????
Your IP : 13.58.172.13
<?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>2.6. Joins Between Tables</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="tutorial-select.html" title="2.5. Querying a Table" /><link rel="next" href="tutorial-agg.html" title="2.7. Aggregate Functions" /></head><body id="docContent" class="container-fluid col-10"><div class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">2.6. Joins Between Tables</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="tutorial-select.html" title="2.5. Querying a Table">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="tutorial-sql.html" title="Chapter 2. The SQL Language">Up</a></td><th width="60%" align="center">Chapter 2. The <acronym class="acronym">SQL</acronym> Language</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 16.3 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="tutorial-agg.html" title="2.7. Aggregate Functions">Next</a></td></tr></table><hr /></div><div class="sect1" id="TUTORIAL-JOIN"><div class="titlepage"><div><div><h2 class="title" style="clear: both">2.6. Joins Between Tables <a href="#TUTORIAL-JOIN" class="id_link">#</a></h2></div></div></div><a id="id-1.4.4.7.2" class="indexterm"></a><p>
Thus far, our queries have only accessed one table at a time.
Queries can access multiple tables at once, or access the same
table in such a way that multiple rows of the table are being
processed at the same time. Queries that access multiple tables
(or multiple instances of the same table) at one time are called
<em class="firstterm">join</em> queries. They combine rows from one table
with rows from a second table, with an expression specifying which rows
are to be paired. For example, to return all the weather records together
with the location of the associated city, the database needs to compare
the <code class="structfield">city</code>
column of each row of the <code class="structname">weather</code> table with the
<code class="structfield">name</code> column of all rows in the <code class="structname">cities</code>
table, and select the pairs of rows where these values match.<a href="#ftn.id-1.4.4.7.3.6" class="footnote"><sup class="footnote" id="id-1.4.4.7.3.6">[4]</sup></a>
This would be accomplished by the following query:
</p><pre class="programlisting">
SELECT * FROM weather JOIN cities ON city = name;
</pre><p>
</p><pre class="screen">
city | temp_lo | temp_hi | prcp | date | name | location
---------------+---------+---------+------+------------+---------------+-----------
San Francisco | 46 | 50 | 0.25 | 1994-11-27 | San Francisco | (-194,53)
San Francisco | 43 | 57 | 0 | 1994-11-29 | San Francisco | (-194,53)
(2 rows)
</pre><p>
</p><p>
Observe two things about the result set:
</p><div class="itemizedlist"><ul class="itemizedlist" style="list-style-type: disc; "><li class="listitem"><p>
There is no result row for the city of Hayward. This is
because there is no matching entry in the
<code class="structname">cities</code> table for Hayward, so the join
ignores the unmatched rows in the <code class="structname">weather</code> table. We will see
shortly how this can be fixed.
</p></li><li class="listitem"><p>
There are two columns containing the city name. This is
correct because the lists of columns from the
<code class="structname">weather</code> and
<code class="structname">cities</code> tables are concatenated. In
practice this is undesirable, though, so you will probably want
to list the output columns explicitly rather than using
<code class="literal">*</code>:
</p><pre class="programlisting">
SELECT city, temp_lo, temp_hi, prcp, date, location
FROM weather JOIN cities ON city = name;
</pre><p>
</p></li></ul></div><p>
</p><p>
Since the columns all had different names, the parser
automatically found which table they belong to. If there
were duplicate column names in the two tables you'd need to
<em class="firstterm">qualify</em> the column names to show which one you
meant, as in:
</p><pre class="programlisting">
SELECT weather.city, weather.temp_lo, weather.temp_hi,
weather.prcp, weather.date, cities.location
FROM weather JOIN cities ON weather.city = cities.name;
</pre><p>
It is widely considered good style to qualify all column names
in a join query, so that the query won't fail if a duplicate
column name is later added to one of the tables.
</p><p>
Join queries of the kind seen thus far can also be written in this
form:
</p><pre class="programlisting">
SELECT *
FROM weather, cities
WHERE city = name;
</pre><p>
This syntax pre-dates the <code class="literal">JOIN</code>/<code class="literal">ON</code>
syntax, which was introduced in SQL-92. The tables are simply listed in
the <code class="literal">FROM</code> clause, and the comparison expression is added
to the <code class="literal">WHERE</code> clause. The results from this older
implicit syntax and the newer explicit
<code class="literal">JOIN</code>/<code class="literal">ON</code> syntax are identical. But
for a reader of the query, the explicit syntax makes its meaning easier to
understand: The join condition is introduced by its own key word whereas
previously the condition was mixed into the <code class="literal">WHERE</code>
clause together with other conditions.
</p><a id="id-1.4.4.7.7" class="indexterm"></a><p>
Now we will figure out how we can get the Hayward records back in.
What we want the query to do is to scan the
<code class="structname">weather</code> table and for each row to find the
matching <code class="structname">cities</code> row(s). If no matching row is
found we want some <span class="quote">“<span class="quote">empty values</span>”</span> to be substituted
for the <code class="structname">cities</code> table's columns. This kind
of query is called an <em class="firstterm">outer join</em>. (The
joins we have seen so far are <em class="firstterm">inner joins</em>.)
The command looks like this:
</p><pre class="programlisting">
SELECT *
FROM weather LEFT OUTER JOIN cities ON weather.city = cities.name;
</pre><p>
</p><pre class="screen">
city | temp_lo | temp_hi | prcp | date | name | location
---------------+---------+---------+------+------------+---------------+-----------
Hayward | 37 | 54 | | 1994-11-29 | |
San Francisco | 46 | 50 | 0.25 | 1994-11-27 | San Francisco | (-194,53)
San Francisco | 43 | 57 | 0 | 1994-11-29 | San Francisco | (-194,53)
(3 rows)
</pre><p>
This query is called a <em class="firstterm">left outer
join</em> because the table mentioned on the left of the
join operator will have each of its rows in the output at least
once, whereas the table on the right will only have those rows
output that match some row of the left table. When outputting a
left-table row for which there is no right-table match, empty (null)
values are substituted for the right-table columns.
</p><p><strong>Exercise: </strong>
There are also right outer joins and full outer joins. Try to
find out what those do.
</p><a id="id-1.4.4.7.10" class="indexterm"></a><a id="id-1.4.4.7.11" class="indexterm"></a><p>
We can also join a table against itself. This is called a
<em class="firstterm">self join</em>. As an example, suppose we wish
to find all the weather records that are in the temperature range
of other weather records. So we need to compare the
<code class="structfield">temp_lo</code> and <code class="structfield">temp_hi</code> columns of
each <code class="structname">weather</code> row to the
<code class="structfield">temp_lo</code> and
<code class="structfield">temp_hi</code> columns of all other
<code class="structname">weather</code> rows. We can do this with the
following query:
</p><pre class="programlisting">
SELECT w1.city, w1.temp_lo AS low, w1.temp_hi AS high,
w2.city, w2.temp_lo AS low, w2.temp_hi AS high
FROM weather w1 JOIN weather w2
ON w1.temp_lo < w2.temp_lo AND w1.temp_hi > w2.temp_hi;
</pre><p>
</p><pre class="screen">
city | low | high | city | low | high
---------------+-----+------+---------------+-----+------
San Francisco | 43 | 57 | San Francisco | 46 | 50
Hayward | 37 | 54 | San Francisco | 46 | 50
(2 rows)
</pre><p>
Here we have relabeled the weather table as <code class="literal">w1</code> and
<code class="literal">w2</code> to be able to distinguish the left and right side
of the join. You can also use these kinds of aliases in other
queries to save some typing, e.g.:
</p><pre class="programlisting">
SELECT *
FROM weather w JOIN cities c ON w.city = c.name;
</pre><p>
You will encounter this style of abbreviating quite frequently.
</p><div class="footnotes"><br /><hr style="width:100; text-align:left;margin-left: 0" /><div id="ftn.id-1.4.4.7.3.6" class="footnote"><p><a href="#id-1.4.4.7.3.6" class="para"><sup class="para">[4] </sup></a>
This is only a conceptual model. The join is usually performed
in a more efficient manner than actually comparing each possible
pair of rows, but this is invisible to the user.
</p></div></div></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="tutorial-select.html" title="2.5. Querying a Table">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="tutorial-sql.html" title="Chapter 2. The SQL Language">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="tutorial-agg.html" title="2.7. Aggregate Functions">Next</a></td></tr><tr><td width="40%" align="left" valign="top">2.5. Querying a Table </td><td width="20%" align="center"><a accesskey="h" href="index.html" title="PostgreSQL 16.3 Documentation">Home</a></td><td width="40%" align="right" valign="top"> 2.7. Aggregate Functions</td></tr></table></div></body></html>