qualify as <TT
CLASS="LITERAL"
>STABLE</TT
>, since their values do
not change within a transaction.
</P
><P
> There is relatively little difference between <TT
CLASS="LITERAL"
>STABLE</TT
> and
<TT
CLASS="LITERAL"
>IMMUTABLE</TT
> categories when considering simple interactive
queries that are planned and immediately executed: it doesn't matter
a lot whether a function is executed once during planning or once during
query execution startup. But there is a big difference if the plan is
saved and reused later. Labeling a function <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> when
it really isn't might allow it to be prematurely folded to a constant during
planning, resulting in a stale value being re-used during subsequent uses
of the plan. This is a hazard when using prepared statements or when
using function languages that cache plans (such as
<SPAN
CLASS="APPLICATION"
>PL/pgSQL</SPAN
>).
</P
><P
> For functions written in SQL or in any of the standard procedural
languages, there is a second important property determined by the
volatility category, namely the visibility of any data changes that have
been made by the SQL command that is calling the function. A
<TT
CLASS="LITERAL"
>VOLATILE</TT
> function will see such changes, a <TT
CLASS="LITERAL"
>STABLE</TT
>
or <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> function will not. This behavior is implemented
using the snapshotting behavior of MVCC (see <A
HREF="mvcc.html"
>Chapter 13</A
>):
<TT
CLASS="LITERAL"
>STABLE</TT
> and <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> functions use a snapshot
established as of the start of the calling query, whereas
<TT
CLASS="LITERAL"
>VOLATILE</TT
> functions obtain a fresh snapshot at the start of
each query they execute.
</P
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> Functions written in C can manage snapshots however they want, but it's
usually a good idea to make C functions work this way too.
</P
></BLOCKQUOTE
></DIV
><P
> Because of this snapshotting behavior,
a function containing only <TT
CLASS="COMMAND"
>SELECT</TT
> commands can safely be
marked <TT
CLASS="LITERAL"
>STABLE</TT
>, even if it selects from tables that might be
undergoing modifications by concurrent queries.
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> will execute all commands of a
<TT
CLASS="LITERAL"
>STABLE</TT
> function using the snapshot established for the
calling query, and so it will see a fixed view of the database throughout
that query.
</P
><P
> The same snapshotting behavior is used for <TT
CLASS="COMMAND"
>SELECT</TT
> commands
within <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> functions. It is generally unwise to select
from database tables within an <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> function at all,
since the immutability will be broken if the table contents ever change.
However, <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> does not enforce that you
do not do that.
</P
><P
> A common error is to label a function <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> when its
results depend on a configuration parameter. For example, a function
that manipulates timestamps might well have results that depend on the
<A
HREF="runtime-config-client.html#GUC-TIMEZONE"
>TimeZone</A
> setting. For safety, such functions should
be labeled <TT
CLASS="LITERAL"
>STABLE</TT
> instead.
</P
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> Before <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> release 8.0, the requirement
that <TT
CLASS="LITERAL"
>STABLE</TT
> and <TT
CLASS="LITERAL"
>IMMUTABLE</TT
> functions cannot modify
the database was not enforced by the system. Releases 8.0 and later enforce it
by requiring SQL functions and procedural language functions of these
categories to contain no SQL commands other than <TT
CLASS="COMMAND"
>SELECT</TT
>.
(This is not a completely bulletproof test, since such functions could
still call <TT
CLASS="LITERAL"
>VOLATILE</TT
> functions that modify the database.
If you do that, you will find that the <TT
CLASS="LITERAL"
>STABLE</TT
> or
<TT
CLASS="LITERAL"
>IMMUTABLE</TT
> function does not notice the database changes
applied by the called function, since they are hidden from its snapshot.)
</P
></BLOCKQUOTE
></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="xfunc-overload.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="xfunc-pl.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
>Function Overloading</TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="extend.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
>Procedural Language Functions</TD
></TR
></TABLE
></DIV
></BODY
></HTML
>
Mini Shell