<?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>SELECT</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="sql-security-label.html" title="SECURITY LABEL" /><link rel="next" href="sql-selectinto.html" title="SELECT INTO" /></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">SELECT</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="sql-security-label.html" title="SECURITY LABEL">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="sql-commands.html" title="SQL Commands">Up</a></td><th width="60%" align="center">SQL Commands</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="sql-selectinto.html" title="SELECT INTO">Next</a></td></tr></table><hr></hr></div><div class="refentry" id="SQL-SELECT"><div class="titlepage"></div><a id="id-1.9.3.171.1" class="indexterm"></a><a id="id-1.9.3.171.2" class="indexterm"></a><a id="id-1.9.3.171.3" class="indexterm"></a><div class="refnamediv"><h2><span class="refentrytitle">SELECT</span></h2><p>SELECT, TABLE, WITH — retrieve rows from a table or view</p></div><div class="refsynopsisdiv"><h2>Synopsis</h2><pre class="synopsis">
[ WITH [ RECURSIVE ] <em class="replaceable"><code>with_query</code></em> [, ...] ]
SELECT [ ALL | DISTINCT [ ON ( <em class="replaceable"><code>expression</code></em> [, ...] ) ] ]
[ * | <em class="replaceable"><code>expression</code></em> [ [ AS ] <em class="replaceable"><code>output_name</code></em> ] [, ...] ]
[ FROM <em class="replaceable"><code>from_item</code></em> [, ...] ]
[ WHERE <em class="replaceable"><code>condition</code></em> ]
[ GROUP BY <em class="replaceable"><code>grouping_element</code></em> [, ...] ]
[ HAVING <em class="replaceable"><code>condition</code></em> [, ...] ]
[ WINDOW <em class="replaceable"><code>window_name</code></em> AS ( <em class="replaceable"><code>window_definition</code></em> ) [, ...] ]
[ { UNION | INTERSECT | EXCEPT } [ ALL | DISTINCT ] <em class="replaceable"><code>select</code></em> ]
[ ORDER BY <em class="replaceable"><code>expression</code></em> [ ASC | DESC | USING <em class="replaceable"><code>operator</code></em> ] [ NULLS { FIRST | LAST } ] [, ...] ]
[ LIMIT { <em class="replaceable"><code>count</code></em> | ALL } ]
[ OFFSET <em class="replaceable"><code>start</code></em> [ ROW | ROWS ] ]
[ FETCH { FIRST | NEXT } [ <em class="replaceable"><code>count</code></em> ] { ROW | ROWS } ONLY ]
[ FOR { UPDATE | NO KEY UPDATE | SHARE | KEY SHARE } [ OF <em class="replaceable"><code>table_name</code></em> [, ...] ] [ NOWAIT | SKIP LOCKED ] [...] ]
<span class="phrase">where <em class="replaceable"><code>from_item</code></em> can be one of:</span>
[ ONLY ] <em class="replaceable"><code>table_name</code></em> [ * ] [ [ AS ] <em class="replaceable"><code>alias</code></em> [ ( <em class="replaceable"><code>column_alias</code></em> [, ...] ) ] ]
[ TABLESAMPLE <em class="replaceable"><code>sampling_method</code></em> ( <em class="replaceable"><code>argument</code></em> [, ...] ) [ REPEATABLE ( <em class="replaceable"><code>seed</code></em> ) ] ]
[ LATERAL ] ( <em class="replaceable"><code>select</code></em> ) [ AS ] <em class="replaceable"><code>alias</code></em> [ ( <em class="replaceable"><code>column_alias</code></em> [, ...] ) ]
<em class="replaceable"><code>with_query_name</code></em> [ [ AS ] <em class="replaceable"><code>alias</code></em> [ ( <em class="replaceable"><code>column_alias</code></em> [, ...] ) ] ]
[ LATERAL ] <em class="replaceable"><code>function_name</code></em> ( [ <em class="replaceable"><code>argument</code></em> [, ...] ] )
[ WITH ORDINALITY ] [ [ AS ] <em class="replaceable"><code>alias</code></em> [ ( <em class="replaceable"><code>column_alias</code></em> [, ...] ) ] ]
[ LATERAL ] <em class="replaceable"><code>function_name</code></em> ( [ <em class="replaceable"><code>argument</code></em> [, ...] ] ) [ AS ] <em class="replaceable"><code>alias</code></em> ( <em class="replaceable"><code>column_definition</code></em> [, ...] )
[ LATERAL ] <em class="replaceable"><code>function_name</code></em> ( [ <em class="replaceable"><code>argument</code></em> [, ...] ] ) AS ( <em class="replaceable"><code>column_definition</code></em> [, ...] )
[ LATERAL ] ROWS FROM( <em class="replaceable"><code>function_name</code></em> ( [ <em class="replaceable"><code>argument</code></em> [, ...] ] ) [ AS ( <em class="replaceable"><code>column_definition</code></em> [, ...] ) ] [, ...] )
[ WITH ORDINALITY ] [ [ AS ] <em class="replaceable"><code>alias</code></em> [ ( <em class="replaceable"><code>column_alias</code></em> [, ...] ) ] ]
<em class="replaceable"><code>from_item</code></em> [ NATURAL ] <em class="replaceable"><code>join_type</code></em> <em class="replaceable"><code>from_item</code></em> [ ON <em class="replaceable"><code>join_condition</code></em> | USING ( <em class="replaceable"><code>join_column</code></em> [, ...] ) ]
<span class="phrase">and <em class="replaceable"><code>grouping_element</code></em> can be one of:</span>
( )
<em class="replaceable"><code>expression</code></em>
( <em class="replaceable"><code>expression</code></em> [, ...] )
ROLLUP ( { <em class="replaceable"><code>expression</code></em> | ( <em class="replaceable"><code>expression</code></em> [, ...] ) } [, ...] )
CUBE ( { <em class="replaceable"><code>expression</code></em> | ( <em class="replaceable"><code>expression</code></em> [, ...] ) } [, ...] )
GROUPING SETS ( <em class="replaceable"><code>grouping_element</code></em> [, ...] )
<span class="phrase">and <em class="replaceable"><code>with_query</code></em> is:</span>
<em class="replaceable"><code>with_query_name</code></em> [ ( <em class="replaceable"><code>column_name</code></em> [, ...] ) ] AS ( <em class="replaceable"><code>select</code></em> | <em class="replaceable"><code>values</code></em> | <em class="replaceable"><code>insert</code></em> | <em class="replaceable"><code>update</code></em> | <em class="replaceable"><code>delete</code></em> )
TABLE [ ONLY ] <em class="replaceable"><code>table_name</code></em> [ * ]
</pre></div><div class="refsect1" id="id-1.9.3.171.7"><h2>Description</h2><p>
<code class="command">SELECT</code> retrieves rows from zero or more tables.
The general processing of <code class="command">SELECT</code> is as follows:
</p><div class="orderedlist"><ol class="orderedlist" type="1"><li class="listitem"><p>
All queries in the <code class="literal">WITH</code> list are computed.
These effectively serve as temporary tables that can be referenced
in the <code class="literal">FROM</code> list. A <code class="literal">WITH</code> query
that is referenced more than once in <code class="literal">FROM</code> is
computed only once.
(See <a class="xref" href="sql-select.html#SQL-WITH" title="WITH Clause"><code class="literal">WITH</code> Clause</a> below.)
</p></li><li class="listitem"><p>
All elements in the <code class="literal">FROM</code> list are computed.
(Each element in the <code class="literal">FROM</code> list is a real or
virtual table.) If more than one element is specified in the
<code class="literal">FROM</code> list, they are cross-joined together.
(See <a class="xref" href="sql-select.html#SQL-FROM" title="FROM Clause"><code class="literal">FROM</code> Clause</a> below.)
</p></li><li class="listitem"><p>
If the <code class="literal">WHERE</code> clause is specified, all rows
that do not satisfy the condition are eliminated from the
output. (See <a class="xref" href="sql-select.html#SQL-WHERE" title="WHERE Clause"><code class="literal">WHERE</code> Clause</a> below.)
</p></li><li class="listitem"><p>
If the <code class="literal">GROUP BY</code> clause is specified,
or if there are aggregate function calls, the
output is combined into groups of rows that match on one or more
values, and the results of aggregate functions are computed.
If the <code class="literal">HAVING</code> clause is present, it
eliminates groups that do not satisfy the given condition. (See
<a class="xref" href="sql-select.html#SQL-GROUPBY" title="GROUP BY Clause"><code class="literal">GROUP BY</code> Clause</a> and
<a class="xref" href="sql-select.html#SQL-HAVING" title="HAVING Clause"><code class="literal">HAVING</code> Clause</a> below.)
</p></li><li class="listitem"><p>
The actual output rows are computed using the
<code class="command">SELECT</code> output expressions for each selected
row or row group. (See
<a class="xref" href="sql-select.html#SQL-SELECT-LIST" title="SELECT List"><code class="command">SELECT</code> List</a>
below.)
</p></li><li class="listitem"><p><code class="literal">SELECT DISTINCT</code> eliminates duplicate rows from the
result. <code class="literal">SELECT DISTINCT ON</code> eliminates rows that
match on all the specified expressions. <code class="literal">SELECT ALL</code>
(the default) will return all candidate rows, including
duplicates. (See <a class="xref" href="sql-select.html#SQL-DISTINCT" title="DISTINCT Clause"><code class="literal">DISTINCT</code> Clause</a> below.)
</p></li><li class="listitem"><p>
Using the operators <code class="literal">UNION</code>,
<code class="literal">INTERSECT</code>, and <code class="literal">EXCEPT</code>, the
output of more than one <code class="command">SELECT</code> statement can
be combined to form a single result set. The
<code class="literal">UNION</code> operator returns all rows that are in
one or both of the result sets. The
<code class="literal">INTERSECT</code> operator returns all rows that are
strictly in both result sets. The <code class="literal">EXCEPT</code>
operator returns the rows that are in the first result set but
not in the second. In all three cases, duplicate rows are
eliminated unless <code class="literal">ALL</code> is specified. The noise
word <code class="literal">DISTINCT</code> can be added to explicitly specify
eliminating duplicate rows. Notice that <code class="literal">DISTINCT</code> is
the default behavior here, even though <code class="literal">ALL</code> is
the default for <code class="command">SELECT</code> itself. (See
<a class="xref" href="sql-select.html#SQL-UNION" title="UNION Clause"><code class="literal">UNION</code> Clause</a>, <a class="xref" href="sql-select.html#SQL-INTERSECT" title="INTERSECT Clause"><code class="literal">INTERSECT</code> Clause</a>, and
<a class="xref" href="sql-select.html#SQL-EXCEPT" title="EXCEPT Clause"><code class="literal">EXCEPT</code> Clause</a> below.)
</p></li><li class="listitem"><p>
If the <code class="literal">ORDER BY</code> clause is specified, the
returned rows are sorted in the specified order. If
<code class="literal">ORDER BY</code> is not given, the rows are returned
in whatever order the system finds fastest to produce. (See
<a class="xref" href="sql-select.html#SQL-ORDERBY" title="ORDER BY Clause"><code class="literal">ORDER BY</code> Clause</a> below.)
</p></li><li class="listitem"><p>
If the <code class="literal">LIMIT</code> (or <code class="literal">FETCH FIRST</code>) or <code class="literal">OFFSET</code>
clause is specified, the <code class="command">SELECT</code> statement
only returns a subset of the result rows. (See <a class="xref" href="sql-select.html#SQL-LIMIT" title="LIMIT Clause"><code class="literal">LIMIT</code> Clause</a> below.)
</p></li><li class="listitem"><p>
If <code class="literal">FOR UPDATE</code>, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR SHARE</code>
or <code class="literal">FOR KEY SHARE</code>
is specified, the
<code class="command">SELECT</code> statement locks the selected rows
against concurrent updates. (See <a class="xref" href="sql-select.html#SQL-FOR-UPDATE-SHARE" title="The Locking Clause">The Locking Clause</a> below.)
</p></li></ol></div><p>
</p><p>
You must have <code class="literal">SELECT</code> privilege on each column used
in a <code class="command">SELECT</code> command. The use of <code class="literal">FOR NO KEY UPDATE</code>,
<code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code> or <code class="literal">FOR KEY SHARE</code> requires
<code class="literal">UPDATE</code> privilege as well (for at least one column
of each table so selected).
</p></div><div class="refsect1" id="id-1.9.3.171.8"><h2>Parameters</h2><div class="refsect2" id="SQL-WITH"><h3><code class="literal">WITH</code> Clause</h3><p>
The <code class="literal">WITH</code> clause allows you to specify one or more
subqueries that can be referenced by name in the primary query.
The subqueries effectively act as temporary tables or views
for the duration of the primary query.
Each subquery can be a <code class="command">SELECT</code>, <code class="command">TABLE</code>, <code class="command">VALUES</code>,
<code class="command">INSERT</code>, <code class="command">UPDATE</code> or
<code class="command">DELETE</code> statement.
When writing a data-modifying statement (<code class="command">INSERT</code>,
<code class="command">UPDATE</code> or <code class="command">DELETE</code>) in
<code class="literal">WITH</code>, it is usual to include a <code class="literal">RETURNING</code> clause.
It is the output of <code class="literal">RETURNING</code>, <span class="emphasis"><em>not</em></span> the underlying
table that the statement modifies, that forms the temporary table that is
read by the primary query. If <code class="literal">RETURNING</code> is omitted, the
statement is still executed, but it produces no output so it cannot be
referenced as a table by the primary query.
</p><p>
A name (without schema qualification) must be specified for each
<code class="literal">WITH</code> query. Optionally, a list of column names
can be specified; if this is omitted,
the column names are inferred from the subquery.
</p><p>
If <code class="literal">RECURSIVE</code> is specified, it allows a
<code class="command">SELECT</code> subquery to reference itself by name. Such a
subquery must have the form
</p><pre class="synopsis">
<em class="replaceable"><code>non_recursive_term</code></em> UNION [ ALL | DISTINCT ] <em class="replaceable"><code>recursive_term</code></em>
</pre><p>
where the recursive self-reference must appear on the right-hand
side of the <code class="literal">UNION</code>. Only one recursive self-reference
is permitted per query. Recursive data-modifying statements are not
supported, but you can use the results of a recursive
<code class="command">SELECT</code> query in
a data-modifying statement. See <a class="xref" href="queries-with.html" title="7.8. WITH Queries (Common Table Expressions)">Section 7.8</a> for
an example.
</p><p>
Another effect of <code class="literal">RECURSIVE</code> is that
<code class="literal">WITH</code> queries need not be ordered: a query
can reference another one that is later in the list. (However,
circular references, or mutual recursion, are not implemented.)
Without <code class="literal">RECURSIVE</code>, <code class="literal">WITH</code> queries
can only reference sibling <code class="literal">WITH</code> queries
that are earlier in the <code class="literal">WITH</code> list.
</p><p>
A key property of <code class="literal">WITH</code> queries is that they
are evaluated only once per execution of the primary query,
even if the primary query refers to them more than once.
In particular, data-modifying statements are guaranteed to be
executed once and only once, regardless of whether the primary query
reads all or any of their output.
</p><p>
The primary query and the <code class="literal">WITH</code> queries are all
(notionally) executed at the same time. This implies that the effects of
a data-modifying statement in <code class="literal">WITH</code> cannot be seen from
other parts of the query, other than by reading its <code class="literal">RETURNING</code>
output. If two such data-modifying statements attempt to modify the same
row, the results are unspecified.
</p><p>
See <a class="xref" href="queries-with.html" title="7.8. WITH Queries (Common Table Expressions)">Section 7.8</a> for additional information.
</p></div><div class="refsect2" id="SQL-FROM"><h3><code class="literal">FROM</code> Clause</h3><p>
The <code class="literal">FROM</code> clause specifies one or more source
tables for the <code class="command">SELECT</code>. If multiple sources are
specified, the result is the Cartesian product (cross join) of all
the sources. But usually qualification conditions are added (via
<code class="literal">WHERE</code>) to restrict the returned rows to a small subset of the
Cartesian product.
</p><p>
The <code class="literal">FROM</code> clause can contain the following
elements:
</p><div class="variablelist"><dl class="variablelist"><dt><span class="term"><em class="replaceable"><code>table_name</code></em></span></dt><dd><p>
The name (optionally schema-qualified) of an existing table or view.
If <code class="literal">ONLY</code> is specified before the table name, only that
table is scanned. If <code class="literal">ONLY</code> is not specified, the table
and all its descendant tables (if any) are scanned. Optionally,
<code class="literal">*</code> can be specified after the table name to explicitly
indicate that descendant tables are included.
</p></dd><dt><span class="term"><em class="replaceable"><code>alias</code></em></span></dt><dd><p>
A substitute name for the <code class="literal">FROM</code> item containing the
alias. An alias is used for brevity or to eliminate ambiguity
for self-joins (where the same table is scanned multiple
times). When an alias is provided, it completely hides the
actual name of the table or function; for example given
<code class="literal">FROM foo AS f</code>, the remainder of the
<code class="command">SELECT</code> must refer to this <code class="literal">FROM</code>
item as <code class="literal">f</code> not <code class="literal">foo</code>. If an alias is
written, a column alias list can also be written to provide
substitute names for one or more columns of the table.
</p></dd><dt><span class="term"><code class="literal">TABLESAMPLE <em class="replaceable"><code>sampling_method</code></em> ( <em class="replaceable"><code>argument</code></em> [, ...] ) [ REPEATABLE ( <em class="replaceable"><code>seed</code></em> ) ]</code></span></dt><dd><p>
A <code class="literal">TABLESAMPLE</code> clause after
a <em class="replaceable"><code>table_name</code></em> indicates that the
specified <em class="replaceable"><code>sampling_method</code></em>
should be used to retrieve a subset of the rows in that table.
This sampling precedes the application of any other filters such
as <code class="literal">WHERE</code> clauses.
The standard <span class="productname">PostgreSQL</span> distribution
includes two sampling methods, <code class="literal">BERNOULLI</code>
and <code class="literal">SYSTEM</code>, and other sampling methods can be
installed in the database via extensions.
</p><p>
The <code class="literal">BERNOULLI</code> and <code class="literal">SYSTEM</code> sampling methods
each accept a single <em class="replaceable"><code>argument</code></em>
which is the fraction of the table to sample, expressed as a
percentage between 0 and 100. This argument can be
any <code class="type">real</code>-valued expression. (Other sampling methods might
accept more or different arguments.) These two methods each return
a randomly-chosen sample of the table that will contain
approximately the specified percentage of the table's rows.
The <code class="literal">BERNOULLI</code> method scans the whole table and
selects or ignores individual rows independently with the specified
probability.
The <code class="literal">SYSTEM</code> method does block-level sampling with
each block having the specified chance of being selected; all rows
in each selected block are returned.
The <code class="literal">SYSTEM</code> method is significantly faster than
the <code class="literal">BERNOULLI</code> method when small sampling
percentages are specified, but it may return a less-random sample of
the table as a result of clustering effects.
</p><p>
The optional <code class="literal">REPEATABLE</code> clause specifies
a <em class="replaceable"><code>seed</code></em> number or expression to use
for generating random numbers within the sampling method. The seed
value can be any non-null floating-point value. Two queries that
specify the same seed and <em class="replaceable"><code>argument</code></em>
values will select the same sample of the table, if the table has
not been changed meanwhile. But different seed values will usually
produce different samples.
If <code class="literal">REPEATABLE</code> is not given then a new random
sample is selected for each query, based upon a system-generated seed.
Note that some add-on sampling methods do not
accept <code class="literal">REPEATABLE</code>, and will always produce new
samples on each use.
</p></dd><dt><span class="term"><em class="replaceable"><code>select</code></em></span></dt><dd><p>
A sub-<code class="command">SELECT</code> can appear in the
<code class="literal">FROM</code> clause. This acts as though its
output were created as a temporary table for the duration of
this single <code class="command">SELECT</code> command. Note that the
sub-<code class="command">SELECT</code> must be surrounded by
parentheses, and an alias <span class="emphasis"><em>must</em></span> be
provided for it. A
<a class="xref" href="sql-values.html" title="VALUES"><span class="refentrytitle">VALUES</span></a> command
can also be used here.
</p></dd><dt><span class="term"><em class="replaceable"><code>with_query_name</code></em></span></dt><dd><p>
A <code class="literal">WITH</code> query is referenced by writing its name,
just as though the query's name were a table name. (In fact,
the <code class="literal">WITH</code> query hides any real table of the same name
for the purposes of the primary query. If necessary, you can
refer to a real table of the same name by schema-qualifying
the table's name.)
An alias can be provided in the same way as for a table.
</p></dd><dt><span class="term"><em class="replaceable"><code>function_name</code></em></span></dt><dd><p>
Function calls can appear in the <code class="literal">FROM</code>
clause. (This is especially useful for functions that return
result sets, but any function can be used.) This acts as
though the function's output were created as a temporary table for the
duration of this single <code class="command">SELECT</code> command.
When the optional <code class="command">WITH ORDINALITY</code> clause is
added to the function call, a new column is appended after
all the function's output columns with numbering for each row.
</p><p>
An alias can be provided in the same way as for a table.
If an alias is written, a column
alias list can also be written to provide substitute names for
one or more attributes of the function's composite return
type, including the column added by <code class="literal">ORDINALITY</code>
if present.
</p><p>
Multiple function calls can be combined into a
single <code class="literal">FROM</code>-clause item by surrounding them
with <code class="literal">ROWS FROM( ... )</code>. The output of such an item is the
concatenation of the first row from each function, then the second
row from each function, etc. If some of the functions produce fewer
rows than others, null values are substituted for the missing data, so
that the total number of rows returned is always the same as for the
function that produced the most rows.
</p><p>
If the function has been defined as returning the
<code class="type">record</code> data type, then an alias or the key word
<code class="literal">AS</code> must be present, followed by a column
definition list in the form <code class="literal">( <em class="replaceable"><code>column_name</code></em> <em class="replaceable"><code>data_type</code></em> [<span class="optional">, ...
</span>])</code>. The column definition list must match the
actual number and types of columns returned by the function.
</p><p>
When using the <code class="literal">ROWS FROM( ... )</code> syntax, if one of the
functions requires a column definition list, it's preferred to put
the column definition list after the function call inside
<code class="literal">ROWS FROM( ... )</code>. A column definition list can be placed
after the <code class="literal">ROWS FROM( ... )</code> construct only if there's just
a single function and no <code class="literal">WITH ORDINALITY</code> clause.
</p><p>
To use <code class="literal">ORDINALITY</code> together with a column definition
list, you must use the <code class="literal">ROWS FROM( ... )</code> syntax and put the
column definition list inside <code class="literal">ROWS FROM( ... )</code>.
</p></dd><dt><span class="term"><em class="replaceable"><code>join_type</code></em></span></dt><dd><p>
One of
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p><code class="literal">[ INNER ] JOIN</code></p></li><li class="listitem"><p><code class="literal">LEFT [ OUTER ] JOIN</code></p></li><li class="listitem"><p><code class="literal">RIGHT [ OUTER ] JOIN</code></p></li><li class="listitem"><p><code class="literal">FULL [ OUTER ] JOIN</code></p></li><li class="listitem"><p><code class="literal">CROSS JOIN</code></p></li></ul></div><p>
For the <code class="literal">INNER</code> and <code class="literal">OUTER</code> join types, a
join condition must be specified, namely exactly one of
<code class="literal">NATURAL</code>, <code class="literal">ON <em class="replaceable"><code>join_condition</code></em></code>, or
<code class="literal">USING (<em class="replaceable"><code>join_column</code></em> [, ...])</code>.
See below for the meaning. For <code class="literal">CROSS JOIN</code>,
none of these clauses can appear.
</p><p>
A <code class="literal">JOIN</code> clause combines two <code class="literal">FROM</code>
items, which for convenience we will refer to as <span class="quote">“<span class="quote">tables</span>”</span>,
though in reality they can be any type of <code class="literal">FROM</code> item.
Use parentheses if necessary to determine the order of nesting.
In the absence of parentheses, <code class="literal">JOIN</code>s nest
left-to-right. In any case <code class="literal">JOIN</code> binds more
tightly than the commas separating <code class="literal">FROM</code>-list items.
</p><p><code class="literal">CROSS JOIN</code> and <code class="literal">INNER JOIN</code>
produce a simple Cartesian product, the same result as you get from
listing the two tables at the top level of <code class="literal">FROM</code>,
but restricted by the join condition (if any).
<code class="literal">CROSS JOIN</code> is equivalent to <code class="literal">INNER JOIN ON
(TRUE)</code>, that is, no rows are removed by qualification.
These join types are just a notational convenience, since they
do nothing you couldn't do with plain <code class="literal">FROM</code> and
<code class="literal">WHERE</code>.
</p><p><code class="literal">LEFT OUTER JOIN</code> returns all rows in the qualified
Cartesian product (i.e., all combined rows that pass its join
condition), plus one copy of each row in the left-hand table
for which there was no right-hand row that passed the join
condition. This left-hand row is extended to the full width
of the joined table by inserting null values for the
right-hand columns. Note that only the <code class="literal">JOIN</code>
clause's own condition is considered while deciding which rows
have matches. Outer conditions are applied afterwards.
</p><p>
Conversely, <code class="literal">RIGHT OUTER JOIN</code> returns all the
joined rows, plus one row for each unmatched right-hand row
(extended with nulls on the left). This is just a notational
convenience, since you could convert it to a <code class="literal">LEFT
OUTER JOIN</code> by switching the left and right tables.
</p><p><code class="literal">FULL OUTER JOIN</code> returns all the joined rows, plus
one row for each unmatched left-hand row (extended with nulls
on the right), plus one row for each unmatched right-hand row
(extended with nulls on the left).
</p></dd><dt><span class="term"><code class="literal">ON <em class="replaceable"><code>join_condition</code></em></code></span></dt><dd><p><em class="replaceable"><code>join_condition</code></em> is
an expression resulting in a value of type
<code class="type">boolean</code> (similar to a <code class="literal">WHERE</code>
clause) that specifies which rows in a join are considered to
match.
</p></dd><dt><span class="term"><code class="literal">USING ( <em class="replaceable"><code>join_column</code></em> [, ...] )</code></span></dt><dd><p>
A clause of the form <code class="literal">USING ( a, b, ... )</code> is
shorthand for <code class="literal">ON left_table.a = right_table.a AND
left_table.b = right_table.b ...</code>. Also,
<code class="literal">USING</code> implies that only one of each pair of
equivalent columns will be included in the join output, not
both.
</p></dd><dt><span class="term"><code class="literal">NATURAL</code></span></dt><dd><p>
<code class="literal">NATURAL</code> is shorthand for a
<code class="literal">USING</code> list that mentions all columns in the two
tables that have matching names. If there are no common
column names, <code class="literal">NATURAL</code> is equivalent
to <code class="literal">ON TRUE</code>.
</p></dd><dt><span class="term"><code class="literal">LATERAL</code></span></dt><dd><p>
The <code class="literal">LATERAL</code> key word can precede a
sub-<code class="command">SELECT</code> <code class="literal">FROM</code> item. This allows the
sub-<code class="command">SELECT</code> to refer to columns of <code class="literal">FROM</code>
items that appear before it in the <code class="literal">FROM</code> list. (Without
<code class="literal">LATERAL</code>, each sub-<code class="command">SELECT</code> is
evaluated independently and so cannot cross-reference any other
<code class="literal">FROM</code> item.)
</p><p><code class="literal">LATERAL</code> can also precede a function-call
<code class="literal">FROM</code> item, but in this case it is a noise word, because
the function expression can refer to earlier <code class="literal">FROM</code> items
in any case.
</p><p>
A <code class="literal">LATERAL</code> item can appear at top level in the
<code class="literal">FROM</code> list, or within a <code class="literal">JOIN</code> tree. In the
latter case it can also refer to any items that are on the left-hand
side of a <code class="literal">JOIN</code> that it is on the right-hand side of.
</p><p>
When a <code class="literal">FROM</code> item contains <code class="literal">LATERAL</code>
cross-references, evaluation proceeds as follows: for each row of the
<code class="literal">FROM</code> item providing the cross-referenced column(s), or
set of rows of multiple <code class="literal">FROM</code> items providing the
columns, the <code class="literal">LATERAL</code> item is evaluated using that
row or row set's values of the columns. The resulting row(s) are
joined as usual with the rows they were computed from. This is
repeated for each row or set of rows from the column source table(s).
</p><p>
The column source table(s) must be <code class="literal">INNER</code> or
<code class="literal">LEFT</code> joined to the <code class="literal">LATERAL</code> item, else
there would not be a well-defined set of rows from which to compute
each set of rows for the <code class="literal">LATERAL</code> item. Thus,
although a construct such as <code class="literal"><em class="replaceable"><code>X</code></em> RIGHT JOIN
LATERAL <em class="replaceable"><code>Y</code></em></code> is syntactically valid, it is
not actually allowed for <em class="replaceable"><code>Y</code></em> to reference
<em class="replaceable"><code>X</code></em>.
</p></dd></dl></div><p>
</p></div><div class="refsect2" id="SQL-WHERE"><h3><code class="literal">WHERE</code> Clause</h3><p>
The optional <code class="literal">WHERE</code> clause has the general form
</p><pre class="synopsis">
WHERE <em class="replaceable"><code>condition</code></em>
</pre><p>
where <em class="replaceable"><code>condition</code></em> is
any expression that evaluates to a result of type
<code class="type">boolean</code>. Any row that does not satisfy this
condition will be eliminated from the output. A row satisfies the
condition if it returns true when the actual row values are
substituted for any variable references.
</p></div><div class="refsect2" id="SQL-GROUPBY"><h3><code class="literal">GROUP BY</code> Clause</h3><p>
The optional <code class="literal">GROUP BY</code> clause has the general form
</p><pre class="synopsis">
GROUP BY <em class="replaceable"><code>grouping_element</code></em> [, ...]
</pre><p>
</p><p>
<code class="literal">GROUP BY</code> will condense into a single row all
selected rows that share the same values for the grouped
expressions. An <em class="replaceable"><code>expression</code></em> used inside a
<em class="replaceable"><code>grouping_element</code></em>
can be an input column name, or the name or ordinal number of an
output column (<code class="command">SELECT</code> list item), or an arbitrary
expression formed from input-column values. In case of ambiguity,
a <code class="literal">GROUP BY</code> name will be interpreted as an
input-column name rather than an output column name.
</p><p>
If any of <code class="literal">GROUPING SETS</code>, <code class="literal">ROLLUP</code> or
<code class="literal">CUBE</code> are present as grouping elements, then the
<code class="literal">GROUP BY</code> clause as a whole defines some number of
independent <em class="replaceable"><code>grouping sets</code></em>. The effect of this is
equivalent to constructing a <code class="literal">UNION ALL</code> between
subqueries with the individual grouping sets as their
<code class="literal">GROUP BY</code> clauses. For further details on the handling
of grouping sets see <a class="xref" href="queries-table-expressions.html#QUERIES-GROUPING-SETS" title="7.2.4. GROUPING SETS, CUBE, and ROLLUP">Section 7.2.4</a>.
</p><p>
Aggregate functions, if any are used, are computed across all rows
making up each group, producing a separate value for each group.
(If there are aggregate functions but no <code class="literal">GROUP BY</code>
clause, the query is treated as having a single group comprising all
the selected rows.)
The set of rows fed to each aggregate function can be further filtered by
attaching a <code class="literal">FILTER</code> clause to the aggregate function
call; see <a class="xref" href="sql-expressions.html#SYNTAX-AGGREGATES" title="4.2.7. Aggregate Expressions">Section 4.2.7</a> for more information. When
a <code class="literal">FILTER</code> clause is present, only those rows matching it
are included in the input to that aggregate function.
</p><p>
When <code class="literal">GROUP BY</code> is present,
or any aggregate functions are present, it is not valid for
the <code class="command">SELECT</code> list expressions to refer to
ungrouped columns except within aggregate functions or when the
ungrouped column is functionally dependent on the grouped columns,
since there would otherwise be more than one possible value to
return for an ungrouped column. A functional dependency exists if
the grouped columns (or a subset thereof) are the primary key of
the table containing the ungrouped column.
</p><p>
Keep in mind that all aggregate functions are evaluated before
evaluating any <span class="quote">“<span class="quote">scalar</span>”</span> expressions in the <code class="literal">HAVING</code>
clause or <code class="literal">SELECT</code> list. This means that, for example,
a <code class="literal">CASE</code> expression cannot be used to skip evaluation of
an aggregate function; see <a class="xref" href="sql-expressions.html#SYNTAX-EXPRESS-EVAL" title="4.2.14. Expression Evaluation Rules">Section 4.2.14</a>.
</p><p>
Currently, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code> and <code class="literal">FOR KEY SHARE</code> cannot be
specified with <code class="literal">GROUP BY</code>.
</p></div><div class="refsect2" id="SQL-HAVING"><h3><code class="literal">HAVING</code> Clause</h3><p>
The optional <code class="literal">HAVING</code> clause has the general form
</p><pre class="synopsis">
HAVING <em class="replaceable"><code>condition</code></em>
</pre><p>
where <em class="replaceable"><code>condition</code></em> is
the same as specified for the <code class="literal">WHERE</code> clause.
</p><p>
<code class="literal">HAVING</code> eliminates group rows that do not
satisfy the condition. <code class="literal">HAVING</code> is different
from <code class="literal">WHERE</code>: <code class="literal">WHERE</code> filters
individual rows before the application of <code class="literal">GROUP
BY</code>, while <code class="literal">HAVING</code> filters group rows
created by <code class="literal">GROUP BY</code>. Each column referenced in
<em class="replaceable"><code>condition</code></em> must
unambiguously reference a grouping column, unless the reference
appears within an aggregate function or the ungrouped column is
functionally dependent on the grouping columns.
</p><p>
The presence of <code class="literal">HAVING</code> turns a query into a grouped
query even if there is no <code class="literal">GROUP BY</code> clause. This is the
same as what happens when the query contains aggregate functions but
no <code class="literal">GROUP BY</code> clause. All the selected rows are considered to
form a single group, and the <code class="command">SELECT</code> list and
<code class="literal">HAVING</code> clause can only reference table columns from
within aggregate functions. Such a query will emit a single row if the
<code class="literal">HAVING</code> condition is true, zero rows if it is not true.
</p><p>
Currently, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code> and <code class="literal">FOR KEY SHARE</code> cannot be
specified with <code class="literal">HAVING</code>.
</p></div><div class="refsect2" id="SQL-WINDOW"><h3><code class="literal">WINDOW</code> Clause</h3><p>
The optional <code class="literal">WINDOW</code> clause has the general form
</p><pre class="synopsis">
WINDOW <em class="replaceable"><code>window_name</code></em> AS ( <em class="replaceable"><code>window_definition</code></em> ) [, ...]
</pre><p>
where <em class="replaceable"><code>window_name</code></em> is
a name that can be referenced from <code class="literal">OVER</code> clauses or
subsequent window definitions, and
<em class="replaceable"><code>window_definition</code></em> is
</p><pre class="synopsis">
[ <em class="replaceable"><code>existing_window_name</code></em> ]
[ PARTITION BY <em class="replaceable"><code>expression</code></em> [, ...] ]
[ ORDER BY <em class="replaceable"><code>expression</code></em> [ ASC | DESC | USING <em class="replaceable"><code>operator</code></em> ] [ NULLS { FIRST | LAST } ] [, ...] ]
[ <em class="replaceable"><code>frame_clause</code></em> ]
</pre><p>
</p><p>
If an <em class="replaceable"><code>existing_window_name</code></em>
is specified it must refer to an earlier entry in the <code class="literal">WINDOW</code>
list; the new window copies its partitioning clause from that entry,
as well as its ordering clause if any. In this case the new window cannot
specify its own <code class="literal">PARTITION BY</code> clause, and it can specify
<code class="literal">ORDER BY</code> only if the copied window does not have one.
The new window always uses its own frame clause; the copied window
must not specify a frame clause.
</p><p>
The elements of the <code class="literal">PARTITION BY</code> list are interpreted in
much the same fashion as elements of a
<a class="xref" href="sql-select.html#SQL-GROUPBY" title="GROUP BY Clause"><code class="literal">GROUP BY</code> Clause</a>, except that
they are always simple expressions and never the name or number of an
output column.
Another difference is that these expressions can contain aggregate
function calls, which are not allowed in a regular <code class="literal">GROUP BY</code>
clause. They are allowed here because windowing occurs after grouping
and aggregation.
</p><p>
Similarly, the elements of the <code class="literal">ORDER BY</code> list are interpreted
in much the same fashion as elements of an
<a class="xref" href="sql-select.html#SQL-ORDERBY" title="ORDER BY Clause"><code class="literal">ORDER BY</code> Clause</a>, except that
the expressions are always taken as simple expressions and never the name
or number of an output column.
</p><p>
The optional <em class="replaceable"><code>frame_clause</code></em> defines
the <em class="firstterm">window frame</em> for window functions that depend on the
frame (not all do). The window frame is a set of related rows for
each row of the query (called the <em class="firstterm">current row</em>).
The <em class="replaceable"><code>frame_clause</code></em> can be one of
</p><pre class="synopsis">
{ RANGE | ROWS | GROUPS } <em class="replaceable"><code>frame_start</code></em> [ <em class="replaceable"><code>frame_exclusion</code></em> ]
{ RANGE | ROWS | GROUPS } BETWEEN <em class="replaceable"><code>frame_start</code></em> AND <em class="replaceable"><code>frame_end</code></em> [ <em class="replaceable"><code>frame_exclusion</code></em> ]
</pre><p>
where <em class="replaceable"><code>frame_start</code></em>
and <em class="replaceable"><code>frame_end</code></em> can be one of
</p><pre class="synopsis">
UNBOUNDED PRECEDING
<em class="replaceable"><code>offset</code></em> PRECEDING
CURRENT ROW
<em class="replaceable"><code>offset</code></em> FOLLOWING
UNBOUNDED FOLLOWING
</pre><p>
and <em class="replaceable"><code>frame_exclusion</code></em> can be one of
</p><pre class="synopsis">
EXCLUDE CURRENT ROW
EXCLUDE GROUP
EXCLUDE TIES
EXCLUDE NO OTHERS
</pre><p>
If <em class="replaceable"><code>frame_end</code></em> is omitted it defaults to <code class="literal">CURRENT
ROW</code>. Restrictions are that
<em class="replaceable"><code>frame_start</code></em> cannot be <code class="literal">UNBOUNDED FOLLOWING</code>,
<em class="replaceable"><code>frame_end</code></em> cannot be <code class="literal">UNBOUNDED PRECEDING</code>,
and the <em class="replaceable"><code>frame_end</code></em> choice cannot appear earlier in the
above list of <em class="replaceable"><code>frame_start</code></em>
and <em class="replaceable"><code>frame_end</code></em> options than
the <em class="replaceable"><code>frame_start</code></em> choice does — for example
<code class="literal">RANGE BETWEEN CURRENT ROW AND <em class="replaceable"><code>offset</code></em>
PRECEDING</code> is not allowed.
</p><p>
The default framing option is <code class="literal">RANGE UNBOUNDED PRECEDING</code>,
which is the same as <code class="literal">RANGE BETWEEN UNBOUNDED PRECEDING AND
CURRENT ROW</code>; it sets the frame to be all rows from the partition start
up through the current row's last <em class="firstterm">peer</em> (a row
that the window's <code class="literal">ORDER BY</code> clause considers
equivalent to the current row; all rows are peers if there
is no <code class="literal">ORDER BY</code>).
In general, <code class="literal">UNBOUNDED PRECEDING</code> means that the frame
starts with the first row of the partition, and similarly
<code class="literal">UNBOUNDED FOLLOWING</code> means that the frame ends with the last
row of the partition, regardless
of <code class="literal">RANGE</code>, <code class="literal">ROWS</code>
or <code class="literal">GROUPS</code> mode.
In <code class="literal">ROWS</code> mode, <code class="literal">CURRENT ROW</code> means
that the frame starts or ends with the current row; but
in <code class="literal">RANGE</code> or <code class="literal">GROUPS</code> mode it means
that the frame starts or ends with the current row's first or last peer
in the <code class="literal">ORDER BY</code> ordering.
The <em class="replaceable"><code>offset</code></em> <code class="literal">PRECEDING</code> and
<em class="replaceable"><code>offset</code></em> <code class="literal">FOLLOWING</code> options
vary in meaning depending on the frame mode.
In <code class="literal">ROWS</code> mode, the <em class="replaceable"><code>offset</code></em>
is an integer indicating that the frame starts or ends that many rows
before or after the current row.
In <code class="literal">GROUPS</code> mode, the <em class="replaceable"><code>offset</code></em>
is an integer indicating that the frame starts or ends that many peer
groups before or after the current row's peer group, where
a <em class="firstterm">peer group</em> is a group of rows that are
equivalent according to the window's <code class="literal">ORDER BY</code> clause.
In <code class="literal">RANGE</code> mode, use of
an <em class="replaceable"><code>offset</code></em> option requires that there be
exactly one <code class="literal">ORDER BY</code> column in the window definition.
Then the frame contains those rows whose ordering column value is no
more than <em class="replaceable"><code>offset</code></em> less than
(for <code class="literal">PRECEDING</code>) or more than
(for <code class="literal">FOLLOWING</code>) the current row's ordering column
value. In these cases the data type of
the <em class="replaceable"><code>offset</code></em> expression depends on the data
type of the ordering column. For numeric ordering columns it is
typically of the same type as the ordering column, but for datetime
ordering columns it is an <code class="type">interval</code>.
In all these cases, the value of the <em class="replaceable"><code>offset</code></em>
must be non-null and non-negative. Also, while
the <em class="replaceable"><code>offset</code></em> does not have to be a simple
constant, it cannot contain variables, aggregate functions, or window
functions.
</p><p>
The <em class="replaceable"><code>frame_exclusion</code></em> option allows rows around
the current row to be excluded from the frame, even if they would be
included according to the frame start and frame end options.
<code class="literal">EXCLUDE CURRENT ROW</code> excludes the current row from the
frame.
<code class="literal">EXCLUDE GROUP</code> excludes the current row and its
ordering peers from the frame.
<code class="literal">EXCLUDE TIES</code> excludes any peers of the current
row from the frame, but not the current row itself.
<code class="literal">EXCLUDE NO OTHERS</code> simply specifies explicitly the
default behavior of not excluding the current row or its peers.
</p><p>
Beware that the <code class="literal">ROWS</code> mode can produce unpredictable
results if the <code class="literal">ORDER BY</code> ordering does not order the rows
uniquely. The <code class="literal">RANGE</code> and <code class="literal">GROUPS</code>
modes are designed to ensure that rows that are peers in
the <code class="literal">ORDER BY</code> ordering are treated alike: all rows of
a given peer group will be in the frame or excluded from it.
</p><p>
The purpose of a <code class="literal">WINDOW</code> clause is to specify the
behavior of <em class="firstterm">window functions</em> appearing in the query's
<a class="xref" href="sql-select.html#SQL-SELECT-LIST" title="SELECT List"><code class="command">SELECT</code> List</a> or
<a class="xref" href="sql-select.html#SQL-ORDERBY" title="ORDER BY Clause"><code class="literal">ORDER BY</code> Clause</a>. These functions
can reference the <code class="literal">WINDOW</code> clause entries by name
in their <code class="literal">OVER</code> clauses. A <code class="literal">WINDOW</code> clause
entry does not have to be referenced anywhere, however; if it is not
used in the query it is simply ignored. It is possible to use window
functions without any <code class="literal">WINDOW</code> clause at all, since
a window function call can specify its window definition directly in
its <code class="literal">OVER</code> clause. However, the <code class="literal">WINDOW</code>
clause saves typing when the same window definition is needed for more
than one window function.
</p><p>
Currently, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code> and <code class="literal">FOR KEY SHARE</code> cannot be
specified with <code class="literal">WINDOW</code>.
</p><p>
Window functions are described in detail in
<a class="xref" href="tutorial-window.html" title="3.5. Window Functions">Section 3.5</a>,
<a class="xref" href="sql-expressions.html#SYNTAX-WINDOW-FUNCTIONS" title="4.2.8. Window Function Calls">Section 4.2.8</a>, and
<a class="xref" href="queries-table-expressions.html#QUERIES-WINDOW" title="7.2.5. Window Function Processing">Section 7.2.5</a>.
</p></div><div class="refsect2" id="SQL-SELECT-LIST"><h3><code class="command">SELECT</code> List</h3><p>
The <code class="command">SELECT</code> list (between the key words
<code class="literal">SELECT</code> and <code class="literal">FROM</code>) specifies expressions
that form the output rows of the <code class="command">SELECT</code>
statement. The expressions can (and usually do) refer to columns
computed in the <code class="literal">FROM</code> clause.
</p><p>
Just as in a table, every output column of a <code class="command">SELECT</code>
has a name. In a simple <code class="command">SELECT</code> this name is just
used to label the column for display, but when the <code class="command">SELECT</code>
is a sub-query of a larger query, the name is seen by the larger query
as the column name of the virtual table produced by the sub-query.
To specify the name to use for an output column, write
<code class="literal">AS</code> <em class="replaceable"><code>output_name</code></em>
after the column's expression. (You can omit <code class="literal">AS</code>,
but only if the desired output name does not match any
<span class="productname">PostgreSQL</span> keyword (see <a class="xref" href="sql-keywords-appendix.html" title="Appendix C. SQL Key Words">Appendix C</a>). For protection against possible
future keyword additions, it is recommended that you always either
write <code class="literal">AS</code> or double-quote the output name.)
If you do not specify a column name, a name is chosen automatically
by <span class="productname">PostgreSQL</span>. If the column's expression
is a simple column reference then the chosen name is the same as that
column's name. In more complex cases a function or type name may be
used, or the system may fall back on a generated name such as
<code class="literal">?column?</code>.
</p><p>
An output column's name can be used to refer to the column's value in
<code class="literal">ORDER BY</code> and <code class="literal">GROUP BY</code> clauses, but not in the
<code class="literal">WHERE</code> or <code class="literal">HAVING</code> clauses; there you must write
out the expression instead.
</p><p>
Instead of an expression, <code class="literal">*</code> can be written in
the output list as a shorthand for all the columns of the selected
rows. Also, you can write <code class="literal"><em class="replaceable"><code>table_name</code></em>.*</code> as a
shorthand for the columns coming from just that table. In these
cases it is not possible to specify new names with <code class="literal">AS</code>;
the output column names will be the same as the table columns' names.
</p><p>
According to the SQL standard, the expressions in the output list should
be computed before applying <code class="literal">DISTINCT</code>, <code class="literal">ORDER
BY</code>, or <code class="literal">LIMIT</code>. This is obviously necessary
when using <code class="literal">DISTINCT</code>, since otherwise it's not clear
what values are being made distinct. However, in many cases it is
convenient if output expressions are computed after <code class="literal">ORDER
BY</code> and <code class="literal">LIMIT</code>; particularly if the output list
contains any volatile or expensive functions. With that behavior, the
order of function evaluations is more intuitive and there will not be
evaluations corresponding to rows that never appear in the output.
<span class="productname">PostgreSQL</span> will effectively evaluate output expressions
after sorting and limiting, so long as those expressions are not
referenced in <code class="literal">DISTINCT</code>, <code class="literal">ORDER BY</code>
or <code class="literal">GROUP BY</code>. (As a counterexample, <code class="literal">SELECT
f(x) FROM tab ORDER BY 1</code> clearly must evaluate <code class="function">f(x)</code>
before sorting.) Output expressions that contain set-returning functions
are effectively evaluated after sorting and before limiting, so
that <code class="literal">LIMIT</code> will act to cut off the output from a
set-returning function.
</p><div class="note"><h3 class="title">Note</h3><p>
<span class="productname">PostgreSQL</span> versions before 9.6 did not provide any
guarantees about the timing of evaluation of output expressions versus
sorting and limiting; it depended on the form of the chosen query plan.
</p></div></div><div class="refsect2" id="SQL-DISTINCT"><h3><code class="literal">DISTINCT</code> Clause</h3><p>
If <code class="literal">SELECT DISTINCT</code> is specified, all duplicate rows are
removed from the result set (one row is kept from each group of
duplicates). <code class="literal">SELECT ALL</code> specifies the opposite: all rows are
kept; that is the default.
</p><p>
<code class="literal">SELECT DISTINCT ON ( <em class="replaceable"><code>expression</code></em> [, ...] )</code>
keeps only the first row of each set of rows where the given
expressions evaluate to equal. The <code class="literal">DISTINCT ON</code>
expressions are interpreted using the same rules as for
<code class="literal">ORDER BY</code> (see above). Note that the <span class="quote">“<span class="quote">first
row</span>”</span> of each set is unpredictable unless <code class="literal">ORDER
BY</code> is used to ensure that the desired row appears first. For
example:
</p><pre class="programlisting">
SELECT DISTINCT ON (location) location, time, report
FROM weather_reports
ORDER BY location, time DESC;
</pre><p>
retrieves the most recent weather report for each location. But
if we had not used <code class="literal">ORDER BY</code> to force descending order
of time values for each location, we'd have gotten a report from
an unpredictable time for each location.
</p><p>
The <code class="literal">DISTINCT ON</code> expression(s) must match the leftmost
<code class="literal">ORDER BY</code> expression(s). The <code class="literal">ORDER BY</code> clause
will normally contain additional expression(s) that determine the
desired precedence of rows within each <code class="literal">DISTINCT ON</code> group.
</p><p>
Currently, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code> and <code class="literal">FOR KEY SHARE</code> cannot be
specified with <code class="literal">DISTINCT</code>.
</p></div><div class="refsect2" id="SQL-UNION"><h3><code class="literal">UNION</code> Clause</h3><p>
The <code class="literal">UNION</code> clause has this general form:
</p><pre class="synopsis">
<em class="replaceable"><code>select_statement</code></em> UNION [ ALL | DISTINCT ] <em class="replaceable"><code>select_statement</code></em>
</pre><p><em class="replaceable"><code>select_statement</code></em> is
any <code class="command">SELECT</code> statement without an <code class="literal">ORDER
BY</code>, <code class="literal">LIMIT</code>, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code>, or <code class="literal">FOR KEY SHARE</code> clause.
(<code class="literal">ORDER BY</code> and <code class="literal">LIMIT</code> can be attached to a
subexpression if it is enclosed in parentheses. Without
parentheses, these clauses will be taken to apply to the result of
the <code class="literal">UNION</code>, not to its right-hand input
expression.)
</p><p>
The <code class="literal">UNION</code> operator computes the set union of
the rows returned by the involved <code class="command">SELECT</code>
statements. A row is in the set union of two result sets if it
appears in at least one of the result sets. The two
<code class="command">SELECT</code> statements that represent the direct
operands of the <code class="literal">UNION</code> must produce the same
number of columns, and corresponding columns must be of compatible
data types.
</p><p>
The result of <code class="literal">UNION</code> does not contain any duplicate
rows unless the <code class="literal">ALL</code> option is specified.
<code class="literal">ALL</code> prevents elimination of duplicates. (Therefore,
<code class="literal">UNION ALL</code> is usually significantly quicker than
<code class="literal">UNION</code>; use <code class="literal">ALL</code> when you can.)
<code class="literal">DISTINCT</code> can be written to explicitly specify the
default behavior of eliminating duplicate rows.
</p><p>
Multiple <code class="literal">UNION</code> operators in the same
<code class="command">SELECT</code> statement are evaluated left to right,
unless otherwise indicated by parentheses.
</p><p>
Currently, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>, <code class="literal">FOR SHARE</code> and
<code class="literal">FOR KEY SHARE</code> cannot be
specified either for a <code class="literal">UNION</code> result or for any input of a
<code class="literal">UNION</code>.
</p></div><div class="refsect2" id="SQL-INTERSECT"><h3><code class="literal">INTERSECT</code> Clause</h3><p>
The <code class="literal">INTERSECT</code> clause has this general form:
</p><pre class="synopsis">
<em class="replaceable"><code>select_statement</code></em> INTERSECT [ ALL | DISTINCT ] <em class="replaceable"><code>select_statement</code></em>
</pre><p><em class="replaceable"><code>select_statement</code></em> is
any <code class="command">SELECT</code> statement without an <code class="literal">ORDER
BY</code>, <code class="literal">LIMIT</code>, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code>, or <code class="literal">FOR KEY SHARE</code> clause.
</p><p>
The <code class="literal">INTERSECT</code> operator computes the set
intersection of the rows returned by the involved
<code class="command">SELECT</code> statements. A row is in the
intersection of two result sets if it appears in both result sets.
</p><p>
The result of <code class="literal">INTERSECT</code> does not contain any
duplicate rows unless the <code class="literal">ALL</code> option is specified.
With <code class="literal">ALL</code>, a row that has <em class="replaceable"><code>m</code></em> duplicates in the
left table and <em class="replaceable"><code>n</code></em> duplicates in the right table will appear
min(<em class="replaceable"><code>m</code></em>,<em class="replaceable"><code>n</code></em>) times in the result set.
<code class="literal">DISTINCT</code> can be written to explicitly specify the
default behavior of eliminating duplicate rows.
</p><p>
Multiple <code class="literal">INTERSECT</code> operators in the same
<code class="command">SELECT</code> statement are evaluated left to right,
unless parentheses dictate otherwise.
<code class="literal">INTERSECT</code> binds more tightly than
<code class="literal">UNION</code>. That is, <code class="literal">A UNION B INTERSECT
C</code> will be read as <code class="literal">A UNION (B INTERSECT
C)</code>.
</p><p>
Currently, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>, <code class="literal">FOR SHARE</code> and
<code class="literal">FOR KEY SHARE</code> cannot be
specified either for an <code class="literal">INTERSECT</code> result or for any input of
an <code class="literal">INTERSECT</code>.
</p></div><div class="refsect2" id="SQL-EXCEPT"><h3><code class="literal">EXCEPT</code> Clause</h3><p>
The <code class="literal">EXCEPT</code> clause has this general form:
</p><pre class="synopsis">
<em class="replaceable"><code>select_statement</code></em> EXCEPT [ ALL | DISTINCT ] <em class="replaceable"><code>select_statement</code></em>
</pre><p><em class="replaceable"><code>select_statement</code></em> is
any <code class="command">SELECT</code> statement without an <code class="literal">ORDER
BY</code>, <code class="literal">LIMIT</code>, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>,
<code class="literal">FOR SHARE</code>, or <code class="literal">FOR KEY SHARE</code> clause.
</p><p>
The <code class="literal">EXCEPT</code> operator computes the set of rows
that are in the result of the left <code class="command">SELECT</code>
statement but not in the result of the right one.
</p><p>
The result of <code class="literal">EXCEPT</code> does not contain any
duplicate rows unless the <code class="literal">ALL</code> option is specified.
With <code class="literal">ALL</code>, a row that has <em class="replaceable"><code>m</code></em> duplicates in the
left table and <em class="replaceable"><code>n</code></em> duplicates in the right table will appear
max(<em class="replaceable"><code>m</code></em>-<em class="replaceable"><code>n</code></em>,0) times in the result set.
<code class="literal">DISTINCT</code> can be written to explicitly specify the
default behavior of eliminating duplicate rows.
</p><p>
Multiple <code class="literal">EXCEPT</code> operators in the same
<code class="command">SELECT</code> statement are evaluated left to right,
unless parentheses dictate otherwise. <code class="literal">EXCEPT</code> binds at
the same level as <code class="literal">UNION</code>.
</p><p>
Currently, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>, <code class="literal">FOR SHARE</code> and
<code class="literal">FOR KEY SHARE</code> cannot be
specified either for an <code class="literal">EXCEPT</code> result or for any input of
an <code class="literal">EXCEPT</code>.
</p></div><div class="refsect2" id="SQL-ORDERBY"><h3><code class="literal">ORDER BY</code> Clause</h3><p>
The optional <code class="literal">ORDER BY</code> clause has this general form:
</p><pre class="synopsis">
ORDER BY <em class="replaceable"><code>expression</code></em> [ ASC | DESC | USING <em class="replaceable"><code>operator</code></em> ] [ NULLS { FIRST | LAST } ] [, ...]
</pre><p>
The <code class="literal">ORDER BY</code> clause causes the result rows to
be sorted according to the specified expression(s). If two rows are
equal according to the leftmost expression, they are compared
according to the next expression and so on. If they are equal
according to all specified expressions, they are returned in
an implementation-dependent order.
</p><p>
Each <em class="replaceable"><code>expression</code></em> can be the
name or ordinal number of an output column
(<code class="command">SELECT</code> list item), or it can be an arbitrary
expression formed from input-column values.
</p><p>
The ordinal number refers to the ordinal (left-to-right) position
of the output column. This feature makes it possible to define an
ordering on the basis of a column that does not have a unique
name. This is never absolutely necessary because it is always
possible to assign a name to an output column using the
<code class="literal">AS</code> clause.
</p><p>
It is also possible to use arbitrary expressions in the
<code class="literal">ORDER BY</code> clause, including columns that do not
appear in the <code class="command">SELECT</code> output list. Thus the
following statement is valid:
</p><pre class="programlisting">
SELECT name FROM distributors ORDER BY code;
</pre><p>
A limitation of this feature is that an <code class="literal">ORDER BY</code>
clause applying to the result of a <code class="literal">UNION</code>,
<code class="literal">INTERSECT</code>, or <code class="literal">EXCEPT</code> clause can only
specify an output column name or number, not an expression.
</p><p>
If an <code class="literal">ORDER BY</code> expression is a simple name that
matches both an output column name and an input column name,
<code class="literal">ORDER BY</code> will interpret it as the output column name.
This is the opposite of the choice that <code class="literal">GROUP BY</code> will
make in the same situation. This inconsistency is made to be
compatible with the SQL standard.
</p><p>
Optionally one can add the key word <code class="literal">ASC</code> (ascending) or
<code class="literal">DESC</code> (descending) after any expression in the
<code class="literal">ORDER BY</code> clause. If not specified, <code class="literal">ASC</code> is
assumed by default. Alternatively, a specific ordering operator
name can be specified in the <code class="literal">USING</code> clause.
An ordering operator must be a less-than or greater-than
member of some B-tree operator family.
<code class="literal">ASC</code> is usually equivalent to <code class="literal">USING <</code> and
<code class="literal">DESC</code> is usually equivalent to <code class="literal">USING ></code>.
(But the creator of a user-defined data type can define exactly what the
default sort ordering is, and it might correspond to operators with other
names.)
</p><p>
If <code class="literal">NULLS LAST</code> is specified, null values sort after all
non-null values; if <code class="literal">NULLS FIRST</code> is specified, null values
sort before all non-null values. If neither is specified, the default
behavior is <code class="literal">NULLS LAST</code> when <code class="literal">ASC</code> is specified
or implied, and <code class="literal">NULLS FIRST</code> when <code class="literal">DESC</code> is specified
(thus, the default is to act as though nulls are larger than non-nulls).
When <code class="literal">USING</code> is specified, the default nulls ordering depends
on whether the operator is a less-than or greater-than operator.
</p><p>
Note that ordering options apply only to the expression they follow;
for example <code class="literal">ORDER BY x, y DESC</code> does not mean
the same thing as <code class="literal">ORDER BY x DESC, y DESC</code>.
</p><p>
Character-string data is sorted according to the collation that applies
to the column being sorted. That can be overridden at need by including
a <code class="literal">COLLATE</code> clause in the
<em class="replaceable"><code>expression</code></em>, for example
<code class="literal">ORDER BY mycolumn COLLATE "en_US"</code>.
For more information see <a class="xref" href="sql-expressions.html#SQL-SYNTAX-COLLATE-EXPRS" title="4.2.10. Collation Expressions">Section 4.2.10</a> and
<a class="xref" href="collation.html" title="23.2. Collation Support">Section 23.2</a>.
</p></div><div class="refsect2" id="SQL-LIMIT"><h3><code class="literal">LIMIT</code> Clause</h3><p>
The <code class="literal">LIMIT</code> clause consists of two independent
sub-clauses:
</p><pre class="synopsis">
LIMIT { <em class="replaceable"><code>count</code></em> | ALL }
OFFSET <em class="replaceable"><code>start</code></em>
</pre><p>
<em class="replaceable"><code>count</code></em> specifies the
maximum number of rows to return, while <em class="replaceable"><code>start</code></em> specifies the number of rows
to skip before starting to return rows. When both are specified,
<em class="replaceable"><code>start</code></em> rows are skipped
before starting to count the <em class="replaceable"><code>count</code></em> rows to be returned.
</p><p>
If the <em class="replaceable"><code>count</code></em> expression
evaluates to NULL, it is treated as <code class="literal">LIMIT ALL</code>, i.e., no
limit. If <em class="replaceable"><code>start</code></em> evaluates
to NULL, it is treated the same as <code class="literal">OFFSET 0</code>.
</p><p>
SQL:2008 introduced a different syntax to achieve the same result,
which <span class="productname">PostgreSQL</span> also supports. It is:
</p><pre class="synopsis">
OFFSET <em class="replaceable"><code>start</code></em> { ROW | ROWS }
FETCH { FIRST | NEXT } [ <em class="replaceable"><code>count</code></em> ] { ROW | ROWS } ONLY
</pre><p>
In this syntax, the <em class="replaceable"><code>start</code></em>
or <em class="replaceable"><code>count</code></em> value is required by
the standard to be a literal constant, a parameter, or a variable name;
as a <span class="productname">PostgreSQL</span> extension, other expressions
are allowed, but will generally need to be enclosed in parentheses to avoid
ambiguity.
If <em class="replaceable"><code>count</code></em> is
omitted in a <code class="literal">FETCH</code> clause, it defaults to 1.
<code class="literal">ROW</code>
and <code class="literal">ROWS</code> as well as <code class="literal">FIRST</code>
and <code class="literal">NEXT</code> are noise words that don't influence
the effects of these clauses.
According to the standard, the <code class="literal">OFFSET</code> clause must come
before the <code class="literal">FETCH</code> clause if both are present; but
<span class="productname">PostgreSQL</span> is laxer and allows either order.
</p><p>
When using <code class="literal">LIMIT</code>, it is a good idea to use an
<code class="literal">ORDER BY</code> clause that constrains the result rows into a
unique order. Otherwise you will get an unpredictable subset of
the query's rows — you might be asking for the tenth through
twentieth rows, but tenth through twentieth in what ordering? You
don't know what ordering unless you specify <code class="literal">ORDER BY</code>.
</p><p>
The query planner takes <code class="literal">LIMIT</code> into account when
generating a query plan, so you are very likely to get different
plans (yielding different row orders) depending on what you use
for <code class="literal">LIMIT</code> and <code class="literal">OFFSET</code>. Thus, using
different <code class="literal">LIMIT</code>/<code class="literal">OFFSET</code> values to select
different subsets of a query result <span class="emphasis"><em>will give
inconsistent results</em></span> unless you enforce a predictable
result ordering with <code class="literal">ORDER BY</code>. This is not a bug; it
is an inherent consequence of the fact that SQL does not promise
to deliver the results of a query in any particular order unless
<code class="literal">ORDER BY</code> is used to constrain the order.
</p><p>
It is even possible for repeated executions of the same <code class="literal">LIMIT</code>
query to return different subsets of the rows of a table, if there
is not an <code class="literal">ORDER BY</code> to enforce selection of a deterministic
subset. Again, this is not a bug; determinism of the results is
simply not guaranteed in such a case.
</p></div><div class="refsect2" id="SQL-FOR-UPDATE-SHARE"><h3>The Locking Clause</h3><p>
<code class="literal">FOR UPDATE</code>, <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR SHARE</code>
and <code class="literal">FOR KEY SHARE</code>
are <em class="firstterm">locking clauses</em>; they affect how <code class="literal">SELECT</code>
locks rows as they are obtained from the table.
</p><p>
The locking clause has the general form
</p><pre class="synopsis">
FOR <em class="replaceable"><code>lock_strength</code></em> [ OF <em class="replaceable"><code>table_name</code></em> [, ...] ] [ NOWAIT | SKIP LOCKED ]
</pre><p>
where <em class="replaceable"><code>lock_strength</code></em> can be one of
</p><pre class="synopsis">
UPDATE
NO KEY UPDATE
SHARE
KEY SHARE
</pre><p>
</p><p>
For more information on each row-level lock mode, refer to
<a class="xref" href="explicit-locking.html#LOCKING-ROWS" title="13.3.2. Row-level Locks">Section 13.3.2</a>.
</p><p>
To prevent the operation from waiting for other transactions to commit,
use either the <code class="literal">NOWAIT</code> or <code class="literal">SKIP LOCKED</code>
option. With <code class="literal">NOWAIT</code>, the statement reports an error, rather
than waiting, if a selected row cannot be locked immediately.
With <code class="literal">SKIP LOCKED</code>, any selected rows that cannot be
immediately locked are skipped. Skipping locked rows provides an
inconsistent view of the data, so this is not suitable for general purpose
work, but can be used to avoid lock contention with multiple consumers
accessing a queue-like table.
Note that <code class="literal">NOWAIT</code> and <code class="literal">SKIP LOCKED</code> apply only
to the row-level lock(s) — the required <code class="literal">ROW SHARE</code>
table-level lock is still taken in the ordinary way (see
<a class="xref" href="mvcc.html" title="Chapter 13. Concurrency Control">Chapter 13</a>). You can use
<a class="xref" href="sql-lock.html" title="LOCK"><span class="refentrytitle">LOCK</span></a>
with the <code class="literal">NOWAIT</code> option first,
if you need to acquire the table-level lock without waiting.
</p><p>
If specific tables are named in a locking clause,
then only rows coming from those tables are locked; any other
tables used in the <code class="command">SELECT</code> are simply read as
usual. A locking
clause without a table list affects all tables used in the statement.
If a locking clause is
applied to a view or sub-query, it affects all tables used in
the view or sub-query.
However, these clauses
do not apply to <code class="literal">WITH</code> queries referenced by the primary query.
If you want row locking to occur within a <code class="literal">WITH</code> query, specify
a locking clause within the <code class="literal">WITH</code> query.
</p><p>
Multiple locking
clauses can be written if it is necessary to specify different locking
behavior for different tables. If the same table is mentioned (or
implicitly affected) by more than one locking clause,
then it is processed as if it was only specified by the strongest one.
Similarly, a table is processed
as <code class="literal">NOWAIT</code> if that is specified in any of the clauses
affecting it. Otherwise, it is processed
as <code class="literal">SKIP LOCKED</code> if that is specified in any of the
clauses affecting it.
</p><p>
The locking clauses cannot be
used in contexts where returned rows cannot be clearly identified with
individual table rows; for example they cannot be used with aggregation.
</p><p>
When a locking clause
appears at the top level of a <code class="command">SELECT</code> query, the rows that
are locked are exactly those that are returned by the query; in the
case of a join query, the rows locked are those that contribute to
returned join rows. In addition, rows that satisfied the query
conditions as of the query snapshot will be locked, although they
will not be returned if they were updated after the snapshot
and no longer satisfy the query conditions. If a
<code class="literal">LIMIT</code> is used, locking stops
once enough rows have been returned to satisfy the limit (but note that
rows skipped over by <code class="literal">OFFSET</code> will get locked). Similarly,
if a locking clause
is used in a cursor's query, only rows actually fetched or stepped past
by the cursor will be locked.
</p><p>
When a locking clause
appears in a sub-<code class="command">SELECT</code>, the rows locked are those
returned to the outer query by the sub-query. This might involve
fewer rows than inspection of the sub-query alone would suggest,
since conditions from the outer query might be used to optimize
execution of the sub-query. For example,
</p><pre class="programlisting">
SELECT * FROM (SELECT * FROM mytable FOR UPDATE) ss WHERE col1 = 5;
</pre><p>
will lock only rows having <code class="literal">col1 = 5</code>, even though that
condition is not textually within the sub-query.
</p><p>
Previous releases failed to preserve a lock which is upgraded by a later
savepoint. For example, this code:
</p><pre class="programlisting">
BEGIN;
SELECT * FROM mytable WHERE key = 1 FOR UPDATE;
SAVEPOINT s;
UPDATE mytable SET ... WHERE key = 1;
ROLLBACK TO s;
</pre><p>
would fail to preserve the <code class="literal">FOR UPDATE</code> lock after the
<code class="command">ROLLBACK TO</code>. This has been fixed in release 9.3.
</p><div class="caution"><h3 class="title">Caution</h3><p>
It is possible for a <code class="command">SELECT</code> command running at the <code class="literal">READ
COMMITTED</code> transaction isolation level and using <code class="literal">ORDER
BY</code> and a locking clause to return rows out of
order. This is because <code class="literal">ORDER BY</code> is applied first.
The command sorts the result, but might then block trying to obtain a lock
on one or more of the rows. Once the <code class="literal">SELECT</code> unblocks, some
of the ordering column values might have been modified, leading to those
rows appearing to be out of order (though they are in order in terms
of the original column values). This can be worked around at need by
placing the <code class="literal">FOR UPDATE/SHARE</code> clause in a sub-query,
for example
</p><pre class="programlisting">
SELECT * FROM (SELECT * FROM mytable FOR UPDATE) ss ORDER BY column1;
</pre><p>
Note that this will result in locking all rows of <code class="structname">mytable</code>,
whereas <code class="literal">FOR UPDATE</code> at the top level would lock only the
actually returned rows. This can make for a significant performance
difference, particularly if the <code class="literal">ORDER BY</code> is combined with
<code class="literal">LIMIT</code> or other restrictions. So this technique is recommended
only if concurrent updates of the ordering columns are expected and a
strictly sorted result is required.
</p><p>
At the <code class="literal">REPEATABLE READ</code> or <code class="literal">SERIALIZABLE</code>
transaction isolation level this would cause a serialization failure (with
a <code class="literal">SQLSTATE</code> of <code class="literal">'40001'</code>), so there is
no possibility of receiving rows out of order under these isolation levels.
</p></div></div><div class="refsect2" id="SQL-TABLE"><h3><code class="literal">TABLE</code> Command</h3><p>
The command
</p><pre class="programlisting">
TABLE <em class="replaceable"><code>name</code></em>
</pre><p>
is equivalent to
</p><pre class="programlisting">
SELECT * FROM <em class="replaceable"><code>name</code></em>
</pre><p>
It can be used as a top-level command or as a space-saving syntax
variant in parts of complex queries. Only the <code class="literal">WITH</code>,
<code class="literal">UNION</code>, <code class="literal">INTERSECT</code>, <code class="literal">EXCEPT</code>,
<code class="literal">ORDER BY</code>, <code class="literal">LIMIT</code>, <code class="literal">OFFSET</code>,
<code class="literal">FETCH</code> and <code class="literal">FOR</code> locking clauses can be used
with <code class="command">TABLE</code>; the <code class="literal">WHERE</code> clause and any form of
aggregation cannot
be used.
</p></div></div><div class="refsect1" id="id-1.9.3.171.9"><h2>Examples</h2><p>
To join the table <code class="literal">films</code> with the table
<code class="literal">distributors</code>:
</p><pre class="programlisting">
SELECT f.title, f.did, d.name, f.date_prod, f.kind
FROM distributors d, films f
WHERE f.did = d.did
title | did | name | date_prod | kind
-------------------+-----+--------------+------------+----------
The Third Man | 101 | British Lion | 1949-12-23 | Drama
The African Queen | 101 | British Lion | 1951-08-11 | Romantic
...
</pre><p>
</p><p>
To sum the column <code class="literal">len</code> of all films and group
the results by <code class="literal">kind</code>:
</p><pre class="programlisting">
SELECT kind, sum(len) AS total FROM films GROUP BY kind;
kind | total
----------+-------
Action | 07:34
Comedy | 02:58
Drama | 14:28
Musical | 06:42
Romantic | 04:38
</pre><p>
</p><p>
To sum the column <code class="literal">len</code> of all films, group
the results by <code class="literal">kind</code> and show those group totals
that are less than 5 hours:
</p><pre class="programlisting">
SELECT kind, sum(len) AS total
FROM films
GROUP BY kind
HAVING sum(len) < interval '5 hours';
kind | total
----------+-------
Comedy | 02:58
Romantic | 04:38
</pre><p>
</p><p>
The following two examples are identical ways of sorting the individual
results according to the contents of the second column
(<code class="literal">name</code>):
</p><pre class="programlisting">
SELECT * FROM distributors ORDER BY name;
SELECT * FROM distributors ORDER BY 2;
did | name
-----+------------------
109 | 20th Century Fox
110 | Bavaria Atelier
101 | British Lion
107 | Columbia
102 | Jean Luc Godard
113 | Luso films
104 | Mosfilm
103 | Paramount
106 | Toho
105 | United Artists
111 | Walt Disney
112 | Warner Bros.
108 | Westward
</pre><p>
</p><p>
The next example shows how to obtain the union of the tables
<code class="literal">distributors</code> and
<code class="literal">actors</code>, restricting the results to those that begin
with the letter W in each table. Only distinct rows are wanted, so the
key word <code class="literal">ALL</code> is omitted.
</p><pre class="programlisting">
distributors: actors:
did | name id | name
-----+-------------- ----+----------------
108 | Westward 1 | Woody Allen
111 | Walt Disney 2 | Warren Beatty
112 | Warner Bros. 3 | Walter Matthau
... ...
SELECT distributors.name
FROM distributors
WHERE distributors.name LIKE 'W%'
UNION
SELECT actors.name
FROM actors
WHERE actors.name LIKE 'W%';
name
----------------
Walt Disney
Walter Matthau
Warner Bros.
Warren Beatty
Westward
Woody Allen
</pre><p>
</p><p>
This example shows how to use a function in the <code class="literal">FROM</code>
clause, both with and without a column definition list:
</p><pre class="programlisting">
CREATE FUNCTION distributors(int) RETURNS SETOF distributors AS $$
SELECT * FROM distributors WHERE did = $1;
$$ LANGUAGE SQL;
SELECT * FROM distributors(111);
did | name
-----+-------------
111 | Walt Disney
CREATE FUNCTION distributors_2(int) RETURNS SETOF record AS $$
SELECT * FROM distributors WHERE did = $1;
$$ LANGUAGE SQL;
SELECT * FROM distributors_2(111) AS (f1 int, f2 text);
f1 | f2
-----+-------------
111 | Walt Disney
</pre><p>
</p><p>
Here is an example of a function with an ordinality column added:
</p><pre class="programlisting">
SELECT * FROM unnest(ARRAY['a','b','c','d','e','f']) WITH ORDINALITY;
unnest | ordinality
--------+----------
a | 1
b | 2
c | 3
d | 4
e | 5
f | 6
(6 rows)
</pre><p>
</p><p>
This example shows how to use a simple <code class="literal">WITH</code> clause:
</p><pre class="programlisting">
WITH t AS (
SELECT random() as x FROM generate_series(1, 3)
)
SELECT * FROM t
UNION ALL
SELECT * FROM t
x
--------------------
0.534150459803641
0.520092216785997
0.0735620250925422
0.534150459803641
0.520092216785997
0.0735620250925422
</pre><p>
Notice that the <code class="literal">WITH</code> query was evaluated only once,
so that we got two sets of the same three random values.
</p><p>
This example uses <code class="literal">WITH RECURSIVE</code> to find all
subordinates (direct or indirect) of the employee Mary, and their
level of indirectness, from a table that shows only direct
subordinates:
</p><pre class="programlisting">
WITH RECURSIVE employee_recursive(distance, employee_name, manager_name) AS (
SELECT 1, employee_name, manager_name
FROM employee
WHERE manager_name = 'Mary'
UNION ALL
SELECT er.distance + 1, e.employee_name, e.manager_name
FROM employee_recursive er, employee e
WHERE er.employee_name = e.manager_name
)
SELECT distance, employee_name FROM employee_recursive;
</pre><p>
Notice the typical form of recursive queries:
an initial condition, followed by <code class="literal">UNION</code>,
followed by the recursive part of the query. Be sure that the
recursive part of the query will eventually return no tuples, or
else the query will loop indefinitely. (See <a class="xref" href="queries-with.html" title="7.8. WITH Queries (Common Table Expressions)">Section 7.8</a>
for more examples.)
</p><p>
This example uses <code class="literal">LATERAL</code> to apply a set-returning function
<code class="function">get_product_names()</code> for each row of the
<code class="structname">manufacturers</code> table:
</p><pre class="programlisting">
SELECT m.name AS mname, pname
FROM manufacturers m, LATERAL get_product_names(m.id) pname;
</pre><p>
Manufacturers not currently having any products would not appear in the
result, since it is an inner join. If we wished to include the names of
such manufacturers in the result, we could do:
</p><pre class="programlisting">
SELECT m.name AS mname, pname
FROM manufacturers m LEFT JOIN LATERAL get_product_names(m.id) pname ON true;
</pre></div><div class="refsect1" id="id-1.9.3.171.10"><h2>Compatibility</h2><p>
Of course, the <code class="command">SELECT</code> statement is compatible
with the SQL standard. But there are some extensions and some
missing features.
</p><div class="refsect2" id="id-1.9.3.171.10.3"><h3>Omitted <code class="literal">FROM</code> Clauses</h3><p>
<span class="productname">PostgreSQL</span> allows one to omit the
<code class="literal">FROM</code> clause. It has a straightforward use to
compute the results of simple expressions:
</p><pre class="programlisting">
SELECT 2+2;
?column?
----------
4
</pre><p>
Some other <acronym class="acronym">SQL</acronym> databases cannot do this except
by introducing a dummy one-row table from which to do the
<code class="command">SELECT</code>.
</p><p>
Note that if a <code class="literal">FROM</code> clause is not specified,
the query cannot reference any database tables. For example, the
following query is invalid:
</p><pre class="programlisting">
SELECT distributors.* WHERE distributors.name = 'Westward';
</pre><p>
<span class="productname">PostgreSQL</span> releases prior to
8.1 would accept queries of this form, and add an implicit entry
to the query's <code class="literal">FROM</code> clause for each table
referenced by the query. This is no longer allowed.
</p></div><div class="refsect2" id="id-1.9.3.171.10.4"><h3>Empty <code class="literal">SELECT</code> Lists</h3><p>
The list of output expressions after <code class="literal">SELECT</code> can be
empty, producing a zero-column result table.
This is not valid syntax according to the SQL standard.
<span class="productname">PostgreSQL</span> allows it to be consistent with
allowing zero-column tables.
However, an empty list is not allowed when <code class="literal">DISTINCT</code> is used.
</p></div><div class="refsect2" id="id-1.9.3.171.10.5"><h3>Omitting the <code class="literal">AS</code> Key Word</h3><p>
In the SQL standard, the optional key word <code class="literal">AS</code> can be
omitted before an output column name whenever the new column name
is a valid column name (that is, not the same as any reserved
keyword). <span class="productname">PostgreSQL</span> is slightly more
restrictive: <code class="literal">AS</code> is required if the new column name
matches any keyword at all, reserved or not. Recommended practice is
to use <code class="literal">AS</code> or double-quote output column names, to prevent
any possible conflict against future keyword additions.
</p><p>
In <code class="literal">FROM</code> items, both the standard and
<span class="productname">PostgreSQL</span> allow <code class="literal">AS</code> to
be omitted before an alias that is an unreserved keyword. But
this is impractical for output column names, because of syntactic
ambiguities.
</p></div><div class="refsect2" id="id-1.9.3.171.10.6"><h3><code class="literal">ONLY</code> and Inheritance</h3><p>
The SQL standard requires parentheses around the table name when
writing <code class="literal">ONLY</code>, for example <code class="literal">SELECT * FROM ONLY
(tab1), ONLY (tab2) WHERE ...</code>. <span class="productname">PostgreSQL</span>
considers these parentheses to be optional.
</p><p>
<span class="productname">PostgreSQL</span> allows a trailing <code class="literal">*</code> to be written to
explicitly specify the non-<code class="literal">ONLY</code> behavior of including
child tables. The standard does not allow this.
</p><p>
(These points apply equally to all SQL commands supporting the
<code class="literal">ONLY</code> option.)
</p></div><div class="refsect2" id="id-1.9.3.171.10.7"><h3><code class="literal">TABLESAMPLE</code> Clause Restrictions</h3><p>
The <code class="literal">TABLESAMPLE</code> clause is currently accepted only on
regular tables and materialized views. According to the SQL standard
it should be possible to apply it to any <code class="literal">FROM</code> item.
</p></div><div class="refsect2" id="id-1.9.3.171.10.8"><h3>Function Calls in <code class="literal">FROM</code></h3><p>
<span class="productname">PostgreSQL</span> allows a function call to be
written directly as a member of the <code class="literal">FROM</code> list. In the SQL
standard it would be necessary to wrap such a function call in a
sub-<code class="command">SELECT</code>; that is, the syntax
<code class="literal">FROM <em class="replaceable"><code>func</code></em>(...) <em class="replaceable"><code>alias</code></em></code>
is approximately equivalent to
<code class="literal">FROM LATERAL (SELECT <em class="replaceable"><code>func</code></em>(...)) <em class="replaceable"><code>alias</code></em></code>.
Note that <code class="literal">LATERAL</code> is considered to be implicit; this is
because the standard requires <code class="literal">LATERAL</code> semantics for an
<code class="literal">UNNEST()</code> item in <code class="literal">FROM</code>.
<span class="productname">PostgreSQL</span> treats <code class="literal">UNNEST()</code> the
same as other set-returning functions.
</p></div><div class="refsect2" id="id-1.9.3.171.10.9"><h3>Namespace Available to <code class="literal">GROUP BY</code> and <code class="literal">ORDER BY</code></h3><p>
In the SQL-92 standard, an <code class="literal">ORDER BY</code> clause can
only use output column names or numbers, while a <code class="literal">GROUP
BY</code> clause can only use expressions based on input column
names. <span class="productname">PostgreSQL</span> extends each of
these clauses to allow the other choice as well (but it uses the
standard's interpretation if there is ambiguity).
<span class="productname">PostgreSQL</span> also allows both clauses to
specify arbitrary expressions. Note that names appearing in an
expression will always be taken as input-column names, not as
output-column names.
</p><p>
SQL:1999 and later use a slightly different definition which is not
entirely upward compatible with SQL-92.
In most cases, however, <span class="productname">PostgreSQL</span>
will interpret an <code class="literal">ORDER BY</code> or <code class="literal">GROUP
BY</code> expression the same way SQL:1999 does.
</p></div><div class="refsect2" id="id-1.9.3.171.10.10"><h3>Functional Dependencies</h3><p>
<span class="productname">PostgreSQL</span> recognizes functional dependency
(allowing columns to be omitted from <code class="literal">GROUP BY</code>) only when
a table's primary key is included in the <code class="literal">GROUP BY</code> list.
The SQL standard specifies additional conditions that should be
recognized.
</p></div><div class="refsect2" id="id-1.9.3.171.10.11"><h3><code class="literal">LIMIT</code> and <code class="literal">OFFSET</code></h3><p>
The clauses <code class="literal">LIMIT</code> and <code class="literal">OFFSET</code>
are <span class="productname">PostgreSQL</span>-specific syntax, also
used by <span class="productname">MySQL</span>. The SQL:2008 standard
has introduced the clauses <code class="literal">OFFSET ... FETCH {FIRST|NEXT}
...</code> for the same functionality, as shown above
in <a class="xref" href="sql-select.html#SQL-LIMIT" title="LIMIT Clause"><code class="literal">LIMIT</code> Clause</a>. This
syntax is also used by <span class="productname">IBM DB2</span>.
(Applications written for <span class="productname">Oracle</span>
frequently use a workaround involving the automatically
generated <code class="literal">rownum</code> column, which is not available in
PostgreSQL, to implement the effects of these clauses.)
</p></div><div class="refsect2" id="id-1.9.3.171.10.12"><h3><code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR UPDATE</code>, <code class="literal">FOR SHARE</code>, <code class="literal">FOR KEY SHARE</code></h3><p>
Although <code class="literal">FOR UPDATE</code> appears in the SQL standard, the
standard allows it only as an option of <code class="command">DECLARE CURSOR</code>.
<span class="productname">PostgreSQL</span> allows it in any <code class="command">SELECT</code>
query as well as in sub-<code class="command">SELECT</code>s, but this is an extension.
The <code class="literal">FOR NO KEY UPDATE</code>, <code class="literal">FOR SHARE</code> and
<code class="literal">FOR KEY SHARE</code> variants, as well as the <code class="literal">NOWAIT</code>
and <code class="literal">SKIP LOCKED</code> options, do not appear in the
standard.
</p></div><div class="refsect2" id="id-1.9.3.171.10.13"><h3>Data-Modifying Statements in <code class="literal">WITH</code></h3><p>
<span class="productname">PostgreSQL</span> allows <code class="command">INSERT</code>,
<code class="command">UPDATE</code>, and <code class="command">DELETE</code> to be used as <code class="literal">WITH</code>
queries. This is not found in the SQL standard.
</p></div><div class="refsect2" id="id-1.9.3.171.10.14"><h3>Nonstandard Clauses</h3><p>
<code class="literal">DISTINCT ON ( ... )</code> is an extension of the
SQL standard.
</p><p>
<code class="literal">ROWS FROM( ... )</code> is an extension of the SQL standard.
</p></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="sql-security-label.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="sql-commands.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="sql-selectinto.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">SECURITY LABEL </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> SELECT INTO</td></tr></table></div></body></html>
