Current File : //opt/alt/postgresql11/usr/share/doc/alt-postgresql11-9.2.24/html/xindex.html
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<HTML
><HEAD
><TITLE
>Interfacing Extensions To Indexes</TITLE
><META
NAME="GENERATOR"
CONTENT="Modular DocBook HTML Stylesheet Version 1.79"><LINK
REV="MADE"
HREF="mailto:pgsql-docs@postgresql.org"><LINK
REL="HOME"
TITLE="PostgreSQL 9.2.24 Documentation"
HREF="index.html"><LINK
REL="UP"
TITLE="Extending SQL"
HREF="extend.html"><LINK
REL="PREVIOUS"
TITLE="Operator Optimization Information"
HREF="xoper-optimization.html"><LINK
REL="NEXT"
TITLE="Packaging Related Objects into an Extension"
HREF="extend-extensions.html"><LINK
REL="STYLESHEET"
TYPE="text/css"
HREF="stylesheet.css"><META
HTTP-EQUIV="Content-Type"
CONTENT="text/html; charset=ISO-8859-1"><META
NAME="creation"
CONTENT="2017-11-06T22:43:11"></HEAD
><BODY
CLASS="SECT1"
><DIV
CLASS="NAVHEADER"
><TABLE
SUMMARY="Header navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TH
COLSPAN="5"
ALIGN="center"
VALIGN="bottom"
><A
HREF="index.html"
>PostgreSQL 9.2.24 Documentation</A
></TH
></TR
><TR
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
TITLE="Operator Optimization Information"
HREF="xoper-optimization.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
HREF="extend.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="60%"
ALIGN="center"
VALIGN="bottom"
>Chapter 35. Extending <ACRONYM
CLASS="ACRONYM"
>SQL</ACRONYM
></TD
><TD
WIDTH="20%"
ALIGN="right"
VALIGN="top"
><A
TITLE="Packaging Related Objects into an Extension"
HREF="extend-extensions.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
></TABLE
><HR
ALIGN="LEFT"
WIDTH="100%"></DIV
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="XINDEX"
>35.14. Interfacing Extensions To Indexes</A
></H1
><P
> The procedures described thus far let you define new types, new
functions, and new operators. However, we cannot yet define an
index on a column of a new data type. To do this, we must define an
<I
CLASS="FIRSTTERM"
>operator class</I
> for the new data type. Later in this
section, we will illustrate this concept in an example: a new
operator class for the B-tree index method that stores and sorts
complex numbers in ascending absolute value order.
</P
><P
> Operator classes can be grouped into <I
CLASS="FIRSTTERM"
>operator families</I
>
to show the relationships between semantically compatible classes.
When only a single data type is involved, an operator class is sufficient,
so we'll focus on that case first and then return to operator families.
</P
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-OPCLASS"
>35.14.1. Index Methods and Operator Classes</A
></H2
><P
> The <CODE
CLASS="CLASSNAME"
>pg_am</CODE
> table contains one row for every
index method (internally known as access method). Support for
regular access to tables is built into
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>, but all index methods are
described in <CODE
CLASS="CLASSNAME"
>pg_am</CODE
>. It is possible to add a
new index method by defining the required interface routines and
then creating a row in <CODE
CLASS="CLASSNAME"
>pg_am</CODE
> — but that is
beyond the scope of this chapter (see <A
HREF="indexam.html"
>Chapter 52</A
>).
</P
><P
> The routines for an index method do not directly know anything
about the data types that the index method will operate on.
Instead, an <I
CLASS="FIRSTTERM"
>operator
class</I
>
identifies the set of operations that the index method needs to use
to work with a particular data type. Operator classes are so
called because one thing they specify is the set of
<TT
CLASS="LITERAL"
>WHERE</TT
>-clause operators that can be used with an index
(i.e., can be converted into an index-scan qualification). An
operator class can also specify some <I
CLASS="FIRSTTERM"
>support
procedures</I
> that are needed by the internal operations of the
index method, but do not directly correspond to any
<TT
CLASS="LITERAL"
>WHERE</TT
>-clause operator that can be used with the index.
</P
><P
> It is possible to define multiple operator classes for the same
data type and index method. By doing this, multiple
sets of indexing semantics can be defined for a single data type.
For example, a B-tree index requires a sort ordering to be defined
for each data type it works on.
It might be useful for a complex-number data type
to have one B-tree operator class that sorts the data by complex
absolute value, another that sorts by real part, and so on.
Typically, one of the operator classes will be deemed most commonly
useful and will be marked as the default operator class for that
data type and index method.
</P
><P
> The same operator class name
can be used for several different index methods (for example, both B-tree
and hash index methods have operator classes named
<TT
CLASS="LITERAL"
>int4_ops</TT
>), but each such class is an independent
entity and must be defined separately.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-STRATEGIES"
>35.14.2. Index Method Strategies</A
></H2
><P
> The operators associated with an operator class are identified by
<SPAN
CLASS="QUOTE"
>"strategy numbers"</SPAN
>, which serve to identify the semantics of
each operator within the context of its operator class.
For example, B-trees impose a strict ordering on keys, lesser to greater,
and so operators like <SPAN
CLASS="QUOTE"
>"less than"</SPAN
> and <SPAN
CLASS="QUOTE"
>"greater than or equal
to"</SPAN
> are interesting with respect to a B-tree.
Because
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> allows the user to define operators,
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> cannot look at the name of an operator
(e.g., <TT
CLASS="LITERAL"
><</TT
> or <TT
CLASS="LITERAL"
>>=</TT
>) and tell what kind of
comparison it is. Instead, the index method defines a set of
<SPAN
CLASS="QUOTE"
>"strategies"</SPAN
>, which can be thought of as generalized operators.
Each operator class specifies which actual operator corresponds to each
strategy for a particular data type and interpretation of the index
semantics.
</P
><P
> The B-tree index method defines five strategies, shown in <A
HREF="xindex.html#XINDEX-BTREE-STRAT-TABLE"
>Table 35-2</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-BTREE-STRAT-TABLE"
></A
><P
><B
>Table 35-2. B-tree Strategies</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Operation</TH
><TH
>Strategy Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
>less than</TD
><TD
>1</TD
></TR
><TR
><TD
>less than or equal</TD
><TD
>2</TD
></TR
><TR
><TD
>equal</TD
><TD
>3</TD
></TR
><TR
><TD
>greater than or equal</TD
><TD
>4</TD
></TR
><TR
><TD
>greater than</TD
><TD
>5</TD
></TR
></TBODY
></TABLE
></DIV
><P
> Hash indexes support only equality comparisons, and so they use only one
strategy, shown in <A
HREF="xindex.html#XINDEX-HASH-STRAT-TABLE"
>Table 35-3</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-HASH-STRAT-TABLE"
></A
><P
><B
>Table 35-3. Hash Strategies</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Operation</TH
><TH
>Strategy Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
>equal</TD
><TD
>1</TD
></TR
></TBODY
></TABLE
></DIV
><P
> GiST indexes are more flexible: they do not have a fixed set of
strategies at all. Instead, the <SPAN
CLASS="QUOTE"
>"consistency"</SPAN
> support routine
of each particular GiST operator class interprets the strategy numbers
however it likes. As an example, several of the built-in GiST index
operator classes index two-dimensional geometric objects, providing
the <SPAN
CLASS="QUOTE"
>"R-tree"</SPAN
> strategies shown in
<A
HREF="xindex.html#XINDEX-RTREE-STRAT-TABLE"
>Table 35-4</A
>. Four of these are true
two-dimensional tests (overlaps, same, contains, contained by);
four of them consider only the X direction; and the other four
provide the same tests in the Y direction.
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-RTREE-STRAT-TABLE"
></A
><P
><B
>Table 35-4. GiST Two-Dimensional <SPAN
CLASS="QUOTE"
>"R-tree"</SPAN
> Strategies</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Operation</TH
><TH
>Strategy Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
>strictly left of</TD
><TD
>1</TD
></TR
><TR
><TD
>does not extend to right of</TD
><TD
>2</TD
></TR
><TR
><TD
>overlaps</TD
><TD
>3</TD
></TR
><TR
><TD
>does not extend to left of</TD
><TD
>4</TD
></TR
><TR
><TD
>strictly right of</TD
><TD
>5</TD
></TR
><TR
><TD
>same</TD
><TD
>6</TD
></TR
><TR
><TD
>contains</TD
><TD
>7</TD
></TR
><TR
><TD
>contained by</TD
><TD
>8</TD
></TR
><TR
><TD
>does not extend above</TD
><TD
>9</TD
></TR
><TR
><TD
>strictly below</TD
><TD
>10</TD
></TR
><TR
><TD
>strictly above</TD
><TD
>11</TD
></TR
><TR
><TD
>does not extend below</TD
><TD
>12</TD
></TR
></TBODY
></TABLE
></DIV
><P
> SP-GiST indexes are similar to GiST indexes in flexibility: they don't have
a fixed set of strategies. Instead the support routines of each operator
class interpret the strategy numbers according to the operator class's
definition. As an example, the strategy numbers used by the built-in
operator classes for points are shown in <A
HREF="xindex.html#XINDEX-SPGIST-POINT-STRAT-TABLE"
>Table 35-5</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-SPGIST-POINT-STRAT-TABLE"
></A
><P
><B
>Table 35-5. SP-GiST Point Strategies</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Operation</TH
><TH
>Strategy Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
>strictly left of</TD
><TD
>1</TD
></TR
><TR
><TD
>strictly right of</TD
><TD
>5</TD
></TR
><TR
><TD
>same</TD
><TD
>6</TD
></TR
><TR
><TD
>contained by</TD
><TD
>8</TD
></TR
><TR
><TD
>strictly below</TD
><TD
>10</TD
></TR
><TR
><TD
>strictly above</TD
><TD
>11</TD
></TR
></TBODY
></TABLE
></DIV
><P
> GIN indexes are similar to GiST and SP-GiST indexes, in that they don't
have a fixed set of strategies either. Instead the support routines of
each operator class interpret the strategy numbers according to the
operator class's definition. As an example, the strategy numbers used by
the built-in operator classes for arrays are shown in
<A
HREF="xindex.html#XINDEX-GIN-ARRAY-STRAT-TABLE"
>Table 35-6</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-GIN-ARRAY-STRAT-TABLE"
></A
><P
><B
>Table 35-6. GIN Array Strategies</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Operation</TH
><TH
>Strategy Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
>overlap</TD
><TD
>1</TD
></TR
><TR
><TD
>contains</TD
><TD
>2</TD
></TR
><TR
><TD
>is contained by</TD
><TD
>3</TD
></TR
><TR
><TD
>equal</TD
><TD
>4</TD
></TR
></TBODY
></TABLE
></DIV
><P
> Notice that all the operators listed above return Boolean values. In
practice, all operators defined as index method search operators must
return type <TT
CLASS="TYPE"
>boolean</TT
>, since they must appear at the top
level of a <TT
CLASS="LITERAL"
>WHERE</TT
> clause to be used with an index.
(Some index access methods also support <I
CLASS="FIRSTTERM"
>ordering operators</I
>,
which typically don't return Boolean values; that feature is discussed
in <A
HREF="xindex.html#XINDEX-ORDERING-OPS"
>Section 35.14.7</A
>.)
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-SUPPORT"
>35.14.3. Index Method Support Routines</A
></H2
><P
> Strategies aren't usually enough information for the system to figure
out how to use an index. In practice, the index methods require
additional support routines in order to work. For example, the B-tree
index method must be able to compare two keys and determine whether one
is greater than, equal to, or less than the other. Similarly, the
hash index method must be able to compute hash codes for key values.
These operations do not correspond to operators used in qualifications in
SQL commands; they are administrative routines used by
the index methods, internally.
</P
><P
> Just as with strategies, the operator class identifies which specific
functions should play each of these roles for a given data type and
semantic interpretation. The index method defines the set
of functions it needs, and the operator class identifies the correct
functions to use by assigning them to the <SPAN
CLASS="QUOTE"
>"support function numbers"</SPAN
>
specified by the index method.
</P
><P
> B-trees require a single support function, and allow a second one to be
supplied at the operator class author's option, as shown in <A
HREF="xindex.html#XINDEX-BTREE-SUPPORT-TABLE"
>Table 35-7</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-BTREE-SUPPORT-TABLE"
></A
><P
><B
>Table 35-7. B-tree Support Functions</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Function</TH
><TH
>Support Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
> Compare two keys and return an integer less than zero, zero, or
greater than zero, indicating whether the first key is less than,
equal to, or greater than the second
</TD
><TD
>1</TD
></TR
><TR
><TD
> Return the addresses of C-callable sort support function(s),
as documented in <TT
CLASS="FILENAME"
>utils/sortsupport.h</TT
> (optional)
</TD
><TD
>2</TD
></TR
></TBODY
></TABLE
></DIV
><P
> Hash indexes require one support function, shown in <A
HREF="xindex.html#XINDEX-HASH-SUPPORT-TABLE"
>Table 35-8</A
>.
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-HASH-SUPPORT-TABLE"
></A
><P
><B
>Table 35-8. Hash Support Functions</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><THEAD
><TR
><TH
>Function</TH
><TH
>Support Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
>Compute the hash value for a key</TD
><TD
>1</TD
></TR
></TBODY
></TABLE
></DIV
><P
> GiST indexes require seven support functions, with an optional eighth, as
shown in <A
HREF="xindex.html#XINDEX-GIST-SUPPORT-TABLE"
>Table 35-9</A
>.
(For more information see <A
HREF="gist.html"
>Chapter 53</A
>.)
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-GIST-SUPPORT-TABLE"
></A
><P
><B
>Table 35-9. GiST Support Functions</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><COL><THEAD
><TR
><TH
>Function</TH
><TH
>Description</TH
><TH
>Support Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
><CODE
CLASS="FUNCTION"
>consistent</CODE
></TD
><TD
>determine whether key satisfies the
query qualifier</TD
><TD
>1</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>union</CODE
></TD
><TD
>compute union of a set of keys</TD
><TD
>2</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>compress</CODE
></TD
><TD
>compute a compressed representation of a key or value
to be indexed</TD
><TD
>3</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>decompress</CODE
></TD
><TD
>compute a decompressed representation of a
compressed key</TD
><TD
>4</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>penalty</CODE
></TD
><TD
>compute penalty for inserting new key into subtree
with given subtree's key</TD
><TD
>5</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>picksplit</CODE
></TD
><TD
>determine which entries of a page are to be moved
to the new page and compute the union keys for resulting pages</TD
><TD
>6</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>equal</CODE
></TD
><TD
>compare two keys and return true if they are equal</TD
><TD
>7</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>distance</CODE
></TD
><TD
>determine distance from key to query value (optional)</TD
><TD
>8</TD
></TR
></TBODY
></TABLE
></DIV
><P
> SP-GiST indexes require five support functions, as
shown in <A
HREF="xindex.html#XINDEX-SPGIST-SUPPORT-TABLE"
>Table 35-10</A
>.
(For more information see <A
HREF="spgist.html"
>Chapter 54</A
>.)
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-SPGIST-SUPPORT-TABLE"
></A
><P
><B
>Table 35-10. SP-GiST Support Functions</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><COL><THEAD
><TR
><TH
>Function</TH
><TH
>Description</TH
><TH
>Support Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
><CODE
CLASS="FUNCTION"
>config</CODE
></TD
><TD
>provide basic information about the operator class</TD
><TD
>1</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>choose</CODE
></TD
><TD
>determine how to insert a new value into an inner tuple</TD
><TD
>2</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>picksplit</CODE
></TD
><TD
>determine how to partition a set of values</TD
><TD
>3</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>inner_consistent</CODE
></TD
><TD
>determine which sub-partitions need to be searched for a
query</TD
><TD
>4</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>leaf_consistent</CODE
></TD
><TD
>determine whether key satisfies the
query qualifier</TD
><TD
>5</TD
></TR
></TBODY
></TABLE
></DIV
><P
> GIN indexes require four support functions, with an optional fifth, as
shown in <A
HREF="xindex.html#XINDEX-GIN-SUPPORT-TABLE"
>Table 35-11</A
>.
(For more information see <A
HREF="gin.html"
>Chapter 55</A
>.)
</P
><DIV
CLASS="TABLE"
><A
NAME="XINDEX-GIN-SUPPORT-TABLE"
></A
><P
><B
>Table 35-11. GIN Support Functions</B
></P
><TABLE
BORDER="1"
CLASS="CALSTABLE"
><COL><COL><COL><THEAD
><TR
><TH
>Function</TH
><TH
>Description</TH
><TH
>Support Number</TH
></TR
></THEAD
><TBODY
><TR
><TD
><CODE
CLASS="FUNCTION"
>compare</CODE
></TD
><TD
> compare two keys and return an integer less than zero, zero,
or greater than zero, indicating whether the first key is less than,
equal to, or greater than the second
</TD
><TD
>1</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>extractValue</CODE
></TD
><TD
>extract keys from a value to be indexed</TD
><TD
>2</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>extractQuery</CODE
></TD
><TD
>extract keys from a query condition</TD
><TD
>3</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>consistent</CODE
></TD
><TD
>determine whether value matches query condition</TD
><TD
>4</TD
></TR
><TR
><TD
><CODE
CLASS="FUNCTION"
>comparePartial</CODE
></TD
><TD
> compare partial key from
query and key from index, and return an integer less than zero, zero,
or greater than zero, indicating whether GIN should ignore this index
entry, treat the entry as a match, or stop the index scan (optional)
</TD
><TD
>5</TD
></TR
></TBODY
></TABLE
></DIV
><P
> Unlike search operators, support functions return whichever data
type the particular index method expects; for example in the case
of the comparison function for B-trees, a signed integer. The number
and types of the arguments to each support function are likewise
dependent on the index method. For B-tree and hash the comparison and
hashing support functions take the same input data types as do the
operators included in the operator class, but this is not the case for
most GiST, SP-GiST, and GIN support functions.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-EXAMPLE"
>35.14.4. An Example</A
></H2
><P
> Now that we have seen the ideas, here is the promised example of
creating a new operator class.
(You can find a working copy of this example in
<TT
CLASS="FILENAME"
>src/tutorial/complex.c</TT
> and
<TT
CLASS="FILENAME"
>src/tutorial/complex.sql</TT
> in the source
distribution.)
The operator class encapsulates
operators that sort complex numbers in absolute value order, so we
choose the name <TT
CLASS="LITERAL"
>complex_abs_ops</TT
>. First, we need
a set of operators. The procedure for defining operators was
discussed in <A
HREF="xoper.html"
>Section 35.12</A
>. For an operator class on
B-trees, the operators we require are:
<P
></P
></P><UL
COMPACT="COMPACT"
><LI
><SPAN
>absolute-value less-than (strategy 1)</SPAN
></LI
><LI
><SPAN
>absolute-value less-than-or-equal (strategy 2)</SPAN
></LI
><LI
><SPAN
>absolute-value equal (strategy 3)</SPAN
></LI
><LI
><SPAN
>absolute-value greater-than-or-equal (strategy 4)</SPAN
></LI
><LI
><SPAN
>absolute-value greater-than (strategy 5)</SPAN
></LI
></UL
><P>
</P
><P
> The least error-prone way to define a related set of comparison operators
is to write the B-tree comparison support function first, and then write the
other functions as one-line wrappers around the support function. This
reduces the odds of getting inconsistent results for corner cases.
Following this approach, we first write:
</P><PRE
CLASS="PROGRAMLISTING"
>#define Mag(c) ((c)->x*(c)->x + (c)->y*(c)->y)
static int
complex_abs_cmp_internal(Complex *a, Complex *b)
{
double amag = Mag(a),
bmag = Mag(b);
if (amag < bmag)
return -1;
if (amag > bmag)
return 1;
return 0;
}</PRE
><P>
Now the less-than function looks like:
</P><PRE
CLASS="PROGRAMLISTING"
>PG_FUNCTION_INFO_V1(complex_abs_lt);
Datum
complex_abs_lt(PG_FUNCTION_ARGS)
{
Complex *a = (Complex *) PG_GETARG_POINTER(0);
Complex *b = (Complex *) PG_GETARG_POINTER(1);
PG_RETURN_BOOL(complex_abs_cmp_internal(a, b) < 0);
}</PRE
><P>
The other four functions differ only in how they compare the internal
function's result to zero.
</P
><P
> Next we declare the functions and the operators based on the functions
to SQL:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE FUNCTION complex_abs_lt(complex, complex) RETURNS bool
AS '<TT
CLASS="REPLACEABLE"
><I
>filename</I
></TT
>', 'complex_abs_lt'
LANGUAGE C IMMUTABLE STRICT;
CREATE OPERATOR < (
leftarg = complex, rightarg = complex, procedure = complex_abs_lt,
commutator = > , negator = >= ,
restrict = scalarltsel, join = scalarltjoinsel
);</PRE
><P>
It is important to specify the correct commutator and negator operators,
as well as suitable restriction and join selectivity
functions, otherwise the optimizer will be unable to make effective
use of the index. Note that the less-than, equal, and
greater-than cases should use different selectivity functions.
</P
><P
> Other things worth noting are happening here:
<P
></P
></P><UL
><LI
><P
> There can only be one operator named, say, <TT
CLASS="LITERAL"
>=</TT
>
and taking type <TT
CLASS="TYPE"
>complex</TT
> for both operands. In this
case we don't have any other operator <TT
CLASS="LITERAL"
>=</TT
> for
<TT
CLASS="TYPE"
>complex</TT
>, but if we were building a practical data
type we'd probably want <TT
CLASS="LITERAL"
>=</TT
> to be the ordinary
equality operation for complex numbers (and not the equality of
the absolute values). In that case, we'd need to use some other
operator name for <CODE
CLASS="FUNCTION"
>complex_abs_eq</CODE
>.
</P
></LI
><LI
><P
> Although <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> can cope with
functions having the same SQL name as long as they have different
argument data types, C can only cope with one global function
having a given name. So we shouldn't name the C function
something simple like <TT
CLASS="FILENAME"
>abs_eq</TT
>. Usually it's
a good practice to include the data type name in the C function
name, so as not to conflict with functions for other data types.
</P
></LI
><LI
><P
> We could have made the SQL name
of the function <TT
CLASS="FILENAME"
>abs_eq</TT
>, relying on
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> to distinguish it by
argument data types from any other SQL function of the same name.
To keep the example simple, we make the function have the same
names at the C level and SQL level.
</P
></LI
></UL
><P>
</P
><P
> The next step is the registration of the support routine required
by B-trees. The example C code that implements this is in the same
file that contains the operator functions. This is how we declare
the function:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE FUNCTION complex_abs_cmp(complex, complex)
RETURNS integer
AS '<TT
CLASS="REPLACEABLE"
><I
>filename</I
></TT
>'
LANGUAGE C IMMUTABLE STRICT;</PRE
><P>
</P
><P
> Now that we have the required operators and support routine,
we can finally create the operator class:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE OPERATOR CLASS complex_abs_ops
DEFAULT FOR TYPE complex USING btree AS
OPERATOR 1 < ,
OPERATOR 2 <= ,
OPERATOR 3 = ,
OPERATOR 4 >= ,
OPERATOR 5 > ,
FUNCTION 1 complex_abs_cmp(complex, complex);</PRE
><P>
</P
><P
> And we're done! It should now be possible to create
and use B-tree indexes on <TT
CLASS="TYPE"
>complex</TT
> columns.
</P
><P
> We could have written the operator entries more verbosely, as in:
</P><PRE
CLASS="PROGRAMLISTING"
> OPERATOR 1 < (complex, complex) ,</PRE
><P>
but there is no need to do so when the operators take the same data type
we are defining the operator class for.
</P
><P
> The above example assumes that you want to make this new operator class the
default B-tree operator class for the <TT
CLASS="TYPE"
>complex</TT
> data type.
If you don't, just leave out the word <TT
CLASS="LITERAL"
>DEFAULT</TT
>.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-OPFAMILY"
>35.14.5. Operator Classes and Operator Families</A
></H2
><P
> So far we have implicitly assumed that an operator class deals with
only one data type. While there certainly can be only one data type in
a particular index column, it is often useful to index operations that
compare an indexed column to a value of a different data type. Also,
if there is use for a cross-data-type operator in connection with an
operator class, it is often the case that the other data type has a
related operator class of its own. It is helpful to make the connections
between related classes explicit, because this can aid the planner in
optimizing SQL queries (particularly for B-tree operator classes, since
the planner contains a great deal of knowledge about how to work with them).
</P
><P
> To handle these needs, <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>
uses the concept of an <I
CLASS="FIRSTTERM"
>operator
family</I
>.
An operator family contains one or more operator classes, and can also
contain indexable operators and corresponding support functions that
belong to the family as a whole but not to any single class within the
family. We say that such operators and functions are <SPAN
CLASS="QUOTE"
>"loose"</SPAN
>
within the family, as opposed to being bound into a specific class.
Typically each operator class contains single-data-type operators
while cross-data-type operators are loose in the family.
</P
><P
> All the operators and functions in an operator family must have compatible
semantics, where the compatibility requirements are set by the index
method. You might therefore wonder why bother to single out particular
subsets of the family as operator classes; and indeed for many purposes
the class divisions are irrelevant and the family is the only interesting
grouping. The reason for defining operator classes is that they specify
how much of the family is needed to support any particular index.
If there is an index using an operator class, then that operator class
cannot be dropped without dropping the index — but other parts of
the operator family, namely other operator classes and loose operators,
could be dropped. Thus, an operator class should be specified to contain
the minimum set of operators and functions that are reasonably needed
to work with an index on a specific data type, and then related but
non-essential operators can be added as loose members of the operator
family.
</P
><P
> As an example, <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> has a built-in
B-tree operator family <TT
CLASS="LITERAL"
>integer_ops</TT
>, which includes operator
classes <TT
CLASS="LITERAL"
>int8_ops</TT
>, <TT
CLASS="LITERAL"
>int4_ops</TT
>, and
<TT
CLASS="LITERAL"
>int2_ops</TT
> for indexes on <TT
CLASS="TYPE"
>bigint</TT
> (<TT
CLASS="TYPE"
>int8</TT
>),
<TT
CLASS="TYPE"
>integer</TT
> (<TT
CLASS="TYPE"
>int4</TT
>), and <TT
CLASS="TYPE"
>smallint</TT
> (<TT
CLASS="TYPE"
>int2</TT
>)
columns respectively. The family also contains cross-data-type comparison
operators allowing any two of these types to be compared, so that an index
on one of these types can be searched using a comparison value of another
type. The family could be duplicated by these definitions:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE OPERATOR FAMILY integer_ops USING btree;
CREATE OPERATOR CLASS int8_ops
DEFAULT FOR TYPE int8 USING btree FAMILY integer_ops AS
-- standard int8 comparisons
OPERATOR 1 < ,
OPERATOR 2 <= ,
OPERATOR 3 = ,
OPERATOR 4 >= ,
OPERATOR 5 > ,
FUNCTION 1 btint8cmp(int8, int8) ,
FUNCTION 2 btint8sortsupport(internal) ;
CREATE OPERATOR CLASS int4_ops
DEFAULT FOR TYPE int4 USING btree FAMILY integer_ops AS
-- standard int4 comparisons
OPERATOR 1 < ,
OPERATOR 2 <= ,
OPERATOR 3 = ,
OPERATOR 4 >= ,
OPERATOR 5 > ,
FUNCTION 1 btint4cmp(int4, int4) ,
FUNCTION 2 btint4sortsupport(internal) ;
CREATE OPERATOR CLASS int2_ops
DEFAULT FOR TYPE int2 USING btree FAMILY integer_ops AS
-- standard int2 comparisons
OPERATOR 1 < ,
OPERATOR 2 <= ,
OPERATOR 3 = ,
OPERATOR 4 >= ,
OPERATOR 5 > ,
FUNCTION 1 btint2cmp(int2, int2) ,
FUNCTION 2 btint2sortsupport(internal) ;
ALTER OPERATOR FAMILY integer_ops USING btree ADD
-- cross-type comparisons int8 vs int2
OPERATOR 1 < (int8, int2) ,
OPERATOR 2 <= (int8, int2) ,
OPERATOR 3 = (int8, int2) ,
OPERATOR 4 >= (int8, int2) ,
OPERATOR 5 > (int8, int2) ,
FUNCTION 1 btint82cmp(int8, int2) ,
-- cross-type comparisons int8 vs int4
OPERATOR 1 < (int8, int4) ,
OPERATOR 2 <= (int8, int4) ,
OPERATOR 3 = (int8, int4) ,
OPERATOR 4 >= (int8, int4) ,
OPERATOR 5 > (int8, int4) ,
FUNCTION 1 btint84cmp(int8, int4) ,
-- cross-type comparisons int4 vs int2
OPERATOR 1 < (int4, int2) ,
OPERATOR 2 <= (int4, int2) ,
OPERATOR 3 = (int4, int2) ,
OPERATOR 4 >= (int4, int2) ,
OPERATOR 5 > (int4, int2) ,
FUNCTION 1 btint42cmp(int4, int2) ,
-- cross-type comparisons int4 vs int8
OPERATOR 1 < (int4, int8) ,
OPERATOR 2 <= (int4, int8) ,
OPERATOR 3 = (int4, int8) ,
OPERATOR 4 >= (int4, int8) ,
OPERATOR 5 > (int4, int8) ,
FUNCTION 1 btint48cmp(int4, int8) ,
-- cross-type comparisons int2 vs int8
OPERATOR 1 < (int2, int8) ,
OPERATOR 2 <= (int2, int8) ,
OPERATOR 3 = (int2, int8) ,
OPERATOR 4 >= (int2, int8) ,
OPERATOR 5 > (int2, int8) ,
FUNCTION 1 btint28cmp(int2, int8) ,
-- cross-type comparisons int2 vs int4
OPERATOR 1 < (int2, int4) ,
OPERATOR 2 <= (int2, int4) ,
OPERATOR 3 = (int2, int4) ,
OPERATOR 4 >= (int2, int4) ,
OPERATOR 5 > (int2, int4) ,
FUNCTION 1 btint24cmp(int2, int4) ;</PRE
><P>
Notice that this definition <SPAN
CLASS="QUOTE"
>"overloads"</SPAN
> the operator strategy and
support function numbers: each number occurs multiple times within the
family. This is allowed so long as each instance of a
particular number has distinct input data types. The instances that have
both input types equal to an operator class's input type are the
primary operators and support functions for that operator class,
and in most cases should be declared as part of the operator class rather
than as loose members of the family.
</P
><P
> In a B-tree operator family, all the operators in the family must sort
compatibly, meaning that the transitive laws hold across all the data types
supported by the family: <SPAN
CLASS="QUOTE"
>"if A = B and B = C, then A = C"</SPAN
>,
and <SPAN
CLASS="QUOTE"
>"if A < B and B < C, then A < C"</SPAN
>. Moreover, implicit
or binary coercion casts between types represented in the operator family
must not change the associated sort ordering. For each
operator in the family there must be a support function having the same
two input data types as the operator. It is recommended that a family be
complete, i.e., for each combination of data types, all operators are
included. Each operator class should include just the non-cross-type
operators and support function for its data type.
</P
><P
> To build a multiple-data-type hash operator family, compatible hash
support functions must be created for each data type supported by the
family. Here compatibility means that the functions are guaranteed to
return the same hash code for any two values that are considered equal
by the family's equality operators, even when the values are of different
types. This is usually difficult to accomplish when the types have
different physical representations, but it can be done in some cases.
Furthermore, casting a value from one data type represented in the operator
family to another data type also represented in the operator family via
an implicit or binary coercion cast must not change the computed hash value.
Notice that there is only one support function per data type, not one
per equality operator. It is recommended that a family be complete, i.e.,
provide an equality operator for each combination of data types.
Each operator class should include just the non-cross-type equality
operator and the support function for its data type.
</P
><P
> GiST, SP-GiST, and GIN indexes do not have any explicit notion of
cross-data-type operations. The set of operators supported is just
whatever the primary support functions for a given operator class can
handle.
</P
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> Prior to <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> 8.3, there was no concept
of operator families, and so any cross-data-type operators intended to be
used with an index had to be bound directly into the index's operator
class. While this approach still works, it is deprecated because it
makes an index's dependencies too broad, and because the planner can
handle cross-data-type comparisons more effectively when both data types
have operators in the same operator family.
</P
></BLOCKQUOTE
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-OPCLASS-DEPENDENCIES"
>35.14.6. System Dependencies on Operator Classes</A
></H2
><P
> <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> uses operator classes to infer the
properties of operators in more ways than just whether they can be used
with indexes. Therefore, you might want to create operator classes
even if you have no intention of indexing any columns of your data type.
</P
><P
> In particular, there are SQL features such as <TT
CLASS="LITERAL"
>ORDER BY</TT
> and
<TT
CLASS="LITERAL"
>DISTINCT</TT
> that require comparison and sorting of values.
To implement these features on a user-defined data type,
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> looks for the default B-tree operator
class for the data type. The <SPAN
CLASS="QUOTE"
>"equals"</SPAN
> member of this operator
class defines the system's notion of equality of values for
<TT
CLASS="LITERAL"
>GROUP BY</TT
> and <TT
CLASS="LITERAL"
>DISTINCT</TT
>, and the sort ordering
imposed by the operator class defines the default <TT
CLASS="LITERAL"
>ORDER BY</TT
>
ordering.
</P
><P
> Comparison of arrays of user-defined types also relies on the semantics
defined by the default B-tree operator class.
</P
><P
> If there is no default B-tree operator class for a data type, the system
will look for a default hash operator class. But since that kind of
operator class only provides equality, in practice it is only enough
to support array equality.
</P
><P
> When there is no default operator class for a data type, you will get
errors like <SPAN
CLASS="QUOTE"
>"could not identify an ordering operator"</SPAN
> if you
try to use these SQL features with the data type.
</P
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Note: </B
> In <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> versions before 7.4,
sorting and grouping operations would implicitly use operators named
<TT
CLASS="LITERAL"
>=</TT
>, <TT
CLASS="LITERAL"
><</TT
>, and <TT
CLASS="LITERAL"
>></TT
>. The new
behavior of relying on default operator classes avoids having to make
any assumption about the behavior of operators with particular names.
</P
></BLOCKQUOTE
></DIV
><P
> Another important point is that an operator that
appears in a hash operator family is a candidate for hash joins,
hash aggregation, and related optimizations. The hash operator family
is essential here since it identifies the hash function(s) to use.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-ORDERING-OPS"
>35.14.7. Ordering Operators</A
></H2
><P
> Some index access methods (currently, only GiST) support the concept of
<I
CLASS="FIRSTTERM"
>ordering operators</I
>. What we have been discussing so far
are <I
CLASS="FIRSTTERM"
>search operators</I
>. A search operator is one for which
the index can be searched to find all rows satisfying
<TT
CLASS="LITERAL"
>WHERE</TT
>
<TT
CLASS="REPLACEABLE"
><I
>indexed_column</I
></TT
>
<TT
CLASS="REPLACEABLE"
><I
>operator</I
></TT
>
<TT
CLASS="REPLACEABLE"
><I
>constant</I
></TT
>.
Note that nothing is promised about the order in which the matching rows
will be returned. In contrast, an ordering operator does not restrict the
set of rows that can be returned, but instead determines their order.
An ordering operator is one for which the index can be scanned to return
rows in the order represented by
<TT
CLASS="LITERAL"
>ORDER BY</TT
>
<TT
CLASS="REPLACEABLE"
><I
>indexed_column</I
></TT
>
<TT
CLASS="REPLACEABLE"
><I
>operator</I
></TT
>
<TT
CLASS="REPLACEABLE"
><I
>constant</I
></TT
>.
The reason for defining ordering operators that way is that it supports
nearest-neighbor searches, if the operator is one that measures distance.
For example, a query like
</P><PRE
CLASS="PROGRAMLISTING"
>SELECT * FROM places ORDER BY location <-> point '(101,456)' LIMIT 10;</PRE
><P>
finds the ten places closest to a given target point. A GiST index
on the location column can do this efficiently because
<TT
CLASS="LITERAL"
><-></TT
> is an ordering operator.
</P
><P
> While search operators have to return Boolean results, ordering operators
usually return some other type, such as float or numeric for distances.
This type is normally not the same as the data type being indexed.
To avoid hard-wiring assumptions about the behavior of different data
types, the definition of an ordering operator is required to name
a B-tree operator family that specifies the sort ordering of the result
data type. As was stated in the previous section, B-tree operator families
define <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>'s notion of ordering, so
this is a natural representation. Since the point <TT
CLASS="LITERAL"
><-></TT
>
operator returns <TT
CLASS="TYPE"
>float8</TT
>, it could be specified in an operator
class creation command like this:
</P><PRE
CLASS="PROGRAMLISTING"
>OPERATOR 15 <-> (point, point) FOR ORDER BY float_ops</PRE
><P>
where <TT
CLASS="LITERAL"
>float_ops</TT
> is the built-in operator family that includes
operations on <TT
CLASS="TYPE"
>float8</TT
>. This declaration states that the index
is able to return rows in order of increasing values of the
<TT
CLASS="LITERAL"
><-></TT
> operator.
</P
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="XINDEX-OPCLASS-FEATURES"
>35.14.8. Special Features of Operator Classes</A
></H2
><P
> There are two special features of operator classes that we have
not discussed yet, mainly because they are not useful
with the most commonly used index methods.
</P
><P
> Normally, declaring an operator as a member of an operator class
(or family) means that the index method can retrieve exactly the set of rows
that satisfy a <TT
CLASS="LITERAL"
>WHERE</TT
> condition using the operator. For example:
</P><PRE
CLASS="PROGRAMLISTING"
>SELECT * FROM table WHERE integer_column < 4;</PRE
><P>
can be satisfied exactly by a B-tree index on the integer column.
But there are cases where an index is useful as an inexact guide to
the matching rows. For example, if a GiST index stores only bounding boxes
for geometric objects, then it cannot exactly satisfy a <TT
CLASS="LITERAL"
>WHERE</TT
>
condition that tests overlap between nonrectangular objects such as
polygons. Yet we could use the index to find objects whose bounding
box overlaps the bounding box of the target object, and then do the
exact overlap test only on the objects found by the index. If this
scenario applies, the index is said to be <SPAN
CLASS="QUOTE"
>"lossy"</SPAN
> for the
operator. Lossy index searches are implemented by having the index
method return a <I
CLASS="FIRSTTERM"
>recheck</I
> flag when a row might or might
not really satisfy the query condition. The core system will then
test the original query condition on the retrieved row to see whether
it should be returned as a valid match. This approach works if
the index is guaranteed to return all the required rows, plus perhaps
some additional rows, which can be eliminated by performing the original
operator invocation. The index methods that support lossy searches
(currently, GiST, SP-GiST and GIN) allow the support functions of individual
operator classes to set the recheck flag, and so this is essentially an
operator-class feature.
</P
><P
> Consider again the situation where we are storing in the index only
the bounding box of a complex object such as a polygon. In this
case there's not much value in storing the whole polygon in the index
entry — we might as well store just a simpler object of type
<TT
CLASS="TYPE"
>box</TT
>. This situation is expressed by the <TT
CLASS="LITERAL"
>STORAGE</TT
>
option in <TT
CLASS="COMMAND"
>CREATE OPERATOR CLASS</TT
>: we'd write something like:
</P><PRE
CLASS="PROGRAMLISTING"
>CREATE OPERATOR CLASS polygon_ops
DEFAULT FOR TYPE polygon USING gist AS
...
STORAGE box;</PRE
><P>
At present, only the GiST and GIN index methods support a
<TT
CLASS="LITERAL"
>STORAGE</TT
> type that's different from the column data type.
The GiST <CODE
CLASS="FUNCTION"
>compress</CODE
> and <CODE
CLASS="FUNCTION"
>decompress</CODE
> support
routines must deal with data-type conversion when <TT
CLASS="LITERAL"
>STORAGE</TT
>
is used. In GIN, the <TT
CLASS="LITERAL"
>STORAGE</TT
> type identifies the type of
the <SPAN
CLASS="QUOTE"
>"key"</SPAN
> values, which normally is different from the type
of the indexed column — for example, an operator class for
integer-array columns might have keys that are just integers. The
GIN <CODE
CLASS="FUNCTION"
>extractValue</CODE
> and <CODE
CLASS="FUNCTION"
>extractQuery</CODE
> support
routines are responsible for extracting keys from indexed values.
</P
></DIV
></DIV
><DIV
CLASS="NAVFOOTER"
><HR
ALIGN="LEFT"
WIDTH="100%"><TABLE
SUMMARY="Footer navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
><A
HREF="xoper-optimization.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="index.html"
ACCESSKEY="H"
>Home</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
><A
HREF="extend-extensions.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
>Operator Optimization Information</TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="extend.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
>Packaging Related Objects into an Extension</TD
></TR
></TABLE
></DIV
></BODY
></HTML
>
Mini Shell