From 01f5b205175530e7189901d6a87157e2fe8a07f3 Mon Sep 17 00:00:00 2001 From: zhongyuankai <872237106@qq.com> Date: Wed, 13 Dec 2023 15:42:43 +0800 Subject: [PATCH] Support ORDER BY ALL --- .../statements/select/order-by.md | 16 +++++++++ .../statements/select/order-by.md | 16 +++++++++ src/Interpreters/TreeRewriter.cpp | 20 +++++++++++ src/Parsers/ASTSelectQuery.cpp | 5 ++- src/Parsers/ASTSelectQuery.h | 1 + src/Parsers/ParserSelectQuery.cpp | 35 +++++++++++-------- .../0_stateless/02943_order_by_all.reference | 9 +++++ .../0_stateless/02943_order_by_all.sql | 15 ++++++++ 8 files changed, 102 insertions(+), 15 deletions(-) create mode 100644 tests/queries/0_stateless/02943_order_by_all.reference create mode 100644 tests/queries/0_stateless/02943_order_by_all.sql diff --git a/docs/en/sql-reference/statements/select/order-by.md b/docs/en/sql-reference/statements/select/order-by.md index 53bdc9041a1..264d8ed323e 100644 --- a/docs/en/sql-reference/statements/select/order-by.md +++ b/docs/en/sql-reference/statements/select/order-by.md @@ -241,6 +241,22 @@ Result: └───┴─────────┘ ``` +## ORDER BY ALL + +`ORDER BY ALL` is sorts all selected columns in ascending order. + +For example: + +``` sql +SELECT a, b, c FROM t ORDER BY ALL +``` + +is the same as + +``` sql +SELECT a, b, c FROM t ORDER BY a, b, c +``` + ## Implementation Details Less RAM is used if a small enough [LIMIT](../../../sql-reference/statements/select/limit.md) is specified in addition to `ORDER BY`. Otherwise, the amount of memory spent is proportional to the volume of data for sorting. For distributed query processing, if [GROUP BY](../../../sql-reference/statements/select/group-by.md) is omitted, sorting is partially done on remote servers, and the results are merged on the requestor server. This means that for distributed sorting, the volume of data to sort can be greater than the amount of memory on a single server. diff --git a/docs/zh/sql-reference/statements/select/order-by.md b/docs/zh/sql-reference/statements/select/order-by.md index 01f702a4b1e..3286fc9f9e7 100644 --- a/docs/zh/sql-reference/statements/select/order-by.md +++ b/docs/zh/sql-reference/statements/select/order-by.md @@ -61,6 +61,22 @@ sidebar_label: ORDER BY 我们只建议使用 `COLLATE` 对于少量行的最终排序，因为排序与 `COLLATE` 比正常的按字节排序效率低。 +## ORDER BY ALL + +`ORDER BY ALL` 对所有选定的列进行升序排序。 + +示例: + +``` sql +SELECT a, b, c FROM t ORDER BY ALL +``` + +等同于： + +``` sql +SELECT a, b, c FROM t ORDER BY a, b, c +``` + ## 实现细节 {#implementation-details} 更少的RAM使用，如果一个足够小 [LIMIT](../../../sql-reference/statements/select/limit.md) 除了指定 `ORDER BY`. 否则，所花费的内存量与用于排序的数据量成正比。 对于分布式查询处理，如果 [GROUP BY](../../../sql-reference/statements/select/group-by.md) 省略排序，在远程服务器上部分完成排序，并将结果合并到请求者服务器上。 这意味着对于分布式排序，要排序的数据量可以大于单个服务器上的内存量。 diff --git a/src/Interpreters/TreeRewriter.cpp b/src/Interpreters/TreeRewriter.cpp index c63aae32090..62d3590f4e2 100644 --- a/src/Interpreters/TreeRewriter.cpp +++ b/src/Interpreters/TreeRewriter.cpp @@ -776,6 +776,22 @@ void expandGroupByAll(ASTSelectQuery * select_query) select_query->setExpression(ASTSelectQuery::Expression::GROUP_BY, group_expression_list); } +void expandOrderByAll(ASTSelectQuery * select_query) +{ + auto order_expression_list = std::make_shared(); + + for (const auto & expr : select_query->select()->children) + { + auto elem = std::make_shared(); + elem->direction = 1; + elem->nulls_direction = 1; + elem->children.push_back(expr); + order_expression_list->children.push_back(elem); + } + + select_query->setExpression(ASTSelectQuery::Expression::ORDER_BY, order_expression_list); +} + ASTs getAggregates(ASTPtr & query, const ASTSelectQuery & select_query) { /// There can not be aggregate functions inside the WHERE and PREWHERE. @@ -1292,6 +1308,10 @@ TreeRewriterResultPtr TreeRewriter::analyzeSelect( if (select_query->group_by_all) expandGroupByAll(select_query); + // expand ORDER BY ALL + if (select_query->order_by_all) + expandOrderByAll(select_query); + /// Remove unneeded columns according to 'required_result_columns'. /// Leave all selected columns in case of DISTINCT; columns that contain arrayJoin function inside. /// Must be after 'normalizeTree' (after expanding aliases, for aliases not get lost) diff --git a/src/Parsers/ASTSelectQuery.cpp b/src/Parsers/ASTSelectQuery.cpp index 7c96db006c4..d2c7de18c8a 100644 --- a/src/Parsers/ASTSelectQuery.cpp +++ b/src/Parsers/ASTSelectQuery.cpp @@ -144,7 +144,10 @@ void ASTSelectQuery::formatImpl(const FormatSettings & s, FormatState & state, F window()->as().formatImplMultiline(s, state, frame); } - if (orderBy()) + if (order_by_all) + s.ostr << (s.hilite ? hilite_keyword : "") << s.nl_or_ws << indent_str << "ORDER BY ALL" << (s.hilite ? hilite_none : ""); + + if (!order_by_all && orderBy()) { s.ostr << (s.hilite ? hilite_keyword : "") << s.nl_or_ws << indent_str << "ORDER BY" << (s.hilite ? hilite_none : ""); s.one_line diff --git a/src/Parsers/ASTSelectQuery.h b/src/Parsers/ASTSelectQuery.h index 57f45a8aacd..eb171dc00ee 100644 --- a/src/Parsers/ASTSelectQuery.h +++ b/src/Parsers/ASTSelectQuery.h @@ -87,6 +87,7 @@ public: bool group_by_with_cube = false; bool group_by_with_constant_keys = false; bool group_by_with_grouping_sets = false; + bool order_by_all = false; bool limit_with_ties = false; ASTPtr & refSelect() { return getExpression(Expression::SELECT); } diff --git a/src/Parsers/ParserSelectQuery.cpp b/src/Parsers/ParserSelectQuery.cpp index 341c1ef60b4..6c94eedc470 100644 --- a/src/Parsers/ParserSelectQuery.cpp +++ b/src/Parsers/ParserSelectQuery.cpp @@ -268,23 +268,30 @@ bool ParserSelectQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected) /// ORDER BY expr ASC|DESC COLLATE 'locale' list if (s_order_by.ignore(pos, expected)) { - if (!order_list.parse(pos, order_expression_list, expected)) - return false; - - /// if any WITH FILL parse possible INTERPOLATE list - if (std::any_of(order_expression_list->children.begin(), order_expression_list->children.end(), - [](auto & child) { return child->template as()->with_fill; })) + if (s_all.ignore(pos, expected)) { - if (s_interpolate.ignore(pos, expected)) + select_query->order_by_all = true; + } + else + { + if (!order_list.parse(pos, order_expression_list, expected)) + return false; + + /// if any WITH FILL parse possible INTERPOLATE list + if (std::any_of(order_expression_list->children.begin(), order_expression_list->children.end(), + [](auto & child) { return child->template as()->with_fill; })) { - if (open_bracket.ignore(pos, expected)) + if (s_interpolate.ignore(pos, expected)) { - if (!interpolate_list.parse(pos, interpolate_expression_list, expected)) - return false; - if (!close_bracket.ignore(pos, expected)) - return false; - } else - interpolate_expression_list = std::make_shared(); + if (open_bracket.ignore(pos, expected)) + { + if (!interpolate_list.parse(pos, interpolate_expression_list, expected)) + return false; + if (!close_bracket.ignore(pos, expected)) + return false; + } else + interpolate_expression_list = std::make_shared(); + } } } } diff --git a/tests/queries/0_stateless/02943_order_by_all.reference b/tests/queries/0_stateless/02943_order_by_all.reference new file mode 100644 index 00000000000..f9dcebc156f --- /dev/null +++ b/tests/queries/0_stateless/02943_order_by_all.reference @@ -0,0 +1,9 @@ +abc1 3 2 +abc2 3 2 +abc3 2 3 +abc4 1 4 +abc1 1 1 +abc2 1 1 +abc3 1 1 +abc4 1 1 +abc 4 diff --git a/tests/queries/0_stateless/02943_order_by_all.sql b/tests/queries/0_stateless/02943_order_by_all.sql new file mode 100644 index 00000000000..1ddd8866533 --- /dev/null +++ b/tests/queries/0_stateless/02943_order_by_all.sql @@ -0,0 +1,15 @@ +DROP TABLE IF EXISTS order_by_all; + +CREATE TABLE order_by_all +( + a String, + b int, + c int +) +engine = Memory; + +insert into order_by_all values ('abc2', 3, 2), ('abc3', 2, 3), ('abc4', 1, 4), ('abc1', 3, 2); + +select a, b, c from order_by_all order by all; +select a, count(b), count(c) from order_by_all group by all order by all; +select substring(a, 1, 3), count(b) from order_by_all group by all order by all;