<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>39.1. Overview of Trigger Behavior</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><link rel="prev" href="triggers.html" title="Chapter 39. Triggers" /><link rel="next" href="trigger-datachanges.html" title="39.2. Visibility of Data Changes" /></head><body><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">39.1. Overview of Trigger Behavior</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="triggers.html" title="Chapter 39. Triggers">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="triggers.html" title="Chapter 39. Triggers">Up</a></td><th width="60%" align="center">Chapter 39. Triggers</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 11.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="trigger-datachanges.html" title="39.2. Visibility of Data Changes">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="TRIGGER-DEFINITION"><div class="titlepage"><div><div><h2 class="title" style="clear: both">39.1. Overview of Trigger Behavior</h2></div></div></div><p>
A trigger is a specification that the database should automatically
execute a particular function whenever a certain type of operation is
performed. Triggers can be attached to tables (partitioned or not),
views, and foreign tables.
</p><p>
On tables and foreign tables, triggers can be defined to execute either
before or after any <code class="command">INSERT</code>, <code class="command">UPDATE</code>,
or <code class="command">DELETE</code> operation, either once per modified row,
or once per <acronym class="acronym">SQL</acronym> statement.
<code class="command">UPDATE</code> triggers can moreover be set to fire only if
certain columns are mentioned in the <code class="literal">SET</code> clause of
the <code class="command">UPDATE</code> statement. Triggers can also fire
for <code class="command">TRUNCATE</code> statements. If a trigger event occurs,
the trigger's function is called at the appropriate time to handle the
event.
</p><p>
On views, triggers can be defined to execute instead of
<code class="command">INSERT</code>, <code class="command">UPDATE</code>, or
<code class="command">DELETE</code> operations.
Such <code class="literal">INSTEAD OF</code> triggers
are fired once for each row that needs to be modified in the view.
It is the responsibility of the
trigger's function to perform the necessary modifications to the view's
underlying base table(s) and, where appropriate, return the modified
row as it will appear in the view. Triggers on views can also be defined
to execute once per <acronym class="acronym">SQL</acronym> statement, before or after
<code class="command">INSERT</code>, <code class="command">UPDATE</code>, or
<code class="command">DELETE</code> operations.
However, such triggers are fired only if there is also
an <code class="literal">INSTEAD OF</code> trigger on the view. Otherwise,
any statement targeting the view must be rewritten into a statement
affecting its underlying base table(s), and then the triggers
that will be fired are the ones attached to the base table(s).
</p><p>
The trigger function must be defined before the trigger itself can be
created. The trigger function must be declared as a
function taking no arguments and returning type <code class="literal">trigger</code>.
(The trigger function receives its input through a specially-passed
<code class="structname">TriggerData</code> structure, not in the form of ordinary function
arguments.)
</p><p>
Once a suitable trigger function has been created, the trigger is
established with
<a class="xref" href="sql-createtrigger.html" title="CREATE TRIGGER"><span class="refentrytitle">CREATE TRIGGER</span></a>.
The same trigger function can be used for multiple triggers.
</p><p>
<span class="productname">PostgreSQL</span> offers both <em class="firstterm">per-row</em>
triggers and <em class="firstterm">per-statement</em> triggers. With a per-row
trigger, the trigger function
is invoked once for each row that is affected by the statement
that fired the trigger. In contrast, a per-statement trigger is
invoked only once when an appropriate statement is executed,
regardless of the number of rows affected by that statement. In
particular, a statement that affects zero rows will still result
in the execution of any applicable per-statement triggers. These
two types of triggers are sometimes called <em class="firstterm">row-level</em>
triggers and <em class="firstterm">statement-level</em> triggers,
respectively. Triggers on <code class="command">TRUNCATE</code> may only be
defined at statement level, not per-row.
</p><p>
Triggers are also classified according to whether they fire
<em class="firstterm">before</em>, <em class="firstterm">after</em>, or
<em class="firstterm">instead of</em> the operation. These are referred to
as <code class="literal">BEFORE</code> triggers, <code class="literal">AFTER</code> triggers, and
<code class="literal">INSTEAD OF</code> triggers respectively.
Statement-level <code class="literal">BEFORE</code> triggers naturally fire before the
statement starts to do anything, while statement-level <code class="literal">AFTER</code>
triggers fire at the very end of the statement. These types of
triggers may be defined on tables, views, or foreign tables. Row-level
<code class="literal">BEFORE</code> triggers fire immediately before a particular row is
operated on, while row-level <code class="literal">AFTER</code> triggers fire at the end of
the statement (but before any statement-level <code class="literal">AFTER</code> triggers).
These types of triggers may only be defined on non-partitioned tables and
foreign tables, not views. <code class="literal">INSTEAD OF</code> triggers may only be
defined on views, and only at row level; they fire immediately as each
row in the view is identified as needing to be operated on.
</p><p>
A statement that targets a parent table in an inheritance or partitioning
hierarchy does not cause the statement-level triggers of affected child
tables to be fired; only the parent table's statement-level triggers are
fired. However, row-level triggers of any affected child tables will be
fired.
</p><p>
If an <code class="command">INSERT</code> contains an <code class="literal">ON CONFLICT
DO UPDATE</code> clause, it is possible that the effects of
row-level <code class="literal">BEFORE</code> <code class="command">INSERT</code> triggers and
row-level <code class="literal">BEFORE</code> <code class="command">UPDATE</code> triggers can
both be applied in a way that is apparent from the final state of
the updated row, if an <code class="varname">EXCLUDED</code> column is referenced.
There need not be an <code class="varname">EXCLUDED</code> column reference for
both sets of row-level <code class="literal">BEFORE</code> triggers to execute,
though. The
possibility of surprising outcomes should be considered when there
are both <code class="literal">BEFORE</code> <code class="command">INSERT</code> and
<code class="literal">BEFORE</code> <code class="command">UPDATE</code> row-level triggers
that change a row being inserted/updated (this can be
problematic even if the modifications are more or less equivalent, if
they're not also idempotent). Note that statement-level
<code class="command">UPDATE</code> triggers are executed when <code class="literal">ON
CONFLICT DO UPDATE</code> is specified, regardless of whether or not
any rows were affected by the <code class="command">UPDATE</code> (and
regardless of whether the alternative <code class="command">UPDATE</code>
path was ever taken). An <code class="command">INSERT</code> with an
<code class="literal">ON CONFLICT DO UPDATE</code> clause will execute
statement-level <code class="literal">BEFORE</code> <code class="command">INSERT</code>
triggers first, then statement-level <code class="literal">BEFORE</code>
<code class="command">UPDATE</code> triggers, followed by statement-level
<code class="literal">AFTER</code> <code class="command">UPDATE</code> triggers and finally
statement-level <code class="literal">AFTER</code> <code class="command">INSERT</code>
triggers.
</p><p>
If an <code class="command">UPDATE</code> on a partitioned table causes a row to move
to another partition, it will be performed as a <code class="command">DELETE</code>
from the original partition followed by an <code class="command">INSERT</code> into
the new partition. In this case, all row-level <code class="literal">BEFORE</code>
<code class="command">UPDATE</code> triggers and all row-level
<code class="literal">BEFORE</code> <code class="command">DELETE</code> triggers are fired on
the original partition. Then all row-level <code class="literal">BEFORE</code>
<code class="command">INSERT</code> triggers are fired on the destination partition.
The possibility of surprising outcomes should be considered when all these
triggers affect the row being moved. As far as <code class="literal">AFTER ROW</code>
triggers are concerned, <code class="literal">AFTER</code> <code class="command">DELETE</code>
and <code class="literal">AFTER</code> <code class="command">INSERT</code> triggers are
applied; but <code class="literal">AFTER</code> <code class="command">UPDATE</code> triggers
are not applied because the <code class="command">UPDATE</code> has been converted to
a <code class="command">DELETE</code> and an <code class="command">INSERT</code>. As far as
statement-level triggers are concerned, none of the
<code class="command">DELETE</code> or <code class="command">INSERT</code> triggers are fired,
even if row movement occurs; only the <code class="command">UPDATE</code> triggers
defined on the target table used in the <code class="command">UPDATE</code> statement
will be fired.
</p><p>
Trigger functions invoked by per-statement triggers should always
return <code class="symbol">NULL</code>. Trigger functions invoked by per-row
triggers can return a table row (a value of
type <code class="structname">HeapTuple</code>) to the calling executor,
if they choose. A row-level trigger fired before an operation has
the following choices:
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
It can return <code class="symbol">NULL</code> to skip the operation for the
current row. This instructs the executor to not perform the
row-level operation that invoked the trigger (the insertion,
modification, or deletion of a particular table row).
</p></li><li class="listitem"><p>
For row-level <code class="command">INSERT</code>
and <code class="command">UPDATE</code> triggers only, the returned row
becomes the row that will be inserted or will replace the row
being updated. This allows the trigger function to modify the
row being inserted or updated.
</p></li></ul></div><p>
A row-level <code class="literal">BEFORE</code> trigger that does not intend to cause
either of these behaviors must be careful to return as its result the same
row that was passed in (that is, the <code class="varname">NEW</code> row
for <code class="command">INSERT</code> and <code class="command">UPDATE</code>
triggers, the <code class="varname">OLD</code> row for
<code class="command">DELETE</code> triggers).
</p><p>
A row-level <code class="literal">INSTEAD OF</code> trigger should either return
<code class="symbol">NULL</code> to indicate that it did not modify any data from
the view's underlying base tables, or it should return the view
row that was passed in (the <code class="varname">NEW</code> row
for <code class="command">INSERT</code> and <code class="command">UPDATE</code>
operations, or the <code class="varname">OLD</code> row for
<code class="command">DELETE</code> operations). A nonnull return value is
used to signal that the trigger performed the necessary data
modifications in the view. This will cause the count of the number
of rows affected by the command to be incremented. For
<code class="command">INSERT</code> and <code class="command">UPDATE</code> operations, the trigger
may modify the <code class="varname">NEW</code> row before returning it. This will
change the data returned by
<code class="command">INSERT RETURNING</code> or <code class="command">UPDATE RETURNING</code>,
and is useful when the view will not show exactly the same data
that was provided.
</p><p>
The return value is ignored for row-level triggers fired after an
operation, and so they can return <code class="symbol">NULL</code>.
</p><p>
If more than one trigger is defined for the same event on the same
relation, the triggers will be fired in alphabetical order by
trigger name. In the case of <code class="literal">BEFORE</code> and
<code class="literal">INSTEAD OF</code> triggers, the possibly-modified row returned by
each trigger becomes the input to the next trigger. If any
<code class="literal">BEFORE</code> or <code class="literal">INSTEAD OF</code> trigger returns
<code class="symbol">NULL</code>, the operation is abandoned for that row and subsequent
triggers are not fired (for that row).
</p><p>
A trigger definition can also specify a Boolean <code class="literal">WHEN</code>
condition, which will be tested to see whether the trigger should
be fired. In row-level triggers the <code class="literal">WHEN</code> condition can
examine the old and/or new values of columns of the row. (Statement-level
triggers can also have <code class="literal">WHEN</code> conditions, although the feature
is not so useful for them.) In a <code class="literal">BEFORE</code> trigger, the
<code class="literal">WHEN</code>
condition is evaluated just before the function is or would be executed,
so using <code class="literal">WHEN</code> is not materially different from testing the
same condition at the beginning of the trigger function. However, in
an <code class="literal">AFTER</code> trigger, the <code class="literal">WHEN</code> condition is evaluated
just after the row update occurs, and it determines whether an event is
queued to fire the trigger at the end of statement. So when an
<code class="literal">AFTER</code> trigger's
<code class="literal">WHEN</code> condition does not return true, it is not necessary
to queue an event nor to re-fetch the row at end of statement. This
can result in significant speedups in statements that modify many
rows, if the trigger only needs to be fired for a few of the rows.
<code class="literal">INSTEAD OF</code> triggers do not support
<code class="literal">WHEN</code> conditions.
</p><p>
Typically, row-level <code class="literal">BEFORE</code> triggers are used for checking or
modifying the data that will be inserted or updated. For example,
a <code class="literal">BEFORE</code> trigger might be used to insert the current time into a
<code class="type">timestamp</code> column, or to check that two elements of the row are
consistent. Row-level <code class="literal">AFTER</code> triggers are most sensibly
used to propagate the updates to other tables, or make consistency
checks against other tables. The reason for this division of labor is
that an <code class="literal">AFTER</code> trigger can be certain it is seeing the final
value of the row, while a <code class="literal">BEFORE</code> trigger cannot; there might
be other <code class="literal">BEFORE</code> triggers firing after it. If you have no
specific reason to make a trigger <code class="literal">BEFORE</code> or
<code class="literal">AFTER</code>, the <code class="literal">BEFORE</code> case is more efficient, since
the information about
the operation doesn't have to be saved until end of statement.
</p><p>
If a trigger function executes SQL commands then these
commands might fire triggers again. This is known as cascading
triggers. There is no direct limitation on the number of cascade
levels. It is possible for cascades to cause a recursive invocation
of the same trigger; for example, an <code class="command">INSERT</code>
trigger might execute a command that inserts an additional row
into the same table, causing the <code class="command">INSERT</code> trigger
to be fired again. It is the trigger programmer's responsibility
to avoid infinite recursion in such scenarios.
</p><p>
<a id="id-1.8.4.5.19.1" class="indexterm"></a>
When a trigger is being defined, arguments can be specified for
it. The purpose of including arguments in the
trigger definition is to allow different triggers with similar
requirements to call the same function. As an example, there
could be a generalized trigger function that takes as its
arguments two column names and puts the current user in one and
the current time stamp in the other. Properly written, this
trigger function would be independent of the specific table it is
triggering on. So the same function could be used for
<code class="command">INSERT</code> events on any table with suitable
columns, to automatically track creation of records in a
transaction table for example. It could also be used to track
last-update events if defined as an <code class="command">UPDATE</code>
trigger.
</p><p>
Each programming language that supports triggers has its own method
for making the trigger input data available to the trigger function.
This input data includes the type of trigger event (e.g.,
<code class="command">INSERT</code> or <code class="command">UPDATE</code>) as well as any
arguments that were listed in <code class="command">CREATE TRIGGER</code>.
For a row-level trigger, the input data also includes the
<code class="varname">NEW</code> row for <code class="command">INSERT</code> and
<code class="command">UPDATE</code> triggers, and/or the <code class="varname">OLD</code> row
for <code class="command">UPDATE</code> and <code class="command">DELETE</code> triggers.
</p><p>
By default, statement-level triggers do not have any way to examine the
individual row(s) modified by the statement. But an <code class="literal">AFTER
STATEMENT</code> trigger can request that <em class="firstterm">transition tables</em>
be created to make the sets of affected rows available to the trigger.
<code class="literal">AFTER ROW</code> triggers can also request transition tables, so
that they can see the total changes in the table as well as the change in
the individual row they are currently being fired for. The method for
examining the transition tables again depends on the programming language
that is being used, but the typical approach is to make the transition
tables act like read-only temporary tables that can be accessed by SQL
commands issued within the trigger function.
</p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="triggers.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="triggers.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="trigger-datachanges.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">Chapter 39. Triggers </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> 39.2. Visibility of Data Changes</td></tr></table></div></body></html>
