Merge pull request #36055 from helifu/master2

Don't materialize external tables if it's explain statement

src/Interpreters/ExpressionAnalyzer.cpp

@@ -210,6 +210,7 @@ ExpressionAnalyzer::ExpressionAnalyzer(
     ContextPtr context_,
     size_t subquery_depth_,
     bool do_global,
+    bool is_explain,
     SubqueriesForSets subqueries_for_sets_,
     PreparedSets prepared_sets_)
     : WithContext(context_)
@@ -223,7 +224,7 @@ ExpressionAnalyzer::ExpressionAnalyzer(
 
     /// external_tables, subqueries_for_sets for global subqueries.
     /// Replaces global subqueries with the generated names of temporary tables that will be sent to remote servers.
-    initGlobalSubqueriesAndExternalTables(do_global);
+    initGlobalSubqueriesAndExternalTables(do_global, is_explain);
 
     auto temp_actions = std::make_shared<ActionsDAG>(sourceColumns());
     columns_after_array_join = getColumnsAfterArrayJoin(temp_actions, sourceColumns());
@@ -391,12 +392,12 @@ void ExpressionAnalyzer::analyzeAggregation(ActionsDAGPtr & temp_actions)
 }
 
 
-void ExpressionAnalyzer::initGlobalSubqueriesAndExternalTables(bool do_global)
+void ExpressionAnalyzer::initGlobalSubqueriesAndExternalTables(bool do_global, bool is_explain)
 {
     if (do_global)
     {
         GlobalSubqueriesVisitor::Data subqueries_data(
-            getContext(), subquery_depth, isRemoteStorage(), external_tables, subqueries_for_sets, has_global_subqueries);
+            getContext(), subquery_depth, isRemoteStorage(), is_explain, external_tables, subqueries_for_sets, has_global_subqueries);
         GlobalSubqueriesVisitor(subqueries_data).visit(query);
     }
 }

src/Interpreters/ExpressionAnalyzer.h

@@ -98,7 +98,7 @@ public:
     /// Ctor for non-select queries. Generally its usage is:
     /// auto actions = ExpressionAnalyzer(query, syntax, context).getActions();
     ExpressionAnalyzer(const ASTPtr & query_, const TreeRewriterResultPtr & syntax_analyzer_result_, ContextPtr context_)
-        : ExpressionAnalyzer(query_, syntax_analyzer_result_, context_, 0, false, {}, {})
+        : ExpressionAnalyzer(query_, syntax_analyzer_result_, context_, 0, false, false, {}, {})
     {
     }
 
@@ -154,6 +154,7 @@ protected:
         ContextPtr context_,
         size_t subquery_depth_,
         bool do_global_,
+        bool is_explain_,
         SubqueriesForSets subqueries_for_sets_,
         PreparedSets prepared_sets_);
 
@@ -168,7 +169,7 @@ protected:
     const NamesAndTypesList & sourceColumns() const { return syntax->required_source_columns; }
     const std::vector<const ASTFunction *> & aggregates() const { return syntax->aggregates; }
     /// Find global subqueries in the GLOBAL IN/JOIN sections. Fills in external_tables.
-    void initGlobalSubqueriesAndExternalTables(bool do_global);
+    void initGlobalSubqueriesAndExternalTables(bool do_global, bool is_explain);
 
     ArrayJoinActionPtr addMultipleArrayJoinAction(ActionsDAGPtr & actions, bool is_left) const;
 
@@ -305,6 +306,7 @@ public:
             context_,
             options_.subquery_depth,
             do_global_,
+            options_.is_explain,
             std::move(subqueries_for_sets_),
             std::move(prepared_sets_))
         , metadata_snapshot(metadata_snapshot_)

src/Interpreters/GlobalSubqueriesVisitor.h

@@ -39,6 +39,7 @@ public:
     {
         size_t subquery_depth;
         bool is_remote;
+        bool is_explain;
         TemporaryTablesMapping & external_tables;
         SubqueriesForSets & subqueries_for_sets;
         bool & has_global_subqueries;
@@ -47,12 +48,14 @@ public:
             ContextPtr context_,
             size_t subquery_depth_,
             bool is_remote_,
+            bool is_explain_,
             TemporaryTablesMapping & tables,
             SubqueriesForSets & subqueries_for_sets_,
             bool & has_global_subqueries_)
             : WithContext(context_)
             , subquery_depth(subquery_depth_)
             , is_remote(is_remote_)
+            , is_explain(is_explain_)
             , external_tables(tables)
             , subqueries_for_sets(subqueries_for_sets_)
             , has_global_subqueries(has_global_subqueries_)
@@ -160,7 +163,11 @@ public:
             /// We need to materialize external tables immediately because reading from distributed
             /// tables might generate local plans which can refer to external tables during index
             /// analysis. It's too late to populate the external table via CreatingSetsTransform.
-            if (getContext()->getSettingsRef().use_index_for_in_with_subqueries)
+            if (is_explain)
+            {
+                /// Do not materialize external tables if it's explain statement.
+            }
+            else if (getContext()->getSettingsRef().use_index_for_in_with_subqueries)
             {
                 auto external_table = external_storage_holder->getTable();
                 auto table_out = external_table->write({}, external_table->getInMemoryMetadataPtr(), getContext());

src/Interpreters/InterpreterExplainQuery.cpp

@@ -269,6 +269,9 @@ QueryPipeline InterpreterExplainQuery::executeImpl()
     bool single_line = false;
     bool insert_buf = true;
 
+    SelectQueryOptions options;
+    options.setExplain();
+
     switch (ast.getKind())
     {
         case ASTExplainQuery::ParsedAST:
@@ -299,7 +302,7 @@ QueryPipeline InterpreterExplainQuery::executeImpl()
             auto settings = checkAndGetSettings<QueryPlanSettings>(ast.getSettings());
             QueryPlan plan;
 
-            InterpreterSelectWithUnionQuery interpreter(ast.getExplainedQuery(), getContext(), SelectQueryOptions());
+            InterpreterSelectWithUnionQuery interpreter(ast.getExplainedQuery(), getContext(), options);
             interpreter.buildQueryPlan(plan);
 
             if (settings.optimize)
@@ -334,7 +337,7 @@ QueryPipeline InterpreterExplainQuery::executeImpl()
                 auto settings = checkAndGetSettings<QueryPipelineSettings>(ast.getSettings());
                 QueryPlan plan;
 
-                InterpreterSelectWithUnionQuery interpreter(ast.getExplainedQuery(), getContext(), SelectQueryOptions());
+                InterpreterSelectWithUnionQuery interpreter(ast.getExplainedQuery(), getContext(), options);
                 interpreter.buildQueryPlan(plan);
                 auto pipeline = plan.buildQueryPipeline(
                     QueryPlanOptimizationSettings::fromContext(getContext()),

src/Interpreters/SelectQueryOptions.h

@@ -49,6 +49,7 @@ struct SelectQueryOptions
     bool is_subquery = false; // non-subquery can also have subquery_depth > 0, e.g. insert select
     bool with_all_cols = false; /// asterisk include materialized and aliased columns
     bool settings_limit_offset_done = false;
+    bool is_explain = false; /// The value is true if it's explain statement.
 
     /// These two fields are used to evaluate shardNum() and shardCount() function when
     /// prefer_localhost_replica == 1 and local instance is selected. They are needed because local
@@ -150,6 +151,12 @@ struct SelectQueryOptions
         shard_count = shard_count_;
         return *this;
     }
+
+    SelectQueryOptions & setExplain(bool value = true)
+    {
+        is_explain = value;
+        return *this;
+    }
 };
 
 }