This is an automated email from the ASF dual-hosted git repository.
krisztiankasa pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/hive.git
The following commit(s) were added to refs/heads/master by this push:
new 08067725bc6 HIVE-28917: NPE in merge statement when checking
nullability of joini… (#5781)
08067725bc6 is described below
commit 08067725bc6e8810579324736a0aac453c06bf7b
Author: Soumyakanti Das <[email protected]>
AuthorDate: Thu May 15 01:43:35 2025 -0700
HIVE-28917: NPE in merge statement when checking nullability of joini…
(#5781)
---
.../merge_with_null_check_on_joining_col.q | 39 +++
.../merge_with_null_check_on_joining_col.q.out | 306 +++++++++++++++++++++
ql/src/java/org/apache/hadoop/hive/ql/Context.java | 12 +-
3 files changed, 355 insertions(+), 2 deletions(-)
diff --git
a/iceberg/iceberg-handler/src/test/queries/positive/merge_with_null_check_on_joining_col.q
b/iceberg/iceberg-handler/src/test/queries/positive/merge_with_null_check_on_joining_col.q
new file mode 100644
index 00000000000..b32670608fe
--- /dev/null
+++
b/iceberg/iceberg-handler/src/test/queries/positive/merge_with_null_check_on_joining_col.q
@@ -0,0 +1,39 @@
+
+create table target(a int, b int, c int) stored by iceberg
tblproperties('format-version'='2', 'write.merge.mode'='copy-on-write');
+create table source(a int, b int, c int) stored by iceberg
tblproperties('format-version'='2', 'write.merge.mode'='copy-on-write');
+
+-- empty plan as joining column cannot be null for matched clause
+explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then delete;
+
+explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then update set b = t.b + 10;
+
+-- non empty plans as condition for all branches are not false
+-- matched and <joining col> is null is always false
+-- not matched and <joining col> is [not] null is not necessarily always false
+explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then delete
+when matched and t.a > 10 then update set b = t.b + 100
+when not matched and s.a > 20 then insert values (s.a, s.b, s.c);
+
+explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a is null then update set b = t.b + 100
+when not matched and s.a > 20 then insert values (s.a, s.b, s.c);
+
+explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a > 20 then update set b = t.b + 100
+when not matched and s.a is null then insert values (s.a, s.b, s.c);
+
+explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a > 20 then update set b = t.b + 100
+when not matched and s.a is not null then insert values (s.a, s.b, s.c);
diff --git
a/iceberg/iceberg-handler/src/test/results/positive/merge_with_null_check_on_joining_col.q.out
b/iceberg/iceberg-handler/src/test/results/positive/merge_with_null_check_on_joining_col.q.out
new file mode 100644
index 00000000000..521eb4e26f2
--- /dev/null
+++
b/iceberg/iceberg-handler/src/test/results/positive/merge_with_null_check_on_joining_col.q.out
@@ -0,0 +1,306 @@
+PREHOOK: query: create table target(a int, b int, c int) stored by iceberg
tblproperties('format-version'='2', 'write.merge.mode'='copy-on-write')
+PREHOOK: type: CREATETABLE
+PREHOOK: Output: database:default
+PREHOOK: Output: default@target
+POSTHOOK: query: create table target(a int, b int, c int) stored by iceberg
tblproperties('format-version'='2', 'write.merge.mode'='copy-on-write')
+POSTHOOK: type: CREATETABLE
+POSTHOOK: Output: database:default
+POSTHOOK: Output: default@target
+PREHOOK: query: create table source(a int, b int, c int) stored by iceberg
tblproperties('format-version'='2', 'write.merge.mode'='copy-on-write')
+PREHOOK: type: CREATETABLE
+PREHOOK: Output: database:default
+PREHOOK: Output: default@source
+POSTHOOK: query: create table source(a int, b int, c int) stored by iceberg
tblproperties('format-version'='2', 'write.merge.mode'='copy-on-write')
+POSTHOOK: type: CREATETABLE
+POSTHOOK: Output: database:default
+POSTHOOK: Output: default@source
+PREHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then delete
+PREHOOK: type: QUERY
+PREHOOK: Input: _dummy_database@_dummy_table
+PREHOOK: Input: default@source
+PREHOOK: Input: default@target
+PREHOOK: Output: default@target
+POSTHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then delete
+POSTHOOK: type: QUERY
+POSTHOOK: Input: _dummy_database@_dummy_table
+POSTHOOK: Input: default@source
+POSTHOOK: Input: default@target
+POSTHOOK: Output: default@target
+CBO PLAN:
+HiveValues(tuples=[[]])
+
+PREHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then update set b = t.b + 10
+PREHOOK: type: QUERY
+PREHOOK: Input: _dummy_database@_dummy_table
+PREHOOK: Input: default@source
+PREHOOK: Input: default@target
+PREHOOK: Output: default@target
+POSTHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then update set b = t.b + 10
+POSTHOOK: type: QUERY
+POSTHOOK: Input: _dummy_database@_dummy_table
+POSTHOOK: Input: default@source
+POSTHOOK: Input: default@target
+POSTHOOK: Output: default@target
+CBO PLAN:
+HiveValues(tuples=[[]])
+
+PREHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then delete
+when matched and t.a > 10 then update set b = t.b + 100
+when not matched and s.a > 20 then insert values (s.a, s.b, s.c)
+PREHOOK: type: QUERY
+PREHOOK: Input: default@source
+PREHOOK: Input: default@target
+PREHOOK: Output: default@target
+POSTHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a is null then delete
+when matched and t.a > 10 then update set b = t.b + 100
+when not matched and s.a > 20 then insert values (s.a, s.b, s.c)
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@source
+POSTHOOK: Input: default@target
+POSTHOOK: Output: default@target
+CBO PLAN:
+HiveUnion(all=[true])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[+($6, 100)], t__c=[$7])
+ HiveJoin(condition=[AND(=($5, $8), =($6, $9))], joinType=[inner],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1], t__c=[$2])
+ HiveFilter(condition=[AND(>($0, 10), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(>($0, 10), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
a=[$7], b=[$8], c=[$9])
+ HiveFilter(condition=[AND(IS NULL($5), IS NULL($6))])
+ HiveJoin(condition=[AND(=($5, $7), =($6, $8))], joinType=[right],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1])
+ HiveFilter(condition=[AND(>($0, 20), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1], c=[$2])
+ HiveFilter(condition=[>($0, 20)])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7])
+ HiveSemiJoin(condition=[=($2, $10)], joinType=[semi])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7], a=[$8], b=[$9])
+ HiveFilter(condition=[OR(IS NULL(OR(AND(=($5, $8), =($6, $9), IS
NULL($5)), >($5, 10), AND(IS NULL($5), IS NULL($6)))), AND(OR(<>($5, $8),
<>($6, $9), IS NOT NULL($5)), <=($5, 10), OR(IS NOT NULL($5), IS NOT
NULL($6))))])
+ HiveJoin(condition=[AND(=($5, $8), =($6, $9))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1], t__c=[$2])
+ HiveFilter(condition=[IS NOT NULL($9)])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__file__path=[$0])
+ HiveFilter(condition=[=($1, 1)])
+ HiveProject(t__file__path=[$0], row_number_window_0=[row_number()
OVER (PARTITION BY $0 ORDER BY $0 NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING
AND UNBOUNDED FOLLOWING)])
+ HiveJoin(condition=[AND(=($1, $3), =($2, $4))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__file__path=[$9], t__a=[$0], t__b=[$1])
+ HiveFilter(condition=[AND(>($0, 10), IS NOT NULL($9))])
+ HiveTableScan(table=[[default, target]],
table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(>($0, 10), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], _o__c3=[-1:BIGINT], t__partition__projection=[$3],
t__a=[$4], t__b=[$5], t__c=[$6])
+ HiveFilter(condition=[=($7, 1)])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__partition__projection=[$3], t__a=[$4], t__b=[$5],
t__c=[$6], row_number_window_0=[row_number() OVER (PARTITION BY $2 ORDER BY $2
NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING)])
+ HiveJoin(condition=[AND(=($4, $7), =($5, $8))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__partition__projection=[$11], t__a=[$0], t__b=[$1],
t__c=[$2])
+ HiveFilter(condition=[>($0, 10)])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(>($0, 10), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+
+PREHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a is null then update set b = t.b + 100
+when not matched and s.a > 20 then insert values (s.a, s.b, s.c)
+PREHOOK: type: QUERY
+PREHOOK: Input: default@source
+PREHOOK: Input: default@target
+PREHOOK: Output: default@target
+POSTHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a is null then update set b = t.b + 100
+when not matched and s.a > 20 then insert values (s.a, s.b, s.c)
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@source
+POSTHOOK: Input: default@target
+POSTHOOK: Output: default@target
+CBO PLAN:
+HiveUnion(all=[true])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
a=[$7], b=[$8], c=[$9])
+ HiveFilter(condition=[AND(IS NULL($5), IS NULL($6))])
+ HiveJoin(condition=[AND(=($5, $7), =($6, $8))], joinType=[right],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1])
+ HiveFilter(condition=[AND(>($0, 20), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1], c=[$2])
+ HiveFilter(condition=[>($0, 20)])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7])
+ HiveSemiJoin(condition=[=($2, $10)], joinType=[semi])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7], a=[$8], b=[$9])
+ HiveFilter(condition=[OR(IS NULL(OR(IS NULL($5), AND(=($5, $8), =($6,
$9), >($5, 10)), AND(IS NULL($5), IS NULL($6)))), AND(OR(<>($5, $8), <>($6,
$9), <=($5, 10)), IS NOT NULL($5)))])
+ HiveJoin(condition=[AND(=($5, $8), =($6, $9))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1], t__c=[$2])
+ HiveFilter(condition=[IS NOT NULL($9)])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__file__path=[$0])
+ HiveFilter(condition=[=($1, 1)])
+ HiveProject(t__file__path=[$0], row_number_window_0=[row_number()
OVER (PARTITION BY $0 ORDER BY $0 NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING
AND UNBOUNDED FOLLOWING)])
+ HiveFilter(condition=[OR(IS NULL($1), AND(=($1, $3), =($2, $4),
>($1, 10)))])
+ HiveJoin(condition=[AND(=($1, $3), =($2, $4))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__file__path=[$9], t__a=[$0], t__b=[$1])
+ HiveFilter(condition=[IS NOT NULL($9)])
+ HiveTableScan(table=[[default, target]],
table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], _o__c3=[-1:BIGINT], t__partition__projection=[$3],
t__a=[$4], t__b=[$5], t__c=[$6])
+ HiveFilter(condition=[=($7, 1)])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__partition__projection=[$3], t__a=[$4], t__b=[$5],
t__c=[$6], row_number_window_0=[row_number() OVER (PARTITION BY $2 ORDER BY $2
NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING)])
+ HiveFilter(condition=[OR(IS NULL($4), AND(=($4, $7), =($5, $8), >($4,
10)))])
+ HiveJoin(condition=[AND(=($4, $7), =($5, $8))], joinType=[full],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__partition__projection=[$11], t__a=[$0], t__b=[$1],
t__c=[$2])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+
+PREHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a > 20 then update set b = t.b + 100
+when not matched and s.a is null then insert values (s.a, s.b, s.c)
+PREHOOK: type: QUERY
+PREHOOK: Input: default@source
+PREHOOK: Input: default@target
+PREHOOK: Output: default@target
+POSTHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a > 20 then update set b = t.b + 100
+when not matched and s.a is null then insert values (s.a, s.b, s.c)
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@source
+POSTHOOK: Input: default@target
+POSTHOOK: Output: default@target
+CBO PLAN:
+HiveUnion(all=[true])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
a=[null:INTEGER], b=[$8], c=[$9])
+ HiveFilter(condition=[AND(IS NULL($5), IS NULL($6), IS NULL($7))])
+ HiveJoin(condition=[AND(=($5, $7), =($6, $8))], joinType=[full],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1], c=[$2])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7])
+ HiveSemiJoin(condition=[=($2, $10)], joinType=[semi])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7], a=[$8], b=[$9])
+ HiveFilter(condition=[OR(IS NULL(OR(AND(=($5, $8), =($6, $9), >($5,
10)), >($5, 20), AND(IS NULL($5), IS NULL($6)))), AND(OR(<>($5, $8), <>($6,
$9), <=($5, 10)), <=($5, 20), OR(IS NOT NULL($5), IS NOT NULL($6))))])
+ HiveJoin(condition=[AND(=($5, $8), =($6, $9))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1], t__c=[$2])
+ HiveFilter(condition=[IS NOT NULL($9)])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__file__path=[$0])
+ HiveFilter(condition=[=($1, 1)])
+ HiveProject(t__file__path=[$0], row_number_window_0=[row_number()
OVER (PARTITION BY $0 ORDER BY $0 NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING
AND UNBOUNDED FOLLOWING)])
+ HiveFilter(condition=[OR(>($1, 20), AND(=($1, $3), =($2, $4),
>($1, 10)))])
+ HiveJoin(condition=[AND(=($1, $3), =($2, $4))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__file__path=[$9], t__a=[$0], t__b=[$1])
+ HiveFilter(condition=[IS NOT NULL($9)])
+ HiveTableScan(table=[[default, target]],
table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], _o__c3=[-1:BIGINT], t__partition__projection=[$3],
t__a=[$4], t__b=[$5], t__c=[$6])
+ HiveFilter(condition=[=($7, 1)])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__partition__projection=[$3], t__a=[$4], t__b=[$5],
t__c=[$6], row_number_window_0=[row_number() OVER (PARTITION BY $2 ORDER BY $2
NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING)])
+ HiveFilter(condition=[OR(>($4, 20), AND(=($4, $7), =($5, $8), >($4,
10)))])
+ HiveJoin(condition=[AND(=($4, $7), =($5, $8))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__partition__projection=[$11], t__a=[$0], t__b=[$1],
t__c=[$2])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+
+PREHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a > 20 then update set b = t.b + 100
+when not matched and s.a is not null then insert values (s.a, s.b, s.c)
+PREHOOK: type: QUERY
+PREHOOK: Input: default@source
+PREHOOK: Input: default@target
+PREHOOK: Output: default@target
+POSTHOOK: query: explain cbo
+merge into target as t using source as s on t.a = s.a and t.b = s.b
+when matched and t.a > 10 then delete
+when matched and t.a > 20 then update set b = t.b + 100
+when not matched and s.a is not null then insert values (s.a, s.b, s.c)
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@source
+POSTHOOK: Input: default@target
+POSTHOOK: Output: default@target
+CBO PLAN:
+HiveUnion(all=[true])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
a=[$7], b=[$8], c=[$9])
+ HiveFilter(condition=[AND(IS NULL($5), IS NULL($6))])
+ HiveJoin(condition=[AND(=($5, $7), =($6, $8))], joinType=[right],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1], c=[$2])
+ HiveFilter(condition=[IS NOT NULL($0)])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7])
+ HiveSemiJoin(condition=[=($2, $10)], joinType=[semi])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__row__position=[$3], t__partition__projection=[$4],
t__a=[$5], t__b=[$6], t__c=[$7], a=[$8], b=[$9])
+ HiveFilter(condition=[OR(IS NULL(OR(AND(=($5, $8), =($6, $9), >($5,
10)), >($5, 20), AND(IS NULL($5), IS NULL($6)))), AND(OR(<>($5, $8), <>($6,
$9), <=($5, 10)), <=($5, 20), OR(IS NOT NULL($5), IS NOT NULL($6))))])
+ HiveJoin(condition=[AND(=($5, $8), =($6, $9))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__row__position=[$10], t__partition__projection=[$11],
t__a=[$0], t__b=[$1], t__c=[$2])
+ HiveFilter(condition=[IS NOT NULL($9)])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__file__path=[$0])
+ HiveFilter(condition=[=($1, 1)])
+ HiveProject(t__file__path=[$0], row_number_window_0=[row_number()
OVER (PARTITION BY $0 ORDER BY $0 NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING
AND UNBOUNDED FOLLOWING)])
+ HiveFilter(condition=[OR(>($1, 20), AND(=($1, $3), =($2, $4),
>($1, 10)))])
+ HiveJoin(condition=[AND(=($1, $3), =($2, $4))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__file__path=[$9], t__a=[$0], t__b=[$1])
+ HiveFilter(condition=[IS NOT NULL($9)])
+ HiveTableScan(table=[[default, target]],
table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], _o__c3=[-1:BIGINT], t__partition__projection=[$3],
t__a=[$4], t__b=[$5], t__c=[$6])
+ HiveFilter(condition=[=($7, 1)])
+ HiveProject(t__partition__spec__id=[$0], t__partition__hash=[$1],
t__file__path=[$2], t__partition__projection=[$3], t__a=[$4], t__b=[$5],
t__c=[$6], row_number_window_0=[row_number() OVER (PARTITION BY $2 ORDER BY $2
NULLS FIRST RANGE BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING)])
+ HiveFilter(condition=[OR(>($4, 20), AND(=($4, $7), =($5, $8), >($4,
10)))])
+ HiveJoin(condition=[AND(=($4, $7), =($5, $8))], joinType=[left],
algorithm=[none], cost=[not available])
+ HiveProject(t__partition__spec__id=[$7], t__partition__hash=[$8],
t__file__path=[$9], t__partition__projection=[$11], t__a=[$0], t__b=[$1],
t__c=[$2])
+ HiveTableScan(table=[[default, target]], table:alias=[target])
+ HiveProject(a=[$0], b=[$1])
+ HiveFilter(condition=[AND(IS NOT NULL($0), IS NOT NULL($1))])
+ HiveTableScan(table=[[default, source]], table:alias=[s])
+
diff --git a/ql/src/java/org/apache/hadoop/hive/ql/Context.java
b/ql/src/java/org/apache/hadoop/hive/ql/Context.java
index 2e6df97c152..138787a733e 100644
--- a/ql/src/java/org/apache/hadoop/hive/ql/Context.java
+++ b/ql/src/java/org/apache/hadoop/hive/ql/Context.java
@@ -361,8 +361,16 @@ private DestClausePrefix getMergeDestClausePrefix(ASTNode
curNode) {
assert insert != null && insert.getType() == HiveParser.TOK_INSERT;
ASTNode query = (ASTNode) insert.getParent();
assert query != null && query.getType() == HiveParser.TOK_QUERY;
-
- int tokFromIdx =
query.getFirstChildWithType(HiveParser.TOK_FROM).getChildIndex();
+ ASTNode from = (ASTNode) query.getFirstChildWithType(HiveParser.TOK_FROM);
+
+ if (from == null) {
+ // We are here when TOK_FROM is missing from the AST.
+ // This can happen for merge queries with a predicate like
`<joining_column> is null`
+ // in the matched clause.
+ return DestClausePrefix.MERGE;
+ }
+
+ int tokFromIdx = from.getChildIndex();
for (int childIdx = tokFromIdx + 1; childIdx < query.getChildCount();
childIdx++) {
assert query.getChild(childIdx).getType() == HiveParser.TOK_INSERT;
if (insert == query.getChild(childIdx)) {