From db2956eb737e137f076f0c6467979621d4edc3a8 Mon Sep 17 00:00:00 2001
From: Alexey Milovidov <milovidov@yandex-team.ru>
Date: Thu, 12 Mar 2015 06:31:15 +0300
Subject: [PATCH] dbms: removing duplicate elements from ORDER BY [#METR-2944].

---
 .../DB/Interpreters/ExpressionAnalyzer.h      |  3 ++
 dbms/src/Interpreters/ExpressionAnalyzer.cpp  | 35 +++++++++++++++++++
 .../Interpreters/InterpreterSelectQuery.cpp   |  8 ++---
 .../00136_duplicate_order_by_elems.reference  |  0
 .../00136_duplicate_order_by_elems.sql        |  1 +
 5 files changed, 42 insertions(+), 5 deletions(-)
 create mode 100644 dbms/tests/queries/0_stateless/00136_duplicate_order_by_elems.reference
 create mode 100644 dbms/tests/queries/0_stateless/00136_duplicate_order_by_elems.sql

diff --git a/dbms/include/DB/Interpreters/ExpressionAnalyzer.h b/dbms/include/DB/Interpreters/ExpressionAnalyzer.h
index c4b01ffe0ba..24ba3c863e7 100644
--- a/dbms/include/DB/Interpreters/ExpressionAnalyzer.h
+++ b/dbms/include/DB/Interpreters/ExpressionAnalyzer.h
@@ -218,6 +218,9 @@ private:
 	///	Eliminates injective function calls and constant expressions from group by statement
 	void optimizeGroupBy();
 
+	/// Удалить из ORDER BY повторяющиеся элементы.
+	void optimizeOrderBy();
+
 	/// Превратить перечисление значений или подзапрос в ASTSet. node - функция in или notIn.
 	void makeSet(ASTFunction * node, const Block & sample_block);
 
diff --git a/dbms/src/Interpreters/ExpressionAnalyzer.cpp b/dbms/src/Interpreters/ExpressionAnalyzer.cpp
index be35443f5cc..ed20ee30054 100644
--- a/dbms/src/Interpreters/ExpressionAnalyzer.cpp
+++ b/dbms/src/Interpreters/ExpressionAnalyzer.cpp
@@ -85,6 +85,9 @@ void ExpressionAnalyzer::init()
 	/// GROUP BY injective function elimination.
 	optimizeGroupBy();
 
+	/// Удалить из ORDER BY повторяющиеся элементы.
+	optimizeOrderBy();
+
 	/// array_join_alias_to_name, array_join_result_to_source.
 	getArrayJoinedColumns();
 
@@ -531,6 +534,38 @@ void ExpressionAnalyzer::optimizeGroupBy()
 }
 
 
+void ExpressionAnalyzer::optimizeOrderBy()
+{
+	if (!(select_query && select_query->order_expression_list))
+		return;
+
+	/// Уникализируем условия сортировки.
+	using NameAndLocale = std::pair<std::string, std::string>;
+	std::set<NameAndLocale> elems_set;
+
+	ASTs & elems = select_query->order_expression_list->children;
+	ASTs unique_elems;
+	unique_elems.reserve(elems.size());
+
+	for (const auto & elem : elems)
+	{
+		String name = elem->children.front()->getColumnName();
+		const ASTOrderByElement & order_by_elem = typeid_cast<const ASTOrderByElement &>(*elem);
+
+		if (elems_set.emplace(
+			std::piecewise_construct,
+			std::forward_as_tuple(name),
+			std::forward_as_tuple(order_by_elem.collator ? order_by_elem.collator->getLocale() : std::string())).second)
+		{
+			unique_elems.emplace_back(elem);
+		}
+	}
+
+	if (unique_elems.size() < elems.size())
+		elems = unique_elems;
+}
+
+
 void ExpressionAnalyzer::makeSetsForIndex()
 {
 	if (storage && ast && storage->supportsIndexForIn())
diff --git a/dbms/src/Interpreters/InterpreterSelectQuery.cpp b/dbms/src/Interpreters/InterpreterSelectQuery.cpp
index 2c676e2b3c1..5952defae10 100644
--- a/dbms/src/Interpreters/InterpreterSelectQuery.cpp
+++ b/dbms/src/Interpreters/InterpreterSelectQuery.cpp
@@ -817,12 +817,10 @@ static SortDescription getSortDescription(ASTSelectQuery & query)
 {
 	SortDescription order_descr;
 	order_descr.reserve(query.order_expression_list->children.size());
-	for (ASTs::iterator it = query.order_expression_list->children.begin();
-		it != query.order_expression_list->children.end();
-		++it)
+	for (const auto & elem : query.order_expression_list->children)
 	{
-		String name = (*it)->children.front()->getColumnName();
-		const ASTOrderByElement & order_by_elem = typeid_cast<const ASTOrderByElement &>(**it);
+		String name = elem->children.front()->getColumnName();
+		const ASTOrderByElement & order_by_elem = typeid_cast<const ASTOrderByElement &>(*elem);
 
 		order_descr.emplace_back(name, order_by_elem.direction, order_by_elem.collator);
 	}
diff --git a/dbms/tests/queries/0_stateless/00136_duplicate_order_by_elems.reference b/dbms/tests/queries/0_stateless/00136_duplicate_order_by_elems.reference
new file mode 100644
index 00000000000..e69de29bb2d
diff --git a/dbms/tests/queries/0_stateless/00136_duplicate_order_by_elems.sql b/dbms/tests/queries/0_stateless/00136_duplicate_order_by_elems.sql
new file mode 100644
index 00000000000..66a0d1a11d5
--- /dev/null
+++ b/dbms/tests/queries/0_stateless/00136_duplicate_order_by_elems.sql
@@ -0,0 +1 @@
+SELECT n FROM (SELECT number AS n FROM system.numbers LIMIT 1000000) ORDER BY n, n, n, n, n, n, n, n, n, n LIMIT 1000000, 1;