mirror of
https://github.com/apache/impala.git
synced 2025-12-30 21:02:41 -05:00
77304530f1
Implements constant propagation within conjuncts and applies the optimization to scan conjuncts and collection conjuncts within Hdfs scan nodes. The optimization is applied during planning. At scan nodes in particular, we want to optimize to enable partition pruning. In certain cases, we might end up with a FALSE conditional, which now will convert to an EmptySet node. Testing: Expanded the test cases for the planner to achieve constant propagation. Added Kudu, datasource, Hdfs and HBase tests to validate we can create EmptySetNodes. Change-Id: I79750a8edb945effee2a519fa3b8192b77042cb4 Reviewed-on: http://gerrit.cloudera.org:8080/6389 Tested-by: Impala Public Jenkins Reviewed-by: Alex Behm <alex.behm@cloudera.com>
132 lines
4.1 KiB
Plaintext
132 lines
4.1 KiB
Plaintext
====
|
|
---- QUERY
|
|
# Gets all types including a row with a NULL value. The predicate pushed to
|
|
# the data source is not actually used, but the second predicate is
|
|
# evaluated by Impala.
|
|
select *
|
|
from alltypes_datasource
|
|
where float_col != 0 and
|
|
int_col >= 1990 limit 5
|
|
---- RESULTS
|
|
1990,true,0,90,1990,19900,2189,1990,1970-01-01 00:00:01.990000000,'NULL',-999998009,-9999998009,-9999999999.9999998009,-9.9999999999999999999999999999999998009,-99999.98009
|
|
1991,false,1,91,1991,19910,2190.10009765625,1991,1970-01-01 00:00:01.991000000,'1991',999998008,9999998008,9999999999.9999998008,9.9999999999999999999999999999999998008,99999.98008
|
|
1992,true,2,92,1992,19920,2191.199951171875,1992,1970-01-01 00:00:01.992000000,'1992',-999998007,-9999998007,-9999999999.9999998007,-9.9999999999999999999999999999999998007,-99999.98007
|
|
1993,false,3,93,1993,19930,2192.300048828125,1993,1970-01-01 00:00:01.993000000,'1993',999998006,9999998006,9999999999.9999998006,9.9999999999999999999999999999999998006,99999.98006
|
|
1994,true,4,94,1994,19940,2193.39990234375,1994,1970-01-01 00:00:01.994000000,'1994',-999998005,-9999998005,-9999999999.9999998005,-9.9999999999999999999999999999999998005,-99999.98005
|
|
---- TYPES
|
|
INT, BOOLEAN, TINYINT, SMALLINT, INT, BIGINT, FLOAT, DOUBLE, TIMESTAMP, STRING, DECIMAL, DECIMAL, DECIMAL, DECIMAL, DECIMAL
|
|
====
|
|
---- QUERY
|
|
# Project a subset of the columns
|
|
select bigint_col, timestamp_col, double_col
|
|
from alltypes_datasource
|
|
where double_col != 0 and int_col >= 1990 limit 3
|
|
---- RESULTS
|
|
19900,1970-01-01 00:00:01.990000000,1990
|
|
19910,1970-01-01 00:00:01.991000000,1991
|
|
19920,1970-01-01 00:00:01.992000000,1992
|
|
---- TYPES
|
|
BIGINT, TIMESTAMP, DOUBLE
|
|
====
|
|
---- QUERY
|
|
# count(*) with a predicate evaluated by Impala
|
|
select count(*) from alltypes_datasource
|
|
where float_col = 0 and
|
|
string_col is not NULL
|
|
---- RESULTS
|
|
4000
|
|
---- TYPES
|
|
BIGINT
|
|
====
|
|
---- QUERY
|
|
# count(*) with no predicates has no materialized slots
|
|
select count(*) from alltypes_datasource
|
|
---- RESULTS
|
|
5000
|
|
---- TYPES
|
|
BIGINT
|
|
====
|
|
---- QUERY
|
|
select string_col from alltypes_datasource
|
|
where string_col = 'VALIDATE_PREDICATES##id LT 1 && id GT 1 && id LE 1 && id GE 1 && int_col EQ 1 && id NE 1'
|
|
and id < 1 and id > 1 and id <= 1 and id >= 1 and int_col = 1 and id != 1
|
|
---- RESULTS
|
|
'SUCCESS'
|
|
---- TYPES
|
|
STRING
|
|
====
|
|
---- QUERY
|
|
select string_col from alltypes_datasource
|
|
where string_col = 'VALIDATE_PREDICATES##id LT 1 && id GT 1 && id LE 1 && id GE 1 && int_col EQ 1 && id NE 1'
|
|
and 1 > id and 1 < id and 1 >= id and 1 <= id and 1 = int_col and 1 != id
|
|
---- RESULTS
|
|
'SUCCESS'
|
|
---- TYPES
|
|
STRING
|
|
====
|
|
---- QUERY
|
|
# Test that <=>, IS DISTINCT FROM, and IS NOT DISTINCT FROM all can be validated
|
|
# Note the duplicate predicate 1 IS NOT DISTINCT FROM id is removed.
|
|
select string_col from alltypes_datasource
|
|
where string_col = 'VALIDATE_PREDICATES##id NOT_DISTINCT 1 && id DISTINCT_FROM 1'
|
|
and 1 <=> id and 1 IS DISTINCT FROM id and 1 IS NOT DISTINCT FROM id
|
|
---- RESULTS
|
|
'SUCCESS'
|
|
---- TYPES
|
|
STRING
|
|
====
|
|
---- QUERY
|
|
# Test that <=>, IS DISTINCT FROM, and IS NOT DISTINCT FROM are evaluated just like their
|
|
# equality counterparts
|
|
select * from
|
|
(select count(*) from alltypes_datasource
|
|
where tinyint_col = 1 and smallint_col = 11) a
|
|
union all
|
|
(select count(*) from alltypes_datasource
|
|
where tinyint_col <=> 1 and smallint_col <=> 11)
|
|
---- RESULTS
|
|
50
|
|
50
|
|
---- TYPES
|
|
BIGINT
|
|
====
|
|
---- QUERY
|
|
select * from
|
|
(select count(*) from alltypes_datasource
|
|
where smallint_col = 11 and tinyint_col = 1) a
|
|
union all
|
|
(select count(*) from alltypes_datasource
|
|
where smallint_col <=> 11 and tinyint_col <=> 1)
|
|
---- RESULTS
|
|
500
|
|
500
|
|
---- TYPES
|
|
BIGINT
|
|
====
|
|
---- QUERY
|
|
select * from
|
|
(select count(*) from alltypes_datasource
|
|
where tinyint_col != 1 and smallint_col != 11) a
|
|
union all
|
|
(select count(*) from alltypes_datasource
|
|
where tinyint_col IS DISTINCT FROM 1 and smallint_col IS DISTINCT FROM 11)
|
|
---- RESULTS
|
|
4950
|
|
4950
|
|
---- TYPES
|
|
BIGINT
|
|
====
|
|
---- QUERY
|
|
select * from
|
|
(select count(*) from alltypes_datasource
|
|
where smallint_col != 11 and tinyint_col != 1) a
|
|
union all
|
|
(select count(*) from alltypes_datasource
|
|
where smallint_col IS DISTINCT FROM 11 and tinyint_col IS DISTINCT FROM 1)
|
|
---- RESULTS
|
|
4096
|
|
4096
|
|
---- TYPES
|
|
BIGINT
|
|
====
|