Skip to content

Commit f48c31f

Browse files
feodorfeodor
committed
Merge branch 'CORE-241-rum' of gitlab.postgrespro.ru:pgpro-dev/postgrespro into CORE-241-rum
2 parents 0a80645 + 00ceebc commit f48c31f

File tree

12 files changed

+402
-35
lines changed

12 files changed

+402
-35
lines changed

‎expected/rum.out

Lines changed: 111 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -5,6 +5,8 @@ BEFORE UPDATE OR INSERT ON test_rum
55
FOR EACH ROW EXECUTE PROCEDURE tsvector_update_trigger('a', 'pg_catalog.english', 't');
66
CREATE INDEX rumidx ON test_rum USING rum (a rum_tsvector_ops);
77
\copy test_rum(t) from 'data/rum.data';
8+
CREATE INDEX failed_rumidx ON test_rum USING rum (a rum_tsvector_timestamp_ops);
9+
ERROR: additional information attribute "a" is not found in index
810
SET enable_seqscan=off;
911
explain (costs off)
1012
SELECT count(*) FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english', 'ever|wrote');
@@ -76,6 +78,30 @@ SELECT count(*) FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english', '(comp
7678
2
7779
(1 row)
7880

81+
SELECT count(*) FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english', '(gave | half) <-> way');
82+
count
83+
-------
84+
2
85+
(1 row)
86+
87+
SELECT count(*) FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english', '(gave | !half) <-> way');
88+
count
89+
-------
90+
3
91+
(1 row)
92+
93+
SELECT count(*) FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english', '!gave & way');
94+
count
95+
-------
96+
3
97+
(1 row)
98+
99+
SELECT count(*) FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english', '!gave & wooded & !look');
100+
count
101+
-------
102+
1
103+
(1 row)
104+
79105
SELECT count(*) FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english',
80106
'def <-> fgr');
81107
count
@@ -124,6 +150,29 @@ SELECT
124150
57.5727 | 57.5727 | thinking--“to go or not to go?” We are this far on the way. Reached | 'far':11 'go':3,7 'reach':15 'think':1 'way':14
125151
(2 rows)
126152

153+
-- Check ranking normalization
154+
SELECT rum_ts_distance(a, to_tsquery('pg_catalog.english', 'way'), 0), *
155+
FROM test_rum
156+
WHERE a @@ to_tsquery('pg_catalog.english', 'way')
157+
ORDER BY a <=> to_tsquery('pg_catalog.english', 'way');
158+
rum_ts_distance | t | a
159+
-----------------+--------------------------------------------------------------------------+---------------------------------------------------------------
160+
16.4493 | my appreciation of you in a more complimentary way than by sending this | 'appreci':2 'complimentari':8 'send':12 'way':9
161+
16.4493 | itself. Put on your “specs” and look at the castle, half way up the | 'castl':10 'half':11 'look':7 'put':2 'spec':5 'way':12
162+
16.4493 | so well that only a fragment, as it were, gave way. It still hangs as if | 'fragment':6 'gave':10 'hang':14 'still':13 'way':11 'well':2
163+
16.4493 | thinking--“to go or not to go?” We are this far on the way. Reached | 'far':11 'go':3,7 'reach':15 'think':1 'way':14
164+
(4 rows)
165+
166+
SELECT rum_ts_distance(a, row(to_tsquery('pg_catalog.english', 'way & (go | half)'), 0)::rum_distance_query), *
167+
FROM test_rum
168+
WHERE a @@ to_tsquery('pg_catalog.english', 'way & (go | half)')
169+
ORDER BY a <=> to_tsquery('pg_catalog.english', 'way & (go | half)');
170+
rum_ts_distance | t | a
171+
-----------------+---------------------------------------------------------------------+---------------------------------------------------------
172+
8.22467 | itself. Put on your “specs” and look at the castle, half way up the | 'castl':10 'half':11 'look':7 'put':2 'spec':5 'way':12
173+
57.5727 | thinking--“to go or not to go?” We are this far on the way. Reached | 'far':11 'go':3,7 'reach':15 'think':1 'way':14
174+
(2 rows)
175+
127176
INSERT INTO test_rum (t) VALUES ('foo bar foo the over foo qq bar');
128177
INSERT INTO test_rum (t) VALUES ('345 qwerty copyright');
129178
INSERT INTO test_rum (t) VALUES ('345 qwerty');
@@ -158,6 +207,68 @@ SELECT a FROM test_rum WHERE a @@ to_tsquery('pg_catalog.english', 'bar') ORDER
158207
'bar':2,8 'foo':1,3,6 'qq':7
159208
(1 row)
160209

210+
-- Check full-index scan with order by
211+
SELECT a <=> to_tsquery('pg_catalog.english', 'ever|wrote') FROM test_rum ORDER BY a <=> to_tsquery('pg_catalog.english', 'ever|wrote');
212+
?column?
213+
----------
214+
16.4493
215+
16.4493
216+
Infinity
217+
Infinity
218+
Infinity
219+
Infinity
220+
Infinity
221+
Infinity
222+
Infinity
223+
Infinity
224+
Infinity
225+
Infinity
226+
Infinity
227+
Infinity
228+
Infinity
229+
Infinity
230+
Infinity
231+
Infinity
232+
Infinity
233+
Infinity
234+
Infinity
235+
Infinity
236+
Infinity
237+
Infinity
238+
Infinity
239+
Infinity
240+
Infinity
241+
Infinity
242+
Infinity
243+
Infinity
244+
Infinity
245+
Infinity
246+
Infinity
247+
Infinity
248+
Infinity
249+
Infinity
250+
Infinity
251+
Infinity
252+
Infinity
253+
Infinity
254+
Infinity
255+
Infinity
256+
Infinity
257+
Infinity
258+
Infinity
259+
Infinity
260+
Infinity
261+
Infinity
262+
Infinity
263+
Infinity
264+
Infinity
265+
Infinity
266+
Infinity
267+
Infinity
268+
Infinity
269+
Infinity
270+
(56 rows)
271+
161272
CREATE TABLE tst (i int4, t tsvector);
162273
INSERT INTO tst SELECT i%10, to_tsvector('simple', substr(md5(i::text), 1, 1)) FROM generate_series(1,100000) i;
163274
CREATE INDEX tstidx ON tst USING rum (t rum_tsvector_ops);

‎rum--1.0.sql

Lines changed: 29 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -6,10 +6,31 @@ LANGUAGE C;
66
-- Access method
77
CREATE ACCESS METHOD rum TYPE INDEX HANDLER rumhandler;
88

9-
-- Opclasses
9+
-- tsvector opclasses
10+
11+
CREATE TYPE rum_distance_query AS (query tsquery, method int);
12+
13+
CREATE FUNCTION tsquery_to_distance_query(tsquery)
14+
RETURNS rum_distance_query
15+
AS 'MODULE_PATHNAME', 'tsquery_to_distance_query'
16+
LANGUAGE C IMMUTABLE STRICT;
17+
18+
CREATE CAST (tsquery AS rum_distance_query)
19+
WITH FUNCTION tsquery_to_distance_query(tsquery) AS IMPLICIT;
20+
1021
CREATE FUNCTION rum_ts_distance(tsvector,tsquery)
1122
RETURNS float4
12-
AS 'MODULE_PATHNAME'
23+
AS 'MODULE_PATHNAME', 'rum_ts_distance_tt'
24+
LANGUAGE C IMMUTABLE STRICT;
25+
26+
CREATE FUNCTION rum_ts_distance(tsvector,tsquery,int)
27+
RETURNS float4
28+
AS 'MODULE_PATHNAME', 'rum_ts_distance_ttf'
29+
LANGUAGE C IMMUTABLE STRICT;
30+
31+
CREATE FUNCTION rum_ts_distance(tsvector,rum_distance_query)
32+
RETURNS float4
33+
AS 'MODULE_PATHNAME', 'rum_ts_distance_td'
1334
LANGUAGE C IMMUTABLE STRICT;
1435

1536
CREATE OPERATOR <=> (
@@ -18,6 +39,12 @@ CREATE OPERATOR <=> (
1839
PROCEDURE = rum_ts_distance
1940
);
2041

42+
CREATE OPERATOR <=> (
43+
LEFTARG = tsvector,
44+
RIGHTARG = rum_distance_query,
45+
PROCEDURE = rum_ts_distance
46+
);
47+
2148
CREATE FUNCTION rum_extract_tsvector(tsvector,internal,internal,internal,internal)
2249
RETURNS internal
2350
AS 'MODULE_PATHNAME'

‎rum.h

Lines changed: 18 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -663,6 +663,13 @@ typedef struct
663663
bool recheck;
664664
} RumOrderingItem;
665665

666+
typedef enum
667+
{
668+
RumFastScan,
669+
RumRegularScan,
670+
RumFullScan
671+
} RumScanType;
672+
666673
typedef struct RumScanOpaqueData
667674
{
668675
MemoryContext tempCtx;
@@ -684,7 +691,7 @@ typedef struct RumScanOpaqueData
684691
RumKey key;
685692
bool firstCall;
686693
bool isVoidRes; /* true if query is unsatisfiable */
687-
bool useFastScan;
694+
RumScanType scanType;
688695
TIDBitmap *tbm;
689696

690697
ScanDirection naturalOrder;
@@ -760,7 +767,12 @@ extern Datum rum_extract_tsquery(PG_FUNCTION_ARGS);
760767
extern Datum rum_tsvector_config(PG_FUNCTION_ARGS);
761768
extern Datum rum_tsquery_pre_consistent(PG_FUNCTION_ARGS);
762769
extern Datum rum_tsquery_distance(PG_FUNCTION_ARGS);
763-
extern Datum rum_ts_distance(PG_FUNCTION_ARGS);
770+
extern Datum rum_ts_distance_tt(PG_FUNCTION_ARGS);
771+
extern Datum rum_ts_distance_ttf(PG_FUNCTION_ARGS);
772+
extern Datum rum_ts_distance_td(PG_FUNCTION_ARGS);
773+
774+
extern Datum tsquery_to_distance_query(PG_FUNCTION_ARGS);
775+
764776

765777
/* GUC parameters */
766778
extern PGDLLIMPORT int RumFuzzySearchLimit;
@@ -870,6 +882,8 @@ rumDataPageLeafRead(Pointer ptr, OffsetNumber attnum, RumKey * item,
870882
{
871883
attr = rumstate->addAttrs[attnum - 1];
872884

885+
Assert(attr);
886+
873887
if (attr->attbyval)
874888
{
875889
/* do not use aligment for pass-by-value types */
@@ -949,6 +963,8 @@ rumDataPageLeafReadPointer(Pointer ptr, OffsetNumber attnum, RumKey * item,
949963
{
950964
attr = rumstate->addAttrs[attnum - 1];
951965

966+
Assert(attr);
967+
952968
if (!attr->attbyval)
953969
ptr = (Pointer) att_align_pointer(ptr, attr->attalign, attr->attlen,
954970
ptr);

0 commit comments

Comments
 (0)