Merge pull request #19937 from ClickHouse/backport/20.12/19884

Backport #19884 to 20.12: BloomFilter index crash fix

src/Interpreters/Set.cpp

@@ -140,7 +140,16 @@ void Set::setHeader(const Block & header)
     ConstNullMapPtr null_map{};
     ColumnPtr null_map_holder;
     if (!transform_null_in)
+    {
+        /// We convert nullable columns to non nullable we also need to update nullable types
+        for (size_t i = 0; i < set_elements_types.size(); ++i)
+        {
+            data_types[i] = removeNullable(data_types[i]);
+            set_elements_types[i] = removeNullable(set_elements_types[i]);
+        }
+
         extractNestedColumnsAndNullMap(key_columns, null_map);
+    }
 
     if (fill_set_elements)
     {
@@ -182,7 +191,7 @@ bool Set::insertFromBlock(const Block & block)
     ConstNullMapPtr null_map{};
     ColumnPtr null_map_holder;
     if (!transform_null_in)
-         null_map_holder = extractNestedColumnsAndNullMap(key_columns, null_map);
+        null_map_holder = extractNestedColumnsAndNullMap(key_columns, null_map);
 
     /// Filter to extract distinct values from the block.
     ColumnUInt8::MutablePtr filter;

tests/queries/0_stateless/01681_bloom_filter_nullable_column.reference

+NullableTuple with transform_null_in=0
+NullableTuple with transform_null_in=1
+NullableColumnFromCast with transform_null_in=0
+1	test
+NullableColumnFromCast with transform_null_in=1
+1	test
+NullableColumnFromTable with transform_null_in=0
+1	test
+NullableColumnFromTable with transform_null_in=1
+1	test

tests/queries/0_stateless/01681_bloom_filter_nullable_column.sql

+DROP TABLE IF EXISTS bloom_filter_nullable_index;
+CREATE TABLE bloom_filter_nullable_index
+    (
+        order_key UInt64,
+        str Nullable(String),
+
+        INDEX idx (str) TYPE bloom_filter GRANULARITY 1
+    )
+    ENGINE = MergeTree() 
+    ORDER BY order_key SETTINGS index_granularity = 6;
+
+INSERT INTO bloom_filter_nullable_index VALUES (1, 'test');
+INSERT INTO bloom_filter_nullable_index VALUES (2, 'test2');
+
+SELECT 'NullableTuple with transform_null_in=0';
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT '1048576', str FROM bloom_filter_nullable_index) SETTINGS transform_null_in = 0;
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT '1048576', str FROM bloom_filter_nullable_index) SETTINGS transform_null_in = 0;
+
+SELECT 'NullableTuple with transform_null_in=1';
+
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT '1048576', str FROM bloom_filter_nullable_index) SETTINGS transform_null_in = 1; -- { serverError 20 }
+
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT '1048576', str FROM bloom_filter_nullable_index) SETTINGS transform_null_in = 1; -- { serverError 20 }
+
+
+SELECT 'NullableColumnFromCast with transform_null_in=0';
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT cast('test', 'Nullable(String)')) SETTINGS transform_null_in = 0;
+
+SELECT 'NullableColumnFromCast with transform_null_in=1';
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT cast('test', 'Nullable(String)')) SETTINGS transform_null_in = 1;
+
+DROP TABLE IF EXISTS nullable_string_value;
+CREATE TABLE nullable_string_value (value Nullable(String)) ENGINE=TinyLog;
+INSERT INTO nullable_string_value VALUES ('test');
+
+SELECT 'NullableColumnFromTable with transform_null_in=0';
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT value FROM nullable_string_value) SETTINGS transform_null_in = 0;
+
+SELECT 'NullableColumnFromTable with transform_null_in=1';
+SELECT * FROM bloom_filter_nullable_index WHERE str IN
+    (SELECT value FROM nullable_string_value) SETTINGS transform_null_in = 1;
+
+DROP TABLE nullable_string_value; 
+DROP TABLE bloom_filter_nullable_index;