path: root/doc/src/sgml/gist.sgml

<!-- doc/src/sgml/gist.sgml -->

<chapter id="gist">
<title>GiST Indexes</title>

   <indexterm>
    <primary>index</primary>
    <secondary>GiST</secondary>
   </indexterm>

<sect1 id="gist-intro">
 <title>Introduction</title>

 <para>
   <acronym>GiST</acronym> stands for Generalized Search Tree.  It is a
   balanced, tree-structured access method, that acts as a base template in
   which to implement arbitrary indexing schemes. B-trees, R-trees and many
   other indexing schemes can be implemented in <acronym>GiST</acronym>.
 </para>

 <para>
  One advantage of <acronym>GiST</acronym> is that it allows the development
  of custom data types with the appropriate access methods, by
  an expert in the domain of the data type, rather than a database expert.
 </para>

  <para>
    Some of the information here is derived from the University of California
    at Berkeley's GiST Indexing Project
    <ulink url="http://gist.cs.berkeley.edu/">web site</ulink> and
    Marcel Kornacker's thesis,
    <ulink url="http://www.sai.msu.su/~megera/postgres/gist/papers/concurrency/access-methods-for-next-generation.pdf.gz">
    Access Methods for Next-Generation Database Systems</ulink>.
    The <acronym>GiST</acronym>
    implementation in <productname>PostgreSQL</productname> is primarily
    maintained by Teodor Sigaev and Oleg Bartunov, and there is more
    information on their
    <ulink url="http://www.sai.msu.su/~megera/postgres/gist/">web site</ulink>.
  </para>

</sect1>

<sect1 id="gist-builtin-opclasses">
 <title>Built-in Operator Classes</title>

 <para>
  The core <productname>PostgreSQL</productname> distribution
  includes the <acronym>GiST</acronym> operator classes shown in
  <xref linkend="gist-builtin-opclasses-table"/>.
  (Some of the optional modules described in <xref linkend="contrib"/>
  provide additional <acronym>GiST</acronym> operator classes.)
 </para>

  <table id="gist-builtin-opclasses-table">
   <title>Built-in <acronym>GiST</acronym> Operator Classes</title>
   <tgroup cols="3">
     <colspec colname="col1" colwidth="2*"/>
     <colspec colname="col2" colwidth="3*"/>
     <colspec colname="col3" colwidth="2*"/>
    <thead>
     <row>
      <entry>Name</entry>
      <entry>Indexable Operators</entry>
      <entry>Ordering Operators</entry>
     </row>
    </thead>
    <tbody>
     <row>
      <entry valign="middle" morerows="11"><literal>box_ops</literal></entry>
      <entry><literal>&lt;&lt; (box, box)</literal></entry>
      <entry valign="middle" morerows="11"><literal>&lt;-&gt; (box, point)</literal></entry>
     </row>
     <row><entry><literal>&amp;&lt; (box, box)</literal></entry></row>
     <row><entry><literal>&amp;&amp; (box, box)</literal></entry></row>
     <row><entry><literal>&amp;&gt; (box, box)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (box, box)</literal></entry></row>
     <row><entry><literal>~= (box, box)</literal></entry></row>
     <row><entry><literal>@&gt; (box, box)</literal></entry></row>
     <row><entry><literal>&lt;@ (box, box)</literal></entry></row>
     <row><entry><literal>&amp;&lt;| (box, box)</literal></entry></row>
     <row><entry><literal>&lt;&lt;| (box, box)</literal></entry></row>
     <row><entry><literal>|&gt;&gt; (box, box)</literal></entry></row>
     <row><entry><literal>|&amp;&gt; (box, box)</literal></entry></row>

     <row>
      <entry valign="middle" morerows="11"><literal>circle_ops</literal></entry>
      <entry><literal>&lt;&lt; (circle, circle)</literal></entry>
      <entry valign="middle" morerows="11"><literal>&lt;-&gt; (circle, point)</literal></entry>
     </row>
     <row><entry><literal>&amp;&lt; (circle, circle)</literal></entry></row>
     <row><entry><literal>&amp;&gt; (circle, circle)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (circle, circle)</literal></entry></row>
     <row><entry><literal>&lt;@ (circle, circle)</literal></entry></row>
     <row><entry><literal>@&gt; (circle, circle)</literal></entry></row>
     <row><entry><literal>~= (circle, circle)</literal></entry></row>
     <row><entry><literal>&amp;&amp; (circle, circle)</literal></entry></row>
     <row><entry><literal>|&gt;&gt; (circle, circle)</literal></entry></row>
     <row><entry><literal>&lt;&lt;| (circle, circle)</literal></entry></row>
     <row><entry><literal>&amp;&lt;| (circle, circle)</literal></entry></row>
     <row><entry><literal>|&amp;&gt; (circle, circle)</literal></entry></row>

     <row>
      <entry valign="middle" morerows="10"><literal>inet_ops</literal></entry>
      <entry><literal>&lt;&lt; (inet, inet)</literal></entry>
      <entry valign="middle" morerows="10"></entry>
     </row>
     <row><entry><literal>&lt;&lt;= (inet, inet)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (inet, inet)</literal></entry></row>
     <row><entry><literal>&gt;&gt;= (inet, inet)</literal></entry></row>
     <row><entry><literal>= (inet, inet)</literal></entry></row>
     <row><entry><literal>&lt;&gt; (inet, inet)</literal></entry></row>
     <row><entry><literal>&lt; (inet, inet)</literal></entry></row>
     <row><entry><literal>&lt;= (inet, inet)</literal></entry></row>
     <row><entry><literal>&gt; (inet, inet)</literal></entry></row>
     <row><entry><literal>&gt;= (inet, inet)</literal></entry></row>
     <row><entry><literal>&amp;&amp; (inet, inet)</literal></entry></row>

     <row>
      <entry valign="middle" morerows="17"><literal>multirange_ops</literal></entry>
      <entry><literal>= (anymultirange, anymultirange)</literal></entry>
      <entry valign="middle" morerows="17"></entry>
     </row>
     <row><entry><literal>&amp;&amp; (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>&amp;&amp; (anymultirange, anyrange)</literal></entry></row>
     <row><entry><literal>@&gt; (anymultirange, anyelement)</literal></entry></row>
     <row><entry><literal>@&gt; (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>@&gt; (anymultirange, anyrange)</literal></entry></row>
     <row><entry><literal>&lt;@ (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>&lt;@ (anymultirange, anyrange)</literal></entry></row>
     <row><entry><literal>&lt;&lt; (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>&lt;&lt; (anymultirange, anyrange)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (anymultirange, anyrange)</literal></entry></row>
     <row><entry><literal>&amp;&lt; (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>&amp;&lt; (anymultirange, anyrange)</literal></entry></row>
     <row><entry><literal>&amp;&gt; (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>&amp;&gt; (anymultirange, anyrange)</literal></entry></row>
     <row><entry><literal>-|- (anymultirange, anymultirange)</literal></entry></row>
     <row><entry><literal>-|- (anymultirange, anyrange)</literal></entry></row>

     <row>
      <entry valign="middle" morerows="7"><literal>point_ops</literal></entry>
      <entry><literal>|&gt;&gt; (point, point)</literal></entry>
      <entry valign="middle" morerows="7"><literal>&lt;-&gt; (point, point)</literal></entry>
     </row>
     <row><entry><literal>&lt;&lt; (point, point)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (point, point)</literal></entry></row>
     <row><entry><literal>&lt;&lt;| (point, point)</literal></entry></row>
     <row><entry><literal>~= (point, point)</literal></entry></row>
     <row><entry><literal>&lt;@ (point, box)</literal></entry></row>
     <row><entry><literal>&lt;@ (point, polygon)</literal></entry></row>
     <row><entry><literal>&lt;@ (point, circle)</literal></entry></row>

     <row>
      <entry valign="middle" morerows="11"><literal>poly_ops</literal></entry>
      <entry><literal>&lt;&lt; (polygon, polygon)</literal></entry>
      <entry valign="middle" morerows="11"><literal>&lt;-&gt; (polygon, point)</literal></entry>
     </row>
     <row><entry><literal>&amp;&lt; (polygon, polygon)</literal></entry></row>
     <row><entry><literal>&amp;&gt; (polygon, polygon)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (polygon, polygon)</literal></entry></row>
     <row><entry><literal>&lt;@ (polygon, polygon)</literal></entry></row>
     <row><entry><literal>@&gt; (polygon, polygon)</literal></entry></row>
     <row><entry><literal>~= (polygon, polygon)</literal></entry></row>
     <row><entry><literal>&amp;&amp; (polygon, polygon)</literal></entry></row>
     <row><entry><literal>&lt;&lt;| (polygon, polygon)</literal></entry></row>
     <row><entry><literal>&amp;&lt;| (polygon, polygon)</literal></entry></row>
     <row><entry><literal>|&amp;&gt; (polygon, polygon)</literal></entry></row>
     <row><entry><literal>|&gt;&gt; (polygon, polygon)</literal></entry></row>

     <row>
      <entry valign="middle" morerows="17"><literal>range_ops</literal></entry>
      <entry><literal>= (anyrange, anyrange)</literal></entry>
      <entry valign="middle" morerows="17"></entry>
     </row>
     <row><entry><literal>&amp;&amp; (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>&amp;&amp; (anyrange, anymultirange)</literal></entry></row>
     <row><entry><literal>@&gt; (anyrange, anyelement)</literal></entry></row>
     <row><entry><literal>@&gt; (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>@&gt; (anyrange, anymultirange)</literal></entry></row>
     <row><entry><literal>&lt;@ (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>&lt;@ (anyrange, anymultirange)</literal></entry></row>
     <row><entry><literal>&lt;&lt; (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>&lt;&lt; (anyrange, anymultirange)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>&gt;&gt; (anyrange, anymultirange)</literal></entry></row>
     <row><entry><literal>&amp;&lt; (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>&amp;&lt; (anyrange, anymultirange)</literal></entry></row>
     <row><entry><literal>&amp;&gt; (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>&amp;&gt; (anyrange, anymultirange)</literal></entry></row>
     <row><entry><literal>-|- (anyrange, anyrange)</literal></entry></row>
     <row><entry><literal>-|- (anyrange, anymultirange)</literal></entry></row>

     <row>
      <entry valign="middle" morerows="1"><literal>tsquery_ops</literal></entry>
      <entry><literal>&lt;@ (tsquery, tsquery)</literal></entry>
      <entry valign="middle" morerows="1"></entry>
     </row>
     <row><entry><literal>@&gt; (tsquery, tsquery)</literal></entry></row>
     <row>
      <entry valign="middle"><literal>tsvector_ops</literal></entry>
      <entry><literal>@@ (tsvector, tsquery)</literal></entry>
      <entry></entry>
     </row>
    </tbody>
   </tgroup>
  </table>

 <para>
  For historical reasons, the <literal>inet_ops</literal> operator class is
  not the default class for types <type>inet</type> and <type>cidr</type>.
  To use it, mention the class name in <command>CREATE INDEX</command>,
  for example
<programlisting>
CREATE INDEX ON my_table USING GIST (my_inet_column inet_ops);
</programlisting>
 </para>

</sect1>

<sect1 id="gist-extensibility">
 <title>Extensibility</title>

 <para>
   Traditionally, implementing a new index access method meant a lot of
   difficult work.  It was necessary to understand the inner workings of the
   database, such as the lock manager and Write-Ahead Log.  The
   <acronym>GiST</acronym> interface has a high level of abstraction,
   requiring the access method implementer only to implement the semantics of
   the data type being accessed.  The <acronym>GiST</acronym> layer itself
   takes care of concurrency, logging and searching the tree structure.
 </para>

 <para>
   This extensibility should not be confused with the extensibility of the
   other standard search trees in terms of the data they can handle.  For
   example, <productname>PostgreSQL</productname> supports extensible B-trees
   and hash indexes. That means that you can use
   <productname>PostgreSQL</productname> to build a B-tree or hash over any
   data type you want. But B-trees only support range predicates
   (<literal>&lt;</literal>, <literal>=</literal>, <literal>&gt;</literal>),
   and hash indexes only support equality queries.
 </para>

 <para>
   So if you index, say, an image collection with a
   <productname>PostgreSQL</productname> B-tree, you can only issue queries
   such as <quote>is imagex equal to imagey</quote>, <quote>is imagex less
   than imagey</quote> and <quote>is imagex greater than imagey</quote>.
   Depending on how you define <quote>equals</quote>, <quote>less than</quote>
   and <quote>greater than</quote> in this context, this could be useful.
   However, by using a <acronym>GiST</acronym> based index, you could create
   ways to ask domain-specific questions, perhaps <quote>find all images of
   horses</quote> or <quote>find all over-exposed images</quote>.
 </para>

 <para>
   All it takes to get a <acronym>GiST</acronym> access method up and running
   is to implement several user-defined methods, which define the behavior of
   keys in the tree. Of course these methods have to be pretty fancy to
   support fancy queries, but for all the standard queries (B-trees,
   R-trees, etc.) they're relatively straightforward. In short,
   <acronym>GiST</acronym> combines extensibility along with generality, code
   reuse, and a clean interface.
  </para>

 <para>
   There are five methods that an index operator class for
   <acronym>GiST</acronym> must provide, and six that are optional.
   Correctness of the index is ensured
   by proper implementation of the <function>same</function>, <function>consistent</function>
   and <function>union</function> methods, while efficiency (size and speed) of the
   index will depend on the <function>penalty</function> and <function>picksplit</function>
   methods.
   Two optional methods are <function>compress</function> and
   <function>decompress</function>, which allow an index to have internal tree data of
   a different type than the data it indexes. The leaves are to be of the
   indexed data type, while the other tree nodes can be of any C struct (but
   you still have to follow <productname>PostgreSQL</productname> data type rules here,
   see about <literal>varlena</literal> for variable sized data). If the tree's
   internal data type exists at the SQL level, the <literal>STORAGE</literal> option
   of the <command>CREATE OPERATOR CLASS</command> command can be used.
   The optional eighth method is <function>distance</function>, which is needed
   if the operator class wishes to support ordered scans (nearest-neighbor
   searches). The optional ninth method <function>fetch</function> is needed if the
   operator class wishes to support index-only scans, except when the
   <function>compress</function> method is omitted. The optional tenth method
   <function>options</function> is needed if the operator class has
   user-specified parameters.
   The optional eleventh method <function>sortsupport</function> is used to
   speed up building a <acronym>GiST</acronym> index.
 </para>

 <variablelist>
    <varlistentry>
     <term><function>consistent</function></term>
     <listitem>
      <para>
       Given an index entry <literal>p</literal> and a query value <literal>q</literal>,
       this function determines whether the index entry is
       <quote>consistent</quote> with the query; that is, could the predicate
       <quote><replaceable>indexed_column</replaceable>
       <replaceable>indexable_operator</replaceable> <literal>q</literal></quote> be true for
       any row represented by the index entry?  For a leaf index entry this is
       equivalent to testing the indexable condition, while for an internal
       tree node this determines whether it is necessary to scan the subtree
       of the index represented by the tree node.  When the result is
       <literal>true</literal>, a <literal>recheck</literal> flag must also be returned.
       This indicates whether the predicate is certainly true or only possibly
       true.  If <literal>recheck</literal> = <literal>false</literal> then the index has
       tested the predicate condition exactly, whereas if <literal>recheck</literal>
       = <literal>true</literal> the row is only a candidate match.  In that case the
       system will automatically evaluate the
       <replaceable>indexable_operator</replaceable> against the actual row value to see
       if it is really a match.  This convention allows
       <acronym>GiST</acronym> to support both lossless and lossy index
       structures.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_consistent(internal, data_type, smallint, oid, internal)
RETURNS bool
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_consistent);

Datum
my_consistent(PG_FUNCTION_ARGS)
{
    GISTENTRY  *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
    data_type  *query = PG_GETARG_DATA_TYPE_P(1);
    StrategyNumber strategy = (StrategyNumber) PG_GETARG_UINT16(2);
    /* Oid subtype = PG_GETARG_OID(3); */
    bool       *recheck = (bool *) PG_GETARG_POINTER(4);
    data_type  *key = DatumGetDataType(entry-&gt;key);
    bool        retval;

    /*
     * determine return value as a function of strategy, key and query.
     *
     * Use GIST_LEAF(entry) to know where you're called in the index tree,
     * which comes handy when supporting the = operator for example (you could
     * check for non empty union() in non-leaf nodes and equality in leaf
     * nodes).
     */

    *recheck = true;        /* or false if check is exact */

    PG_RETURN_BOOL(retval);
}
</programlisting>

       Here, <varname>key</varname> is an element in the index and <varname>query</varname>
       the value being looked up in the index. The <literal>StrategyNumber</literal>
       parameter indicates which operator of your operator class is being
       applied &mdash; it matches one of the operator numbers in the
       <command>CREATE OPERATOR CLASS</command> command.
      </para>

      <para>
       Depending on which operators you have included in the class, the data
       type of <varname>query</varname> could vary with the operator, since it will
       be whatever type is on the right-hand side of the operator, which might
       be different from the indexed data type appearing on the left-hand side.
       (The above code skeleton assumes that only one type is possible; if
       not, fetching the <varname>query</varname> argument value would have to depend
       on the operator.)  It is recommended that the SQL declaration of
       the <function>consistent</function> function use the opclass's indexed data
       type for the <varname>query</varname> argument, even though the actual type
       might be something else depending on the operator.
      </para>

     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>union</function></term>
     <listitem>
      <para>
       This method consolidates information in the tree.  Given a set of
       entries, this function generates a new index entry that represents
       all the given entries.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_union(internal, internal)
RETURNS storage_type
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_union);

Datum
my_union(PG_FUNCTION_ARGS)
{
    GistEntryVector *entryvec = (GistEntryVector *) PG_GETARG_POINTER(0);
    GISTENTRY  *ent = entryvec-&gt;vector;
    data_type  *out,
               *tmp,
               *old;
    int         numranges,
                i = 0;

    numranges = entryvec-&gt;n;
    tmp = DatumGetDataType(ent[0].key);
    out = tmp;

    if (numranges == 1)
    {
        out = data_type_deep_copy(tmp);

        PG_RETURN_DATA_TYPE_P(out);
    }

    for (i = 1; i &lt; numranges; i++)
    {
        old = out;
        tmp = DatumGetDataType(ent[i].key);
        out = my_union_implementation(out, tmp);
    }

    PG_RETURN_DATA_TYPE_P(out);
}
</programlisting>
      </para>

      <para>
        As you can see, in this skeleton we're dealing with a data type
        where <literal>union(X, Y, Z) = union(union(X, Y), Z)</literal>. It's easy
        enough to support data types where this is not the case, by
        implementing the proper union algorithm in this
        <acronym>GiST</acronym> support method.
      </para>

      <para>
        The result of the <function>union</function> function must be a value of the
        index's storage type, whatever that is (it might or might not be
        different from the indexed column's type).  The <function>union</function>
        function should return a pointer to newly <function>palloc()</function>ed
        memory. You can't just return the input value as-is, even if there is
        no type change.
      </para>

      <para>
       As shown above, the <function>union</function> function's
       first <type>internal</type> argument is actually
       a <structname>GistEntryVector</structname> pointer.  The second argument is a
       pointer to an integer variable, which can be ignored.  (It used to be
       required that the <function>union</function> function store the size of its
       result value into that variable, but this is no longer necessary.)
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>compress</function></term>
     <listitem>
      <para>
       Converts a data item into a format suitable for physical storage in
       an index page.
       If the <function>compress</function> method is omitted, data items are stored
       in the index without modification.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_compress(internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_compress);

Datum
my_compress(PG_FUNCTION_ARGS)
{
    GISTENTRY  *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
    GISTENTRY  *retval;

    if (entry-&gt;leafkey)
    {
        /* replace entry-&gt;key with a compressed version */
        compressed_data_type *compressed_data = palloc(sizeof(compressed_data_type));

        /* fill *compressed_data from entry-&gt;key ... */

        retval = palloc(sizeof(GISTENTRY));
        gistentryinit(*retval, PointerGetDatum(compressed_data),
                      entry-&gt;rel, entry-&gt;page, entry-&gt;offset, FALSE);
    }
    else
    {
        /* typically we needn't do anything with non-leaf entries */
        retval = entry;
    }

    PG_RETURN_POINTER(retval);
}
</programlisting>
      </para>

      <para>
       You have to adapt <replaceable>compressed_data_type</replaceable> to the specific
       type you're converting to in order to compress your leaf nodes, of
       course.
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>decompress</function></term>
     <listitem>
      <para>
       Converts the stored representation of a data item into a format that
       can be manipulated by the other GiST methods in the operator class.
       If the <function>decompress</function> method is omitted, it is assumed that
       the other GiST methods can work directly on the stored data format.
       (<function>decompress</function> is not necessarily the reverse of
       the <function>compress</function> method; in particular,
       if <function>compress</function> is lossy then it's impossible
       for <function>decompress</function> to exactly reconstruct the original
       data.  <function>decompress</function> is not necessarily equivalent
       to <function>fetch</function>, either, since the other GiST methods might not
       require full reconstruction of the data.)
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_decompress(internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_decompress);

Datum
my_decompress(PG_FUNCTION_ARGS)
{
    PG_RETURN_POINTER(PG_GETARG_POINTER(0));
}
</programlisting>

        The above skeleton is suitable for the case where no decompression
        is needed.  (But, of course, omitting the method altogether is even
        easier, and is recommended in such cases.)
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>penalty</function></term>
     <listitem>
      <para>
       Returns a value indicating the <quote>cost</quote> of inserting the new
       entry into a particular branch of the tree.  Items will be inserted
       down the path of least <function>penalty</function> in the tree.
       Values returned by <function>penalty</function> should be non-negative.
       If a negative value is returned, it will be treated as zero.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_penalty(internal, internal, internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;  -- in some cases penalty functions need not be strict
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_penalty);

Datum
my_penalty(PG_FUNCTION_ARGS)
{
    GISTENTRY  *origentry = (GISTENTRY *) PG_GETARG_POINTER(0);
    GISTENTRY  *newentry = (GISTENTRY *) PG_GETARG_POINTER(1);
    float      *penalty = (float *) PG_GETARG_POINTER(2);
    data_type  *orig = DatumGetDataType(origentry-&gt;key);
    data_type  *new = DatumGetDataType(newentry-&gt;key);

    *penalty = my_penalty_implementation(orig, new);
    PG_RETURN_POINTER(penalty);
}
</programlisting>

        For historical reasons, the <function>penalty</function> function doesn't
        just return a <type>float</type> result; instead it has to store the value
        at the location indicated by the third argument.  The return
        value per se is ignored, though it's conventional to pass back the
        address of that argument.
      </para>

      <para>
        The <function>penalty</function> function is crucial to good performance of
        the index. It'll get used at insertion time to determine which branch
        to follow when choosing where to add the new entry in the tree. At
        query time, the more balanced the index, the quicker the lookup.
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>picksplit</function></term>
     <listitem>
      <para>
       When an index page split is necessary, this function decides which
       entries on the page are to stay on the old page, and which are to move
       to the new page.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_picksplit(internal, internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_picksplit);

Datum
my_picksplit(PG_FUNCTION_ARGS)
{
    GistEntryVector *entryvec = (GistEntryVector *) PG_GETARG_POINTER(0);
    GIST_SPLITVEC *v = (GIST_SPLITVEC *) PG_GETARG_POINTER(1);
    OffsetNumber maxoff = entryvec-&gt;n - 1;
    GISTENTRY  *ent = entryvec-&gt;vector;
    int         i,
                nbytes;
    OffsetNumber *left,
               *right;
    data_type  *tmp_union;
    data_type  *unionL;
    data_type  *unionR;
    GISTENTRY **raw_entryvec;

    maxoff = entryvec-&gt;n - 1;
    nbytes = (maxoff + 1) * sizeof(OffsetNumber);

    v-&gt;spl_left = (OffsetNumber *) palloc(nbytes);
    left = v-&gt;spl_left;
    v-&gt;spl_nleft = 0;

    v-&gt;spl_right = (OffsetNumber *) palloc(nbytes);
    right = v-&gt;spl_right;
    v-&gt;spl_nright = 0;

    unionL = NULL;
    unionR = NULL;

    /* Initialize the raw entry vector. */
    raw_entryvec = (GISTENTRY **) malloc(entryvec-&gt;n * sizeof(void *));
    for (i = FirstOffsetNumber; i &lt;= maxoff; i = OffsetNumberNext(i))
        raw_entryvec[i] = &amp;(entryvec-&gt;vector[i]);

    for (i = FirstOffsetNumber; i &lt;= maxoff; i = OffsetNumberNext(i))
    {
        int         real_index = raw_entryvec[i] - entryvec-&gt;vector;

        tmp_union = DatumGetDataType(entryvec-&gt;vector[real_index].key);
        Assert(tmp_union != NULL);

        /*
         * Choose where to put the index entries and update unionL and unionR
         * accordingly. Append the entries to either v-&gt;spl_left or
         * v-&gt;spl_right, and care about the counters.
         */

        if (my_choice_is_left(unionL, curl, unionR, curr))
        {
            if (unionL == NULL)
                unionL = tmp_union;
            else
                unionL = my_union_implementation(unionL, tmp_union);

            *left = real_index;
            ++left;
            ++(v-&gt;spl_nleft);
        }
        else
        {
            /*
             * Same on the right
             */
        }
    }

    v-&gt;spl_ldatum = DataTypeGetDatum(unionL);
    v-&gt;spl_rdatum = DataTypeGetDatum(unionR);
    PG_RETURN_POINTER(v);
}
</programlisting>

       Notice that the <function>picksplit</function> function's result is delivered
       by modifying the passed-in <structname>v</structname> structure.  The return
       value per se is ignored, though it's conventional to pass back the
       address of <structname>v</structname>.
      </para>

      <para>
        Like <function>penalty</function>, the <function>picksplit</function> function
        is crucial to good performance of the index.  Designing suitable
        <function>penalty</function> and <function>picksplit</function> implementations
        is where the challenge of implementing well-performing
        <acronym>GiST</acronym> indexes lies.
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>same</function></term>
     <listitem>
      <para>
       Returns true if two index entries are identical, false otherwise.
       (An <quote>index entry</quote> is a value of the index's storage type,
       not necessarily the original indexed column's type.)
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_same(storage_type, storage_type, internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_same);

Datum
my_same(PG_FUNCTION_ARGS)
{
    prefix_range *v1 = PG_GETARG_PREFIX_RANGE_P(0);
    prefix_range *v2 = PG_GETARG_PREFIX_RANGE_P(1);
    bool       *result = (bool *) PG_GETARG_POINTER(2);

    *result = my_eq(v1, v2);
    PG_RETURN_POINTER(result);
}
</programlisting>

        For historical reasons, the <function>same</function> function doesn't
        just return a Boolean result; instead it has to store the flag
        at the location indicated by the third argument.  The return
        value per se is ignored, though it's conventional to pass back the
        address of that argument.
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>distance</function></term>
     <listitem>
      <para>
       Given an index entry <literal>p</literal> and a query value <literal>q</literal>,
       this function determines the index entry's
       <quote>distance</quote> from the query value.  This function must be
       supplied if the operator class contains any ordering operators.
       A query using the ordering operator will be implemented by returning
       index entries with the smallest <quote>distance</quote> values first,
       so the results must be consistent with the operator's semantics.
       For a leaf index entry the result just represents the distance to
       the index entry; for an internal tree node, the result must be the
       smallest distance that any child entry could have.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_distance(internal, data_type, smallint, oid, internal)
RETURNS float8
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        And the matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_distance);

Datum
my_distance(PG_FUNCTION_ARGS)
{
    GISTENTRY  *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
    data_type  *query = PG_GETARG_DATA_TYPE_P(1);
    StrategyNumber strategy = (StrategyNumber) PG_GETARG_UINT16(2);
    /* Oid subtype = PG_GETARG_OID(3); */
    /* bool *recheck = (bool *) PG_GETARG_POINTER(4); */
    data_type  *key = DatumGetDataType(entry-&gt;key);
    double      retval;

    /*
     * determine return value as a function of strategy, key and query.
     */

    PG_RETURN_FLOAT8(retval);
}
</programlisting>

       The arguments to the <function>distance</function> function are identical to
       the arguments of the <function>consistent</function> function.
      </para>

      <para>
       Some approximation is allowed when determining the distance, so long
       as the result is never greater than the entry's actual distance. Thus,
       for example, distance to a bounding box is usually sufficient in
       geometric applications.  For an internal tree node, the distance
       returned must not be greater than the distance to any of the child
       nodes. If the returned distance is not exact, the function must set
       <literal>*recheck</literal> to true. (This is not necessary for internal tree
       nodes; for them, the calculation is always assumed to be inexact.) In
       this case the executor will calculate the accurate distance after
       fetching the tuple from the heap, and reorder the tuples if necessary.
      </para>

      <para>
       If the distance function returns <literal>*recheck = true</literal> for any
       leaf node, the original ordering operator's return type must
       be <type>float8</type> or <type>float4</type>, and the distance function's
       result values must be comparable to those of the original ordering
       operator, since the executor will sort using both distance function
       results and recalculated ordering-operator results.  Otherwise, the
       distance function's result values can be any finite <type>float8</type>
       values, so long as the relative order of the result values matches the
       order returned by the ordering operator.  (Infinity and minus infinity
       are used internally to handle cases such as nulls, so it is not
       recommended that <function>distance</function> functions return these values.)
      </para>

     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>fetch</function></term>
     <listitem>
      <para>
       Converts the compressed index representation of a data item into the
       original data type, for index-only scans. The returned data must be an
       exact, non-lossy copy of the originally indexed value.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_fetch(internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

        The argument is a pointer to a <structname>GISTENTRY</structname> struct. On
        entry, its <structfield>key</structfield> field contains a non-NULL leaf datum in
        compressed form. The return value is another <structname>GISTENTRY</structname>
        struct, whose <structfield>key</structfield> field contains the same datum in its
        original, uncompressed form. If the opclass's compress function does
        nothing for leaf entries, the <function>fetch</function> method can return the
        argument as-is.  Or, if the opclass does not have a compress function,
        the <function>fetch</function> method can be omitted as well, since it would
        necessarily be a no-op.
       </para>

       <para>
        The matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_fetch);

Datum
my_fetch(PG_FUNCTION_ARGS)
{
    GISTENTRY  *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
    input_data_type *in = DatumGetPointer(entry->key);
    fetched_data_type *fetched_data;
    GISTENTRY  *retval;

    retval = palloc(sizeof(GISTENTRY));
    fetched_data = palloc(sizeof(fetched_data_type));

    /*
     * Convert 'fetched_data' into the a Datum of the original datatype.
     */

    /* fill *retval from fetched_data. */
    gistentryinit(*retval, PointerGetDatum(converted_datum),
                  entry->rel, entry->page, entry->offset, FALSE);

    PG_RETURN_POINTER(retval);
}
</programlisting>
      </para>

      <para>
       If the compress method is lossy for leaf entries, the operator class
       cannot support index-only scans, and must not define
       a <function>fetch</function> function.
      </para>

     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>options</function></term>
     <listitem>
      <para>
       Allows definition of user-visible parameters that control operator
       class behavior.
      </para>

      <para>
        The <acronym>SQL</acronym> declaration of the function must look like this:

<programlisting>
CREATE OR REPLACE FUNCTION my_options(internal)
RETURNS void
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>
      </para>

      <para>
       The function is passed a pointer to a <structname>local_relopts</structname>
       struct, which needs to be filled with a set of operator class
       specific options.  The options can be accessed from other support
       functions using the <literal>PG_HAS_OPCLASS_OPTIONS()</literal> and
       <literal>PG_GET_OPCLASS_OPTIONS()</literal> macros.
      </para>

       <para>
        An example implementation of my_options() and parameters use
        from other support functions are given below:

<programlisting>
typedef enum MyEnumType
{
    MY_ENUM_ON,
    MY_ENUM_OFF,
    MY_ENUM_AUTO
} MyEnumType;

typedef struct
{
    int32   vl_len_;    /* varlena header (do not touch directly!) */
    int     int_param;  /* integer parameter */
    double  real_param; /* real parameter */
    MyEnumType enum_param; /* enum parameter */
    int     str_param;  /* string parameter */
} MyOptionsStruct;

/* String representation of enum values */
static relopt_enum_elt_def myEnumValues[] =
{
    {"on", MY_ENUM_ON},
    {"off", MY_ENUM_OFF},
    {"auto", MY_ENUM_AUTO},
    {(const char *) NULL}   /* list terminator */
};

static char *str_param_default = "default";

/*
 * Sample validator: checks that string is not longer than 8 bytes.
 */
static void
validate_my_string_relopt(const char *value)
{
    if (strlen(value) > 8)
        ereport(ERROR,
                (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
                 errmsg("str_param must be at most 8 bytes")));
}

/*
 * Sample filler: switches characters to lower case.
 */
static Size
fill_my_string_relopt(const char *value, void *ptr)
{
    char   *tmp = str_tolower(value, strlen(value), DEFAULT_COLLATION_OID);
    int     len = strlen(tmp);

    if (ptr)
        strcpy((char *) ptr, tmp);

    pfree(tmp);
    return len + 1;
}

PG_FUNCTION_INFO_V1(my_options);

Datum
my_options(PG_FUNCTION_ARGS)
{
    local_relopts *relopts = (local_relopts *) PG_GETARG_POINTER(0);

    init_local_reloptions(relopts, sizeof(MyOptionsStruct));
    add_local_int_reloption(relopts, "int_param", "integer parameter",
                            100, 0, 1000000,
                            offsetof(MyOptionsStruct, int_param));
    add_local_real_reloption(relopts, "real_param", "real parameter",
                             1.0, 0.0, 1000000.0,
                             offsetof(MyOptionsStruct, real_param));
    add_local_enum_reloption(relopts, "enum_param", "enum parameter",
                             myEnumValues, MY_ENUM_ON,
                             "Valid values are: \"on\", \"off\" and \"auto\".",
                             offsetof(MyOptionsStruct, enum_param));
    add_local_string_reloption(relopts, "str_param", "string parameter",
                               str_param_default,
                               &amp;validate_my_string_relopt,
                               &amp;fill_my_string_relopt,
                               offsetof(MyOptionsStruct, str_param));

    PG_RETURN_VOID();
}

PG_FUNCTION_INFO_V1(my_compress);

Datum
my_compress(PG_FUNCTION_ARGS)
{
    int     int_param = 100;
    double  real_param = 1.0;
    MyEnumType enum_param = MY_ENUM_ON;
    char   *str_param = str_param_default;

    /*
     * Normally, when opclass contains 'options' method, then options are always
     * passed to support functions.  However, if you add 'options' method to
     * existing opclass, previously defined indexes have no options, so the
     * check is required.
     */
    if (PG_HAS_OPCLASS_OPTIONS())
    {
        MyOptionsStruct *options = (MyOptionsStruct *) PG_GET_OPCLASS_OPTIONS();

        int_param = options->int_param;
        real_param = options->real_param;
        enum_param = options->enum_param;
        str_param = GET_STRING_RELOPTION(options, str_param);
    }

    /* the rest implementation of support function */
}

</programlisting>
      </para>

      <para>
       Since the representation of the key in <acronym>GiST</acronym> is
       flexible, it may depend on user-specified parameters.  For instance,
       the length of key signature may be specified.  See
       <literal>gtsvector_options()</literal> for example.
      </para>
     </listitem>
    </varlistentry>

    <varlistentry>
     <term><function>sortsupport</function></term>
     <listitem>
      <para>
       Returns a comparator function to sort data in a way that preserves
       locality. It is used by <command>CREATE INDEX</command> and
       <command>REINDEX</command> commands. The quality of the created index
       depends on how well the sort order determined by the comparator function
       preserves locality of the inputs.
      </para>
      <para>
       The <function>sortsupport</function> method is optional. If it is not
       provided, <command>CREATE INDEX</command> builds the index by inserting
       each tuple to the tree using the <function>penalty</function> and
       <function>picksplit</function> functions, which is much slower.
      </para>

      <para>
       The <acronym>SQL</acronym> declaration of the function must look like
       this:

<programlisting>
CREATE OR REPLACE FUNCTION my_sortsupport(internal)
RETURNS void
AS 'MODULE_PATHNAME'
LANGUAGE C STRICT;
</programlisting>

       The argument is a pointer to a <structname>SortSupport</structname>
       struct. At a minimum, the function must fill in its comparator field.
       The comparator takes three arguments: two Datums to compare, and
       a pointer to the <structname>SortSupport</structname> struct. The
       Datums are the two indexed values in the format that they are stored
       in the index; that is, in the format returned by the
       <function>compress</function> method. The full API is defined in
       <filename>src/include/utils/sortsupport.h</filename>.
       </para>

       <para>
        The matching code in the C module could then follow this skeleton:

<programlisting>
PG_FUNCTION_INFO_V1(my_sortsupport);

static int
my_fastcmp(Datum x, Datum y, SortSupport ssup)
{
  /* establish order between x and y by computing some sorting value z */

  int z1 = ComputeSpatialCode(x);
  int z2 = ComputeSpatialCode(y);

  return z1 == z2 ? 0 : z1 > z2 ? 1 : -1;
}

Datum
my_sortsupport(PG_FUNCTION_ARGS)
{
  SortSupport ssup = (SortSupport) PG_GETARG_POINTER(0);

  ssup->comparator = my_fastcmp;
  PG_RETURN_VOID();
}
</programlisting>
      </para>
     </listitem>
    </varlistentry>
  </variablelist>

  <para>
   All the GiST support methods are normally called in short-lived memory
   contexts; that is, <varname>CurrentMemoryContext</varname> will get reset after
   each tuple is processed.  It is therefore not very important to worry about
   pfree'ing everything you palloc.  However, in some cases it's useful for a
   support method to cache data across repeated calls.  To do that, allocate
   the longer-lived data in <literal>fcinfo-&gt;flinfo-&gt;fn_mcxt</literal>, and
   keep a pointer to it in <literal>fcinfo-&gt;flinfo-&gt;fn_extra</literal>.  Such
   data will survive for the life of the index operation (e.g., a single GiST
   index scan, index build, or index tuple insertion).  Be careful to pfree
   the previous value when replacing a <literal>fn_extra</literal> value, or the leak
   will accumulate for the duration of the operation.
  </para>

</sect1>

<sect1 id="gist-implementation">
 <title>Implementation</title>

 <sect2 id="gist-buffering-build">
  <title>GiST Index Build Methods</title>

  <para>
   The simplest way to build a GiST index is just to insert all the entries,
   one by one.  This tends to be slow for large indexes, because if the
   index tuples are scattered across the index and the index is large enough
   to not fit in cache, a lot of random I/O will be
   needed.  <productname>PostgreSQL</productname> supports two alternative
   methods for initial build of a GiST index: <firstterm>sorted</firstterm>
   and <firstterm>buffered</firstterm> modes.
  </para>

  <para>
   The sorted method is only available if each of the opclasses used by the
   index provides a <function>sortsupport</function> function, as described
   in <xref linkend="gist-extensibility"/>.  If they do, this method is
   usually the best, so it is used by default.
  </para>

  <para>
   The buffered method works by not inserting tuples directly into the index
   right away.  It can dramatically reduce the amount of random I/O needed
   for non-ordered data sets.  For well-ordered data sets the benefit is
   smaller or non-existent, because only a small number of pages receive new
   tuples at a time, and those pages fit in cache even if the index as a
   whole does not.
  </para>

  <para>
   The buffered method needs to call the <function>penalty</function>
   function more often than the simple method does, which consumes some
   extra CPU resources. Also, the buffers need temporary disk space, up to
   the size of the resulting index. Buffering can also influence the quality
   of the resulting index, in both positive and negative directions. That
   influence depends on various factors, like the distribution of the input
   data and the operator class implementation.
  </para>

  <para>
   If sorting is not possible, then by default a GiST index build switches
   to the buffering method when the index size reaches
   <xref linkend="guc-effective-cache-size"/>.  Buffering can be manually
   forced or prevented by the <literal>buffering</literal> parameter to the
   CREATE INDEX command.  The default behavior is good for most cases, but
   turning buffering off might speed up the build somewhat if the input data
   is ordered.
  </para>

 </sect2>
</sect1>

<sect1 id="gist-examples">
 <title>Examples</title>

 <para>
  The <productname>PostgreSQL</productname> source distribution includes
  several examples of index methods implemented using
  <acronym>GiST</acronym>.  The core system currently provides text search
  support (indexing for <type>tsvector</type> and <type>tsquery</type>) as well as
  R-Tree equivalent functionality for some of the built-in geometric data types
  (see <filename>src/backend/access/gist/gistproc.c</filename>).  The following
  <filename>contrib</filename> modules also contain <acronym>GiST</acronym>
  operator classes:

 <variablelist>
  <varlistentry>
   <term><filename>btree_gist</filename></term>
   <listitem>
    <para>B-tree equivalent functionality for several data types</para>
   </listitem>
  </varlistentry>

  <varlistentry>
   <term><filename>cube</filename></term>
   <listitem>
    <para>Indexing for multidimensional cubes</para>
   </listitem>
  </varlistentry>

  <varlistentry>
   <term><filename>hstore</filename></term>
   <listitem>
    <para>Module for storing (key, value) pairs</para>
   </listitem>
  </varlistentry>

  <varlistentry>
   <term><filename>intarray</filename></term>
   <listitem>
    <para>RD-Tree for one-dimensional array of int4 values</para>
   </listitem>
  </varlistentry>

  <varlistentry>
   <term><filename>ltree</filename></term>
   <listitem>
    <para>Indexing for tree-like structures</para>
   </listitem>
  </varlistentry>

  <varlistentry>
   <term><filename>pg_trgm</filename></term>
   <listitem>
    <para>Text similarity using trigram matching</para>
   </listitem>
  </varlistentry>

  <varlistentry>
   <term><filename>seg</filename></term>
   <listitem>
    <para>Indexing for <quote>float ranges</quote></para>
   </listitem>
  </varlistentry>
 </variablelist>
 </para>

</sect1>

</chapter>