Skip to content

Commit a9bd6b7

Browse files
xudong963xudong963
committed
add slt test
1 parent 8edce7e commit a9bd6b7

File tree

1 file changed

+212
-0
lines changed

1 file changed

+212
-0
lines changed

‎datafusion/sqllogictest/test_files/simplify_predicates.slt

Lines changed: 212 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,212 @@
1+
# Licensed to the Apache Software Foundation (ASF) under one
2+
# or more contributor license agreements. See the NOTICE file
3+
# distributed with this work for additional information
4+
# regarding copyright ownership. The ASF licenses this file
5+
# to you under the Apache License, Version 2.0 (the
6+
# "License"); you may not use this file except in compliance
7+
# with the License. You may obtain a copy of the License at
8+
9+
# http://www.apache.org/licenses/LICENSE-2.0
10+
11+
# Unless required by applicable law or agreed to in writing,
12+
# software distributed under the License is distributed on an
13+
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
14+
# KIND, either express or implied. See the License for the
15+
# specific language governing permissions and limitations
16+
# under the License.
17+
18+
# Test cases for predicate simplification feature
19+
# Basic redundant comparison simplification
20+
21+
statement ok
22+
set datafusion.explain.logical_plan_only=true;
23+
24+
statement ok
25+
CREATE TABLE test_data (
26+
int_col INT,
27+
float_col FLOAT,
28+
str_col VARCHAR,
29+
date_col DATE,
30+
bool_col BOOLEAN
31+
);
32+
33+
# x > 5 AND x > 6 should simplify to x > 6
34+
query TT
35+
EXPLAIN SELECT * FROM test_data WHERE int_col > 5 AND int_col > 6;
36+
----
37+
logical_plan
38+
01)Filter: test_data.int_col > Int32(6)
39+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
40+
41+
# x > 5 AND x >= 6 should simplify to x >= 6
42+
query TT
43+
EXPLAIN SELECT * FROM test_data WHERE int_col > 5 AND int_col >= 6;
44+
----
45+
logical_plan
46+
01)Filter: test_data.int_col >= Int32(6)
47+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
48+
49+
# x < 10 AND x <= 8 should simplify to x <= 8
50+
query TT
51+
EXPLAIN SELECT * FROM test_data WHERE int_col < 10 AND int_col <= 8;
52+
----
53+
logical_plan
54+
01)Filter: test_data.int_col <= Int32(8)
55+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
56+
57+
# x > 5 AND x > 6 AND x > 7 should simplify to x > 7
58+
query TT
59+
EXPLAIN SELECT * FROM test_data WHERE int_col > 5 AND int_col > 6 AND int_col > 7;
60+
----
61+
logical_plan
62+
01)Filter: test_data.int_col > Int32(7)
63+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
64+
65+
# x > 5 AND y < 10 AND x > 6 AND y < 8 should simplify to x > 6 AND y < 8
66+
query TT
67+
EXPLAIN SELECT * FROM test_data WHERE int_col > 5 AND float_col < 10 AND int_col > 6 AND float_col < 8;
68+
----
69+
logical_plan
70+
01)Filter: test_data.float_col < Float32(8) AND test_data.int_col > Int32(6)
71+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
72+
73+
74+
# x = 7 AND x > 5 should simplify to x = 7
75+
query TT
76+
EXPLAIN SELECT * FROM test_data WHERE int_col = 7 AND int_col > 5;
77+
----
78+
logical_plan
79+
01)Filter: test_data.int_col = Int32(7)
80+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
81+
82+
# str_col > 'apple' AND str_col > 'banana' should simplify to str_col > 'banana'
83+
query TT
84+
EXPLAIN SELECT * FROM test_data WHERE str_col > 'apple' AND str_col > 'banana';
85+
----
86+
logical_plan
87+
01)Filter: test_data.str_col > Utf8("banana")
88+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
89+
90+
# date_col > '2023-01-01' AND date_col > '2023-02-01' should simplify to date_col > '2023-02-01'
91+
query TT
92+
EXPLAIN SELECT * FROM test_data WHERE date_col > '2023-01-01' AND date_col > '2023-02-01';
93+
----
94+
logical_plan
95+
01)Filter: test_data.date_col > Date32("2023-02-01")
96+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
97+
98+
query TT
99+
EXPLAIN SELECT * FROM test_data WHERE bool_col = true AND bool_col = false;
100+
----
101+
logical_plan
102+
01)Filter: test_data.bool_col AND NOT test_data.bool_col
103+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
104+
105+
106+
# This shouldn't be simplified since they're different relationships
107+
query TT
108+
EXPLAIN SELECT * FROM test_data WHERE int_col > float_col AND int_col > 5;
109+
----
110+
logical_plan
111+
01)Filter: CAST(test_data.int_col AS Float32) > test_data.float_col AND test_data.int_col > Int32(5)
112+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
113+
114+
# Should simplify the int_col predicates but preserve the others
115+
query TT
116+
EXPLAIN SELECT * FROM test_data
117+
WHERE int_col > 5
118+
AND int_col > 10
119+
AND str_col LIKE 'A%'
120+
AND float_col BETWEEN 1 AND 100;
121+
----
122+
logical_plan
123+
01)Filter: test_data.str_col LIKE Utf8("A%") AND test_data.float_col >= Float32(1) AND test_data.float_col <= Float32(100) AND test_data.int_col > Int32(10)
124+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
125+
126+
statement ok
127+
CREATE TABLE test_data2 (
128+
id INT,
129+
value INT
130+
);
131+
132+
query TT
133+
EXPLAIN SELECT t1.int_col, t2.value
134+
FROM test_data t1
135+
JOIN test_data2 t2 ON t1.int_col = t2.id
136+
WHERE t1.int_col > 5
137+
AND t1.int_col > 10
138+
AND t2.value < 100
139+
AND t2.value < 50;
140+
----
141+
logical_plan
142+
01)Projection: t1.int_col, t2.value
143+
02)--Inner Join: t1.int_col = t2.id
144+
03)----SubqueryAlias: t1
145+
04)------Filter: test_data.int_col > Int32(10)
146+
05)--------TableScan: test_data projection=[int_col]
147+
06)----SubqueryAlias: t2
148+
07)------Filter: test_data2.value < Int32(50) AND test_data2.id > Int32(10)
149+
08)--------TableScan: test_data2 projection=[id, value]
150+
151+
# Case 13: Handling negated predicates
152+
# NOT (x < 10) AND NOT (x < 5) should simplify to NOT (x < 10)
153+
query TT
154+
EXPLAIN SELECT * FROM test_data WHERE NOT (int_col < 10) AND NOT (int_col < 5);
155+
----
156+
logical_plan
157+
01)Filter: test_data.int_col >= Int32(10)
158+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
159+
160+
# x > 5 AND x < 10 should be preserved (can't be simplified)
161+
query TT
162+
EXPLAIN SELECT * FROM test_data WHERE int_col > 5 AND int_col < 10;
163+
----
164+
logical_plan
165+
01)Filter: test_data.int_col > Int32(5) AND test_data.int_col < Int32(10)
166+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
167+
168+
# 5 < x AND 3 < x should simplify to 5 < x
169+
query TT
170+
EXPLAIN SELECT * FROM test_data WHERE 5 < int_col AND 3 < int_col;
171+
----
172+
logical_plan
173+
01)Filter: test_data.int_col > Int32(5)
174+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
175+
176+
# CAST(x AS FLOAT) > 5.0 AND CAST(x AS FLOAT) > 6.0 should simplify
177+
query TT
178+
EXPLAIN SELECT * FROM test_data WHERE CAST(int_col AS FLOAT) > 5.0 AND CAST(int_col AS FLOAT) > 6.0;
179+
----
180+
logical_plan
181+
01)Filter: CAST(CAST(test_data.int_col AS Float32) AS Float64) > Float64(6)
182+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
183+
184+
# x = 5 AND x = 6 (logically impossible)
185+
query TT
186+
EXPLAIN SELECT * FROM test_data WHERE int_col = 5 AND int_col = 6;
187+
----
188+
logical_plan EmptyRelation
189+
190+
# (x > 5 OR y < 10) AND (x > 6 OR y < 8)
191+
# This is more complex but could still benefit from some simplification
192+
query TT
193+
EXPLAIN SELECT * FROM test_data
194+
WHERE (int_col > 5 OR float_col < 10)
195+
AND (int_col > 6 OR float_col < 8);
196+
----
197+
logical_plan
198+
01)Filter: (test_data.int_col > Int32(5) OR test_data.float_col < Float32(10)) AND (test_data.int_col > Int32(6) OR test_data.float_col < Float32(8))
199+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
200+
201+
# Case 20: Combination of AND and OR with simplifiable predicates
202+
query TT
203+
EXPLAIN SELECT * FROM test_data
204+
WHERE (int_col > 5 AND int_col > 6)
205+
OR (float_col < 10 AND float_col < 8);
206+
----
207+
logical_plan
208+
01)Filter: test_data.int_col > Int32(5) AND test_data.int_col > Int32(6) OR test_data.float_col < Float32(10) AND test_data.float_col < Float32(8)
209+
02)--TableScan: test_data projection=[int_col, float_col, str_col, date_col, bool_col]
210+
211+
statement ok
212+
set datafusion.explain.logical_plan_only=false;

0 commit comments

Comments
 (0)