diff options
Diffstat (limited to 'doc/src/sgml/xoper.sgml')
| -rw-r--r-- | doc/src/sgml/xoper.sgml | 19 |
1 files changed, 18 insertions, 1 deletions
diff --git a/doc/src/sgml/xoper.sgml b/doc/src/sgml/xoper.sgml index ab5f38be751..80e6793397e 100644 --- a/doc/src/sgml/xoper.sgml +++ b/doc/src/sgml/xoper.sgml @@ -1,5 +1,5 @@ <!-- -$Header: /cvsroot/pgsql/doc/src/sgml/xoper.sgml,v 1.26 2003/08/31 17:32:21 petere Exp $ +$Header: /cvsroot/pgsql/doc/src/sgml/xoper.sgml,v 1.27 2003/10/21 22:51:14 tgl Exp $ --> <sect1 id="xoper"> @@ -121,6 +121,23 @@ SELECT (a + b) AS c FROM test_complex; </para> <para> + It's critical to provide commutator information for operators that + will be used in indexes and join clauses, because this allows the + query optimizer to <quote>flip around</> such a clause to the forms + needed for different plan types. For example, consider a query with + a WHERE clause like <literal>tab1.x = tab2.y</>, where <literal>tab1.x</> + and <literal>tab2.y</> are of a user-defined type, and suppose that + <literal>tab2.y</> is indexed. The optimizer cannot generate an + indexscan unless it can determine how to flip the clause around to + <literal>tab2.y = tab1.x</>, because the indexscan machinery expects + to see the indexed column on the left of the operator it is given. + <ProductName>PostgreSQL</ProductName> will <emphasis>not</> simply + assume that this is a valid transformation --- the definer of the + <literal>=</> operator must specify that it is valid, by marking the + operator with commutator information. + </para> + + <para> When you are defining a self-commutative operator, you just do it. When you are defining a pair of commutative operators, things are a little trickier: how can the first one to be defined refer to the |