Internals//GiST Indexes: Extensibility

Given an index entry p and a query value q, this function determines whether the index entry is "consistent" with the query; that is, could the predicate "indexed_column indexable_operator q" 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 true, a recheck flag must also be returned. This indicates whether the predicate is certainly true or only possibly true. If recheck = false then the index has tested the predicate condition exactly, whereas if recheck = true the row is only a candidate match. In that case the system will automatically evaluate the indexable_operator against the actual row value to see if it is really a match. This convention allows GiST to support both lossless and lossy index structures.

The SQL declaration of the function must look like this:

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

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

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->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);
}

Here, key is an element in the index and query the value being looked up in the index. The StrategyNumber parameter indicates which operator of your operator class is being applied — it matches one of the operator numbers in the CREATE OPERATOR CLASS command.

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

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.

The SQL declaration of the function must look like this:

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

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

PG_FUNCTION_INFO_V1(my_union);

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

    numranges = entryvec->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 < numranges; i++)
    {
        old = out;
        tmp = DatumGetDataType(ent[i].key);
        out = my_union_implementation(out, tmp);
    }

    PG_RETURN_DATA_TYPE_P(out);
}

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

The result of the union 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 union function should return a pointer to newly palloc()ed memory. You can't just return the input value as-is, even if there is no type change.

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

Converts the data item into a format suitable for physical storage in an index page.

The SQL declaration of the function must look like this:

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

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

PG_FUNCTION_INFO_V1(my_compress);

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

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

        /* fill *compressed_data from entry->key ... */

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

    PG_RETURN_POINTER(retval);
}

You have to adapt compressed_data_type to the specific type you're converting to in order to compress your leaf nodes, of course.

The reverse of the compress method. Converts the index representation of the data item into a format that can be manipulated by the other GiST methods in the operator class.

The SQL declaration of the function must look like this:

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

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

PG_FUNCTION_INFO_V1(my_decompress);

Datum
my_decompress(PG_FUNCTION_ARGS)
{
    PG_RETURN_POINTER(PG_GETARG_POINTER(0));
}

The above skeleton is suitable for the case where no decompression is needed.

Returns a value indicating the "cost" of inserting the new entry into a particular branch of the tree. Items will be inserted down the path of least penalty in the tree. Values returned by penalty should be non-negative. If a negative value is returned, it will be treated as zero.

The SQL declaration of the function must look like this:

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

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

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->key);
    data_type  *new = DatumGetDataType(newentry->key);

    *penalty = my_penalty_implementation(orig, new);
    PG_RETURN_POINTER(penalty);
}

For historical reasons, the penalty function doesn't just return a float 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.

The penalty 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.

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.

The SQL declaration of the function must look like this:

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

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

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->n - 1;
    GISTENTRY  *ent = entryvec->vector;
    int         i,
                nbytes;
    OffsetNumber *left,
               *right;
    data_type  *tmp_union;
    data_type  *unionL;
    data_type  *unionR;
    GISTENTRY **raw_entryvec;

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

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

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

    unionL = NULL;
    unionR = NULL;

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

    for (i = FirstOffsetNumber; i <= maxoff; i = OffsetNumberNext(i))
    {
        int         real_index = raw_entryvec[i] - entryvec->vector;

        tmp_union = DatumGetDataType(entryvec->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_spl_left or
         * v_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->spl_nleft);
        }
        else
        {
            /*
             * Same on the right
             */
        }
    }

    v->spl_ldatum = DataTypeGetDatum(unionL);
    v->spl_rdatum = DataTypeGetDatum(unionR);
    PG_RETURN_POINTER(v);
}

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

Like penalty, the picksplit function is crucial to good performance of the index. Designing suitable penalty and picksplit implementations is where the challenge of implementing well-performing GiST indexes lies.

Returns true if two index entries are identical, false otherwise. (An "index entry" is a value of the index's storage type, not necessarily the original indexed column's type.)

The SQL declaration of the function must look like this:

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

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

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);
}

For historical reasons, the same 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.

Given an index entry p and a query value q, this function determines the index entry's "distance" 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 "distance" 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.

The SQL declaration of the function must look like this:

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

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

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->key);
    double      retval;

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

    PG_RETURN_FLOAT8(retval);
}

The arguments to the distance function are identical to the arguments of the consistent function.

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 *recheck 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.

If the distance function returns *recheck = true for any leaf node, the original ordering operator's return type must be float8 or float4, 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 float8 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 distance functions return these values.)

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.

The SQL declaration of the function must look like this:

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

The argument is a pointer to a GISTENTRY struct. On entry, its key field contains a non-NULL leaf datum in compressed form. The return value is another GISTENTRY struct, whose key field contains the same datum in its original, uncompressed form. If the opclass's compress function does nothing for leaf entries, the fetch method can return the argument as-is.

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

PG_FUNCTION_INFO_V1(my_fetch);

Datum
my_fetch(PG_FUNCTION_ARGS)
{
    GISTENTRY  *entry = (GISTENTRY *) PG_GETARG_POINTER(0);
    input_data_type *in = DatumGetP(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 fetch_data. */
    gistentryinit(*retval, PointerGetDatum(converted_datum),
                  entry->rel, entry->page, entry->offset, FALSE);

    PG_RETURN_POINTER(retval);
}

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