Merge pull request ClickHouse#86959 from ClickHouse/backport/25.8/85985

Backport ClickHouse#85985 to 25.8: Fix alter update of a column with a subcolumn used in other column materialized expression

Author: clickhouse-gh[bot]

src/Interpreters/MutationsInterpreter.cpp

@@ -6,6 +6,7 @@
 #include <Interpreters/MutationsInterpreter.h>
 #include <Interpreters/TreeRewriter.h>
 #include <Interpreters/MutationsNonDeterministicHelpers.h>
+#include <Interpreters/replaceSubcolumnsToGetSubcolumnFunctionInQuery.h>
 #include <Storages/MergeTree/MergeTreeData.h>
 #include <Storages/MergeTree/StorageFromMergeTreeDataPart.h>
 #include <Storages/StorageMergeTree.h>
@@ -528,7 +529,7 @@ static void validateUpdateColumns(
                 {
                     throw Exception(ErrorCodes::CANNOT_UPDATE_COLUMN,
                                     "Updated column {} affects MATERIALIZED column {}, which is a key column. "
-                                    "Cannot UPDATE it.", backQuote(column_name), backQuote(materialized));
+                                    "Cannot UPDATE it", backQuote(column_name), backQuote(materialized));
                 }
             }
         }
@@ -676,6 +677,8 @@ void MutationsInterpreter::prepare(bool dry_run)
             if (column.default_desc.kind == ColumnDefaultKind::Materialized && available_columns_set.contains(column.name))
             {
                 auto query = column.default_desc.expression->clone();
+                /// Replace all subcolumns to the getSubcolumn() to get only top level columns as required source columns.
+                replaceSubcolumnsToGetSubcolumnFunctionInQuery(query, all_columns);
                 auto syntax_result = TreeRewriter(context).analyze(query, all_columns);
                 for (const auto & dependency : syntax_result->requiredSourceColumns())
                     if (updated_columns.contains(dependency))
@@ -855,10 +858,15 @@ void MutationsInterpreter::prepare(bool dry_run)
                     {
                         auto type_literal = std::make_shared<ASTLiteral>(column.type->getName());
 
-                        auto materialized_column = makeASTFunction("_CAST",
+                        ASTPtr materialized_column = makeASTFunction("_CAST",
                             column.default_desc.expression->clone(),
                             type_literal);
 
+                        /// We need to replace all subcolumns used in materialized expression to getSubcolumn() function,
+                        /// because otherwise subcolumns are extracted before the source column is updated and we get
+                        /// old subcolumns values.
+                        replaceSubcolumnsToGetSubcolumnFunctionInQuery(materialized_column, all_columns);
+
                         stages.back().column_to_updated.emplace(
                             column.name,
                             materialized_column);

src/Interpreters/replaceSubcolumnsToGetSubcolumnFunctionInQuery.cpp

@@ -0,0 +1,40 @@
+#include <Interpreters/replaceSubcolumnsToGetSubcolumnFunctionInQuery.h>
+#include <Parsers/ASTFunction.h>
+#include <Parsers/ASTIdentifier.h>
+#include <Parsers/ASTLiteral.h>
+#include <DataTypes/IDataType.h>
+#include <DataTypes/NestedUtils.h>
+
+namespace DB
+{
+
+void replaceSubcolumnsToGetSubcolumnFunctionInQuery(ASTPtr & ast, const NamesAndTypesList & columns)
+{
+    if (auto * identifier = ast->as<ASTIdentifier>())
+    {
+        if (columns.contains(identifier->getColumnName()))
+            return;
+
+        auto [column_name, subcolumn_name] = Nested::splitName(identifier->getColumnName());
+        auto column = columns.tryGetByName(column_name);
+        if (!column || !column->type->hasSubcolumn(subcolumn_name))
+            return;
+
+        ast = makeASTFunction("getSubcolumn", std::make_shared<ASTIdentifier>(column_name), std::make_shared<ASTLiteral>(subcolumn_name));
+    }
+    else if (auto * node = ast->as<ASTFunction>())
+    {
+        if (node->arguments)
+        {
+            for (auto & child : node->arguments->children)
+                replaceSubcolumnsToGetSubcolumnFunctionInQuery(child, columns);
+        }
+    }
+    else
+    {
+        for (auto & child : ast->children)
+            replaceSubcolumnsToGetSubcolumnFunctionInQuery(child, columns);
+    }
+}
+
+}

src/Interpreters/replaceSubcolumnsToGetSubcolumnFunctionInQuery.h

@@ -0,0 +1,14 @@
+#pragma once
+
+#include <Core/NamesAndTypes.h>
+#include <Parsers/IAST_fwd.h>
+#include <base/types.h>
+
+namespace DB
+{
+
+/// Replace subcolumns to getSubcolumn() function.
+void replaceSubcolumnsToGetSubcolumnFunctionInQuery(ASTPtr & ast, const NamesAndTypesList & columns);
+
+}
+

tests/queries/0_stateless/03596_alter_update_column_with_subcolumn_used_in_materialized_expression.reference

@@ -0,0 +1,14 @@
+(1)	1
+(2)	2
+(3)	3
+(1)	43
+(2)	44
+(3)	45
+(1)	1
+{"a":1}	1
+{"a":2}	2
+{"a":3}	3
+{"a":1}	43
+{"a":2}	44
+{"a":3}	45
+{"a":1}	1

tests/queries/0_stateless/03596_alter_update_column_with_subcolumn_used_in_materialized_expression.sql

@@ -0,0 +1,55 @@
+drop table if exists test;
+create table test (t Tuple(a UInt32), a UInt32 materialized t.a) engine=MergeTree() order by tuple();
+insert into test select tuple(1);
+select t, a from test;
+alter table test update t = tuple(2) where 1 settings mutations_sync=1;
+select t, a from test;
+alter table test update t = tuple(3) where 1 settings mutations_sync=1;
+select t, a from test;
+drop table test;
+
+drop table if exists test;
+create table test (t Tuple(a UInt32), a UInt32 materialized t.a + 42) engine=MergeTree() order by tuple();
+insert into test select tuple(1);
+select t, a from test;
+alter table test update t = tuple(2) where 1 settings mutations_sync=1;
+select t, a from test;
+alter table test update t = tuple(3) where 1 settings mutations_sync=1;
+select t, a from test;
+drop table test;
+
+
+create table test (t Tuple(a UInt32), a UInt32 materialized t.a) engine=MergeTree() order by a;
+insert into test select tuple(1);
+select t, a from test;
+alter table test update t = tuple(2) where 1 settings mutations_sync=1; -- {serverError CANNOT_UPDATE_COLUMN}
+drop table test;
+
+drop table if exists test;
+create table test (json JSON, a UInt32 materialized json.a::UInt32) engine=MergeTree() order by tuple();
+insert into test select '{"a" : 1}';
+select json, a from test;
+alter table test update json = '{"a" : 2}' where 1 settings mutations_sync=1;
+select json, a from test;
+alter table test update json = '{"a" : 3}' where 1 settings mutations_sync=1;
+select json, a from test;
+drop table test;
+
+drop table if exists test;
+create table test (json JSON, a UInt32 materialized json.a::UInt32 + 42) engine=MergeTree() order by tuple();
+insert into test select '{"a" : 1}';
+select json, a from test;
+alter table test update json = '{"a" : 2}' where 1 settings mutations_sync=1;
+select json, a from test;
+alter table test update json = '{"a" : 3}' where 1 settings mutations_sync=1;
+select json, a from test;
+drop table test;
+
+
+
+create table test (json JSON, a UInt32 materialized json.a::UInt32) engine=MergeTree() order by a;
+insert into test select '{"a" : 1}';
+select json, a from test;
+alter table test update json = '{"a" : 2}' where 1 settings mutations_sync=1; -- {serverError CANNOT_UPDATE_COLUMN}
+drop table test;
+