Commit e6a8eba3 authored by Tom Lane's avatar Tom Lane

Remove contrib version of rtree_gist --- now in core system.

parent e7e16942
# $PostgreSQL: pgsql/contrib/Makefile,v 1.56 2005/06/23 02:33:25 tgl Exp $ # $PostgreSQL: pgsql/contrib/Makefile,v 1.57 2005/07/01 19:23:04 tgl Exp $
subdir = contrib subdir = contrib
top_builddir = .. top_builddir = ..
...@@ -27,7 +27,6 @@ WANTED_DIRS = \ ...@@ -27,7 +27,6 @@ WANTED_DIRS = \
pgbench \ pgbench \
pgcrypto \ pgcrypto \
pgstattuple \ pgstattuple \
rtree_gist \
seg \ seg \
spi \ spi \
tablefunc \ tablefunc \
......
...@@ -136,10 +136,6 @@ reindexdb - ...@@ -136,10 +136,6 @@ reindexdb -
Reindexes a database Reindexes a database
by Shaun Thomas <sthomas@townnews.com> by Shaun Thomas <sthomas@townnews.com>
rtree_gist -
Support for emulating RTREE indexing in GiST
by Oleg Bartunov <oleg@sai.msu.su> and Teodor Sigaev <teodor@sigaev.ru>
seg - seg -
Confidence-interval datatype (GiST indexing example) Confidence-interval datatype (GiST indexing example)
by Gene Selkov, Jr. <selkovjr@mcs.anl.gov> by Gene Selkov, Jr. <selkovjr@mcs.anl.gov>
......
# $PostgreSQL: pgsql/contrib/rtree_gist/Makefile,v 1.5 2004/08/20 20:13:07 momjian Exp $
MODULES = rtree_gist
DATA_built = rtree_gist.sql
DOCS = README.rtree_gist
REGRESS = rtree_gist
ifdef USE_PGXS
PGXS = $(shell pg_config --pgxs)
include $(PGXS)
else
subdir = contrib/rtree_gist
top_builddir = ../..
include $(top_builddir)/src/Makefile.global
include $(top_srcdir)/contrib/contrib-global.mk
endif
This is a R-Tree implementation using GiST.
Code (for PG95) are taken from http://s2k-ftp.cs.berkeley.edu:8000/gist/pggist/
and changed according to new version of GiST (7.1 and above)
All work was done by Teodor Sigaev (teodor@stack.net) and Oleg Bartunov
(oleg@sai.msu.su). See http://www.sai.msu.su/~megera/postgres/gist
for additional information.
CHANGES:
Oct 10 MSD 2001
1. Implemented new linear algorithm for picksplit
ref. ( 'New Linear Node Splitting Algorithm for R-tree',
C.H.Ang and T.C.Tan )
Tue May 29 17:04:16 MSD 2001
1. Small fixes in polygon code
Thanks to Dave Blasby <dblasby@refractions.net>
Mon May 28 19:42:14 MSD 2001
1. Full implementation of R-tree using GiST - gist_box_ops,gist_poly_ops
2. gist_poly_ops is lossy
3. NULLs support
4. works with multi-key GiST
NOTICE:
This version will only work with PostgreSQL version 7.1 and above
because of changes in the function call interface.
INSTALLATION:
gmake
gmake install
-- load functions
psql <database> < rtree_gist.sql
REGRESSION TEST:
gmake installcheck
EXAMPLE USAGE:
create table boxtmp (b box);
-- create index
create index bix on boxtmp using gist (b gist_box_ops);
-- query
select * from boxtmp where b && '(1000,1000,0,0)'::box;
BENCHMARKS:
subdirectory bench contains benchmark suite.
Prerequisities: perl, DBI, DBD:Pg, Time::HiRes
cd ./bench
1. createdb TEST
2. psql TEST < ../rtree_gist.sql
3. ./create_test.pl | psql TEST
-- change $NUM - number of rows in test dataset
4. ./bench.pl - perl script to benchmark queries.
Run script without arguments to see available options.
a)test without GiST index, using built-in R-Tree
./bench.pl -d TEST
b)test R-Tree using GiST index
./bench.pl -d TEST -g
RESULTS:
1. One interesting thing is that insertion time for built-in R-Tree is
about 8 times more than ones for GiST implementation of R-Tree !!!
2. Postmaster requires much more memory for built-in R-Tree
3. Search time depends on dataset. In our case we got:
+------------+-----------+--------------+
|Number boxes|R-tree, sec|R-tree using |
| | | GiST, sec |
+------------+-----------+--------------+
| 10| 0.002| 0.002|
+------------+-----------+--------------+
| 100| 0.002| 0.002|
+------------+-----------+--------------+
| 1000| 0.002| 0.002|
+------------+-----------+--------------+
| 10000| 0.015| 0.025|
+------------+-----------+--------------+
| 20000| 0.029| 0.048|
+------------+-----------+--------------+
| 40000| 0.055| 0.092|
+------------+-----------+--------------+
| 80000| 0.113| 0.178|
+------------+-----------+--------------+
| 160000| 0.338| 0.337|
+------------+-----------+--------------+
| 320000| 0.674| 0.673|
+------------+-----------+--------------+
#!/usr/bin/perl -w
use strict;
# make sure we are in a sane environment.
use DBI();
use DBD::Pg();
use Time::HiRes qw( usleep ualarm gettimeofday tv_interval );
use Getopt::Std;
my %opt;
getopts('d:b:gv', \%opt);
if ( !( scalar %opt ) ) {
print <<EOT;
Usage:
$0 -d DATABASE -b N [-v] [-g]
-d DATABASE - DATABASE name
-b N -number of cycles
-v - print sql
-g -use GiST index( default built-in R-tree )
EOT
exit;
}
$opt{d} ||= 'TEST';
my $dbi=DBI->connect('DBI:Pg:dbname='.$opt{d}) || die "Couldn't connect DB: $opt{d} !\n";
my $setsql = qq{
SET search_path = public;
};
my $sth = $dbi->prepare($setsql);
$sth->execute();
my $sql;
my $notice;
my $sss = '(3000,3000,2990,2990)';
if ( $opt{g} ) {
$notice = "Testing GiST implementation of R-Tree";
$sql = "select count(*) from boxtmp where b && '$sss'::box;";
} else {
$notice = "Testing built-in implementation of R-Tree";
$sql = "select count(*) from boxtmp2 where b && '$sss'::box;";
}
my $t0 = [gettimeofday];
my $count=0;
my $b=$opt{b};
$b ||=1;
foreach ( 1..$b ) {
my @a=exec_sql($dbi,$sql);
$count=$#a;
}
my $elapsed = tv_interval ( $t0, [gettimeofday]);
print "$notice:\n";
print "$sql\n" if ( $opt{v} );
print "Done\n";
print sprintf("total: %.02f sec; number: %d; for one: %.03f sec; found %d docs\n", $elapsed, $b, $elapsed/$b, $count+1 );
$dbi -> disconnect;
sub exec_sql {
my ($dbi, $sql, @keys) = @_;
my $sth=$dbi->prepare($sql) || die;
$sth->execute( @keys ) || die;
my $r;
my @row;
while ( defined ( $r=$sth->fetchrow_hashref ) ) {
push @row, $r;
}
$sth->finish;
return @row;
}
#!/usr/bin/perl
use strict;
my $NUM = 20000;
print "DROP TABLE boxtmp;\n";
print "DROP TABLE boxtmp2;\n";
print "CREATE TABLE boxtmp (b box);\n";
print "CREATE TABLE boxtmp2 (b box);\n";
srand(1);
open(DAT,">bbb.dat") || die;
foreach ( 1..$NUM ) {
#print DAT '(',int( 500+500*rand() ),',',int( 500+500*rand() ),',',int( 500*rand() ),',',int( 500*rand() ),")\n";
my ( $x1,$y1, $x2,$y2 ) = (
10000*rand(),
10000*rand(),
10000*rand(),
10000*rand()
);
print DAT '(',
max($x1,$x2),',',
max($y1,$y2),',',
min($x1,$x2),',',
min($y1,$y2),")\n";
}
close DAT;
print "COPY boxtmp FROM stdin;\n";
open(DAT,"bbb.dat") || die;
while(<DAT>) { print; }
close DAT;
print "\\.\n";
print "COPY boxtmp2 FROM stdin;\n";
open(DAT,"bbb.dat") || die;
while(<DAT>) { print; }
close DAT;
print "\\.\n";
print "CREATE INDEX bix ON boxtmp USING gist (b gist_box_ops);\n";
print "CREATE INDEX bix2 ON boxtmp2 USING rtree (b box_ops);\n";
sub min {
return ( $_[0] < $_[1] ) ? $_[0] : $_[1];
}
sub max {
return ( $_[0] > $_[1] ) ? $_[0] : $_[1];
}
This diff is collapsed.
--
-- first, define the datatype. Turn off echoing so that expected file
-- does not depend on contents of seg.sql.
--
\set ECHO none
CREATE TABLE boxtmp (b box);
\copy boxtmp from 'data/test_box.data'
SELECT count(*)
FROM boxtmp
WHERE b && '(1000,1000,0,0)'::box;
count
-------
2
(1 row)
CREATE INDEX bix ON boxtmp USING rtree (b);
SELECT count(*)
FROM boxtmp
WHERE b && '(1000,1000,0,0)'::box;
count
-------
2
(1 row)
DROP INDEX bix;
CREATE INDEX bix ON boxtmp USING gist (b);
SELECT count(*)
FROM boxtmp
WHERE b && '(1000,1000,0,0)'::box;
count
-------
2
(1 row)
CREATE TABLE polytmp (p polygon);
\copy polytmp from 'data/test_box.data'
CREATE INDEX pix ON polytmp USING rtree (p);
SELECT count(*)
FROM polytmp
WHERE p && '(1000,1000),(0,0)'::polygon;
count
-------
2
(1 row)
DROP INDEX pix;
CREATE INDEX pix ON polytmp USING gist (p);
SELECT count(*)
FROM polytmp
WHERE p && '(1000,1000),(0,0)'::polygon;
count
-------
2
(1 row)
This diff is collapsed.
-- Adjust this setting to control where the objects get created.
SET search_path = public;
--
--
--
-- BOX ops
--
--
--
-- define the GiST support methods
CREATE FUNCTION gbox_consistent(internal,box,int4)
RETURNS bool
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
CREATE FUNCTION gbox_compress(internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
CREATE FUNCTION rtree_decompress(internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
CREATE FUNCTION gbox_penalty(internal,internal,internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE 'C' STRICT;
CREATE FUNCTION gbox_picksplit(internal, internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
CREATE FUNCTION gbox_union(internal, internal)
RETURNS box
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
CREATE FUNCTION gbox_same(box, box, internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
-- create the operator class
CREATE OPERATOR CLASS gist_box_ops
DEFAULT FOR TYPE box USING gist
AS
OPERATOR 1 << ,
OPERATOR 2 &< ,
OPERATOR 3 && ,
OPERATOR 4 &> ,
OPERATOR 5 >> ,
OPERATOR 6 ~= ,
OPERATOR 7 ~ ,
OPERATOR 8 @ ,
OPERATOR 9 &<| ,
OPERATOR 10 <<| ,
OPERATOR 11 |>> ,
OPERATOR 12 |&> ,
FUNCTION 1 gbox_consistent (internal, box, int4),
FUNCTION 2 gbox_union (internal, internal),
FUNCTION 3 gbox_compress (internal),
FUNCTION 4 rtree_decompress (internal),
FUNCTION 5 gbox_penalty (internal, internal, internal),
FUNCTION 6 gbox_picksplit (internal, internal),
FUNCTION 7 gbox_same (box, box, internal);
--
--
--
-- POLYGON ops
--
--
--
-- define the GiST support methods
CREATE FUNCTION gpoly_consistent(internal,polygon,int4)
RETURNS bool
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
CREATE FUNCTION gpoly_compress(internal)
RETURNS internal
AS 'MODULE_PATHNAME'
LANGUAGE 'C';
-- create the operator class
CREATE OPERATOR CLASS gist_poly_ops
DEFAULT FOR TYPE polygon USING gist
AS
OPERATOR 1 << RECHECK,
OPERATOR 2 &< RECHECK,
OPERATOR 3 && RECHECK,
OPERATOR 4 &> RECHECK,
OPERATOR 5 >> RECHECK,
OPERATOR 6 ~= RECHECK,
OPERATOR 7 ~ RECHECK,
OPERATOR 8 @ RECHECK,
OPERATOR 9 &<| RECHECK,
OPERATOR 10 <<| RECHECK,
OPERATOR 11 |>> RECHECK,
OPERATOR 12 |&> RECHECK,
FUNCTION 1 gpoly_consistent (internal, polygon, int4),
FUNCTION 2 gbox_union (internal, internal),
FUNCTION 3 gpoly_compress (internal),
FUNCTION 4 rtree_decompress (internal),
FUNCTION 5 gbox_penalty (internal, internal, internal),
FUNCTION 6 gbox_picksplit (internal, internal),
FUNCTION 7 gbox_same (box, box, internal),
STORAGE box;
--
-- first, define the datatype. Turn off echoing so that expected file
-- does not depend on contents of seg.sql.
--
\set ECHO none
\i rtree_gist.sql
\set ECHO all
CREATE TABLE boxtmp (b box);
\copy boxtmp from 'data/test_box.data'
SELECT count(*)
FROM boxtmp
WHERE b && '(1000,1000,0,0)'::box;
CREATE INDEX bix ON boxtmp USING rtree (b);
SELECT count(*)
FROM boxtmp
WHERE b && '(1000,1000,0,0)'::box;
DROP INDEX bix;
CREATE INDEX bix ON boxtmp USING gist (b);
SELECT count(*)
FROM boxtmp
WHERE b && '(1000,1000,0,0)'::box;
CREATE TABLE polytmp (p polygon);
\copy polytmp from 'data/test_box.data'
CREATE INDEX pix ON polytmp USING rtree (p);
SELECT count(*)
FROM polytmp
WHERE p && '(1000,1000),(0,0)'::polygon;
DROP INDEX pix;
CREATE INDEX pix ON polytmp USING gist (p);
SELECT count(*)
FROM polytmp
WHERE p && '(1000,1000),(0,0)'::polygon;
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment