thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-09-20 08:40:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #16123 from ucasFL/agg-func-setting-null-for-empty

Browse Source 
Add setting `aggregate_functions_null_for_empty`
This commit is contained in:
alexey-milovidov 2020-11-06 21:42:30 +03:00 committed by GitHub
commit 4a872a5297
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 87 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
src/Core/Settings.h
View File

@ -374,7 +374,6 @@ class IColumn;
M(Bool, optimize_monotonous_functions_in_order_by, true, "Replace monotonous function with its argument in ORDER BY", 0) \
M(Bool, allow_experimental_alter_materialized_view_structure, false, "Allow atomic alter on Materialized views. Work in progress.", 0) \
M(Bool, enable_early_constant_folding, true, "Enable query optimization where we analyze function and subqueries results and rewrite query if there're constants there", 0) \
\
M(Bool, deduplicate_blocks_in_dependent_materialized_views, false, "Should deduplicate blocks for materialized views if the block is not a duplicate for the table. Use true to always deduplicate in dependent tables.", 0) \
M(Bool, use_compact_format_in_distributed_parts_names, false, "Changes format of directories names for distributed table insert parts.", 0) \
M(Bool, validate_polygons, true, "Throw exception if polygon is invalid in function pointInPolygon (e.g. self-tangent, self-intersecting). If the setting is false, the function will accept invalid polygons but may silently return wrong result.", 0) \
@ -385,7 +384,6 @@ class IColumn;
M(Seconds, lock_acquire_timeout, DBMS_DEFAULT_LOCK_ACQUIRE_TIMEOUT_SEC, "How long locking request should wait before failing", 0) \
M(Bool, materialize_ttl_after_modify, true, "Apply TTL for old data, after ALTER MODIFY TTL query", 0) \
M(String, function_implementation, "", "Choose function implementation for specific target or variant (experimental). If empty enable all of them.", 0) \
\
M(Bool, allow_experimental_geo_types, false, "Allow geo data types such as Point, Ring, Polygon, MultiPolygon", 0) \
M(Bool, allow_experimental_bigint_types, false, "Allow Int128, Int256, UInt256 and Decimal256 types", 0) \
M(Bool, data_type_default_nullable, false, "Data types without NULL or NOT NULL will make Nullable", 0) \
@ -394,20 +392,18 @@ class IColumn;
M(Bool, allow_experimental_database_materialize_mysql, false, "Allow to create database with Engine=MaterializeMySQL(...).", 0) \
M(Bool, system_events_show_zero_values, false, "Include all metrics, even with zero values", 0) \
M(MySQLDataTypesSupport, mysql_datatypes_support_level, 0, "Which MySQL types should be converted to corresponding ClickHouse types (rather than being represented as String). Can be empty or any combination of 'decimal' or 'datetime64'. When empty MySQL's DECIMAL and DATETIME/TIMESTAMP with non-zero precision are seen as String on ClickHouse's side.", 0) \
M(Bool, optimize_trivial_insert_select, true, "Optimize trivial 'INSERT INTO table SELECT ... FROM TABLES' query", 0) \
M(Bool, allow_non_metadata_alters, true, "Allow to execute alters which affects not only tables metadata, but also data on disk", 0) \
M(Bool, enable_global_with_statement, false, "Propagate WITH statements to UNION queries and all subqueries", 0) \
M(Bool, aggregate_functions_null_for_empty, false, "Rewrite all aggregate functions in a query, adding -OrNull suffix to them", 0) \
\
/** Obsolete settings that do nothing but left for compatibility reasons. Remove each one after half a year of obsolescence. */ \
\
M(Bool, partial_merge_join, false, "Obsolete. Use join_algorithm='prefer_partial_merge' instead.", 0) \
M(UInt64, max_memory_usage_for_all_queries, 0, "Obsolete. Will be removed after 2020-10-20", 0) \
M(UInt64, multiple_joins_rewriter_version, 0, "Obsolete setting, does nothing. Will be removed after 2021-03-31", 0) \
\
M(Bool, force_optimize_skip_unused_shards_no_nested, false, "Obsolete setting, does nothing. Will be removed after 2020-12-01. Use force_optimize_skip_unused_shards_nesting instead.", 0) \
M(Bool, experimental_use_processors, true, "Obsolete setting, does nothing. Will be removed after 2020-11-29.", 0) \
M(Bool, optimize_trivial_insert_select, true, "Optimize trivial 'INSERT INTO table SELECT ... FROM TABLES' query", 0) \
M(Bool, force_optimize_skip_unused_shards_no_nested, false, "Obsolete setting, does nothing. Will be removed after 2020-12-01. Use force_optimize_skip_unused_shards_nesting instead.", 0) \
M(Bool, enable_debug_queries, false, "Enabled debug queries, but now is obsolete", 0) \
M(Bool, allow_experimental_database_atomic, true, "Obsolete setting, does nothing. Will be removed after 2021-02-12", 0) \
M(Bool, allow_non_metadata_alters, true, "Allow to execute alters which affects not only tables metadata, but also data on disk", 0) \
M(Bool, enable_global_with_statement, false, "Propagate WITH statements to UNION queries and all subqueries", 0) \
M(Bool, enable_debug_queries, false, "Enabled debug queries, but now is obsolete", 0)
// End of COMMON_SETTINGS
// Please add settings related to formats into the FORMAT_FACTORY_SETTINGS below.

29
src/Interpreters/TreeRewriter.cpp
View File

@ -31,6 +31,7 @@
#include <IO/WriteHelpers.h>
#include <Storages/IStorage.h>
#include <AggregateFunctions/AggregateFunctionFactory.h>
namespace DB
{
@ -110,6 +111,27 @@ struct CustomizeFunctionsSuffixData
char ifDistinct[] = "ifdistinct";
using CustomizeIfDistinctVisitor = InDepthNodeVisitor<OneTypeMatcher<CustomizeFunctionsSuffixData<ifDistinct>>, true>;
/// Used to rewrite all aggregate functions to add -OrNull suffix to them if setting `aggregate_functions_null_for_empty` is set.
struct CustomizeAggregateFunctionsSuffixData
{
using TypeToVisit = ASTFunction;
const String & customized_func_suffix;
void visit(ASTFunction & func, ASTPtr &)
{
const auto & instance = AggregateFunctionFactory::instance();
if (instance.isAggregateFunctionName(func.name) && !endsWith(func.name, customized_func_suffix))
{
auto properties = instance.tryGetProperties(func.name);
if (properties && !properties->returns_default_when_only_null)
func.name = func.name + customized_func_suffix;
}
}
};
using CustomizeAggregateFunctionsOrNullVisitor = InDepthNodeVisitor<OneTypeMatcher<CustomizeAggregateFunctionsSuffixData>, true>;
/// Translate qualified names such as db.table.column, table.column, table_alias.column to names' normal form.
/// Expand asterisks and qualified asterisks with column names.
/// There would be columns in normal form & column aliases after translation. Column & column alias would be normalized in QueryNormalizer.
@ -710,6 +732,13 @@ void TreeRewriter::normalize(ASTPtr & query, Aliases & aliases, const Settings &
CustomizeGlobalNotInVisitor(data_global_not_null_in).visit(query);
}
// Rewrite all aggregate functions to add -OrNull suffix to them
if (settings.aggregate_functions_null_for_empty)
{
CustomizeAggregateFunctionsOrNullVisitor::Data data_or_null{"OrNull"};
CustomizeAggregateFunctionsOrNullVisitor(data_or_null).visit(query);
}
/// Creates a dictionary `aliases`: alias -> ASTPtr
QueryAliasesVisitor(aliases).visit(query);

16
tests/queries/0_stateless/01528_setting_aggregate_functions_null_for_empty.reference Normal file
View File

@ -0,0 +1,16 @@
0
\N
0
\N
\N
\N
0
\N
45
45
10
10
45
45
10
10

36
tests/queries/0_stateless/01528_setting_aggregate_functions_null_for_empty.sql Normal file
View File

@ -0,0 +1,36 @@
DROP TABLE IF EXISTS defaults;
CREATE TABLE defaults
(
n Int8
)ENGINE = Memory();
SELECT sum(n) FROM defaults;
SELECT sumOrNull(n) FROM defaults;
SELECT count(n) FROM defaults;
SELECT countOrNull(n) FROM defaults;
SET aggregate_functions_null_for_empty=1;
SELECT sum(n) FROM defaults;
SELECT sumOrNull(n) FROM defaults;
SELECT count(n) FROM defaults;
SELECT countOrNull(n) FROM defaults;
INSERT INTO defaults SELECT * FROM numbers(10);
SET aggregate_functions_null_for_empty=0;
SELECT sum(n) FROM defaults;
SELECT sumOrNull(n) FROM defaults;
SELECT count(n) FROM defaults;
SELECT countOrNull(n) FROM defaults;
SET aggregate_functions_null_for_empty=1;
SELECT sum(n) FROM defaults;
SELECT sumOrNull(n) FROM defaults;
SELECT count(n) FROM defaults;
SELECT countOrNull(n) FROM defaults;
DROP TABLE defaults;