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

dbms: allowed to hashing tuples [#METR-18564].

Browse Source
This commit is contained in:
Alexey Milovidov 2015-10-29 03:12:04 +03:00
parent c9a8f19def
commit 9f16f6fb6e
3 changed files with 49 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
dbms/include/DB/Functions/FunctionsHashing.h
View File

@ -19,6 +19,7 @@
#include <DB/Columns/ColumnConst.h>
#include <DB/Columns/ColumnFixedString.h>
#include <DB/Columns/ColumnArray.h>
#include <DB/Columns/ColumnTuple.h>
#include <DB/Common/HashTable/Hash.h>
#include <DB/Functions/IFunction.h>
@ -399,6 +400,9 @@ UInt64 toInteger<Float64>(Float64 x)
}
/** Используются хэш-функции под названием CityHash, FarmHash, MetroHash.
* В связи с этим, этот шаблон назван со словами NeighbourhoodHash.
*/
template <typename Impl>
class FunctionNeighbourhoodHash64 : public IFunction
{
@ -565,6 +569,29 @@ private:
ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT);
}
void executeForArgument(const IDataType * type, const IColumn * column, ColumnUInt64::Container_t & vec_to, bool & is_first)
{
/// Раскрытие кортежей.
if (const ColumnTuple * tuple = typeid_cast<const ColumnTuple *>(column))
{
const Block & tuple_data = tuple->getData();
for (size_t i = 0, size = tuple_data.columns(); i < size; ++i)
{
const ColumnWithTypeAndName & col = tuple_data.unsafeGetByPosition(i);
executeForArgument(col.type.get(), col.column.get(), vec_to, is_first);
}
}
else
{
if (is_first)
executeAny<true>(type, column, vec_to);
else
executeAny<false>(type, column, vec_to);
}
is_first = false;
}
public:
/// Получить имя функции.
String getName() const override
@ -589,20 +616,17 @@ public:
if (arguments.empty())
{
/// Случайное число из /dev/urandom используется как хеш пустого кортежа.
/// Случайное число из /dev/urandom используется как хеш пустого количества аргументов.
vec_to.assign(rows, 0xe28dbde7fe22e41c);
}
/// Функция поддерживает произвольное количество аргументов всевозможных типов.
bool is_first_argument = true;
for (size_t i = 0; i < arguments.size(); ++i)
{
const ColumnWithTypeAndName & column = block.getByPosition(arguments[i]);
const IDataType * from_type = &*column.type;
const IColumn * icolumn = &*column.column;
if (i == 0)
executeAny<true>(from_type, icolumn, vec_to);
else
executeAny<false>(from_type, icolumn, vec_to);
const ColumnWithTypeAndName & col = block.getByPosition(arguments[i]);
executeForArgument(col.type.get(), col.column.get(), vec_to, is_first_argument);
}
}
};

8
dbms/tests/queries/0_stateless/00259_hashing_tuples.reference Normal file
View File

@ -0,0 +1,8 @@
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012
3151614220437955012 3151614220437955012 3151614220437955012 3151614220437955012

8
dbms/tests/queries/0_stateless/00259_hashing_tuples.sql Normal file
View File

@ -0,0 +1,8 @@
SELECT cityHash64(1, 2, '') AS x1, cityHash64((1, 2), '') AS x2, cityHash64(1, (2, '')) AS x3, cityHash64((1, 2, '')) AS x4;
SELECT cityHash64(materialize(1), 2, '') AS x1, cityHash64((materialize(1), 2), '') AS x2, cityHash64(materialize(1), (2, '')) AS x3, cityHash64((materialize(1), 2, '')) AS x4;
SELECT cityHash64(1, materialize(2), '') AS x1, cityHash64((1, materialize(2)), '') AS x2, cityHash64(1, (materialize(2), '')) AS x3, cityHash64((1, materialize(2), '')) AS x4;
SELECT cityHash64(1, 2, materialize('')) AS x1, cityHash64((1, 2), materialize('')) AS x2, cityHash64(1, (2, materialize(''))) AS x3, cityHash64((1, 2, materialize(''))) AS x4;
SELECT cityHash64(materialize(1), materialize(2), '') AS x1, cityHash64((materialize(1), materialize(2)), '') AS x2, cityHash64(materialize(1), (materialize(2), '')) AS x3, cityHash64((materialize(1), materialize(2), '')) AS x4;
SELECT cityHash64(1, materialize(2), materialize('')) AS x1, cityHash64((1, materialize(2)), materialize('')) AS x2, cityHash64(1, (materialize(2), materialize(''))) AS x3, cityHash64((1, materialize(2), materialize(''))) AS x4;
SELECT cityHash64(materialize(1), 2, materialize('')) AS x1, cityHash64((materialize(1), 2), materialize('')) AS x2, cityHash64(materialize(1), (2, materialize(''))) AS x3, cityHash64((materialize(1), 2, materialize(''))) AS x4;
SELECT cityHash64(materialize(1), materialize(2), materialize('')) AS x1, cityHash64((materialize(1), materialize(2)), materialize('')) AS x2, cityHash64(materialize(1), (materialize(2), materialize(''))) AS x3, cityHash64((materialize(1), materialize(2), materialize(''))) AS x4;