Merge pull request #39688 from ClickHouse/const_columns_in_distinct

Fix DISTINCT: handle all const columns case correctly

src/Processors/Transforms/DistinctSortedTransform.cpp

@@ -8,6 +8,20 @@ namespace ErrorCodes
     extern const int SET_SIZE_LIMIT_EXCEEDED;
 }
 
+static void handleAllColumnsConst(Chunk & chunk)
+{
+    const size_t rows = chunk.getNumRows();
+    IColumn::Filter filter(rows);
+
+    Chunk res_chunk;
+    std::fill(filter.begin(), filter.end(), 0);
+    filter[0] = 1;
+    for (const auto & column : chunk.getColumns())
+        res_chunk.addColumn(column->filter(filter, -1));
+
+    chunk = std::move(res_chunk);
+}
+
 DistinctSortedTransform::DistinctSortedTransform(
     Block header_, SortDescription sort_description, const SizeLimits & set_size_limits_, UInt64 limit_hint_, const Names & columns)
     : ISimpleTransform(header_, header_, true)
@@ -23,9 +37,12 @@ DistinctSortedTransform::DistinctSortedTransform(
     for (size_t i = 0; i < num_columns; ++i)
     {
         auto pos = column_names.empty() ? i : header.getPositionByName(column_names[i]);
-        const auto & col = header.getByPosition(pos).column;
-        if (col && !isColumnConst(*col))
+        const auto & column = header.getByPosition(pos).column;
+        if (column && !isColumnConst(*column))
+        {
             column_positions.emplace_back(pos);
+            all_columns_const = false;
+        }
     }
     column_ptrs.reserve(column_positions.size());
 
@@ -52,6 +69,14 @@ void DistinctSortedTransform::transform(Chunk & chunk)
     if (unlikely(!chunk.hasRows()))
         return;
 
+    /// special case - all column constant
+    if (unlikely(all_columns_const))
+    {
+        handleAllColumnsConst(chunk);
+        stopReading();
+        return;
+    }
+
     /// get DISTINCT columns from chunk
     column_ptrs.clear();
     for (const auto pos : column_positions)

src/Processors/Transforms/DistinctSortedTransform.h

@@ -66,6 +66,7 @@ private:
 
     /// Restrictions on the maximum size of the output data.
     SizeLimits set_size_limits;
+    bool all_columns_const = true;
 };
 
 }

tests/queries/0_stateless/02317_distinct_in_order_optimization.reference

@@ -77,4 +77,30 @@
 2	2
 1	1
 0	0
+-- distinct with constants columns
+-- { echoOn }
+select distinct 1 as x, 2 as y from distinct_in_order;
+1	2
+select distinct 1 as x, 2 as y from distinct_in_order order by x;
+1	2
+select distinct 1 as x, 2 as y from distinct_in_order order by x, y;
+1	2
+select distinct a, 1 as x from distinct_in_order order by x;
+0	1
+select distinct a, 1 as x, 2 as y from distinct_in_order order by a;
+0	1	2
+select distinct a, b, 1 as x, 2 as y from distinct_in_order order by a;
+0	0	1	2
+0	1	1	2
+0	2	1	2
+0	3	1	2
+0	4	1	2
+select distinct x, y from (select 1 as x, 2 as y from distinct_in_order order by x) order by y;
+1	2
+select distinct a, b, x, y from (select a, b, 1 as x, 2 as y from distinct_in_order order by a) order by b;
+0	0	1	2
+0	1	1	2
+0	2	1	2
+0	3	1	2
+0	4	1	2
 -- check that distinct in order has the same result as ordinary distinct

tests/queries/0_stateless/02317_distinct_in_order_optimization.sql

@@ -43,6 +43,18 @@ select distinct b,c from distinct_in_order order by c;
 select '-- distinct with non-key prefix and non-sorted column, order by non-sorted desc';
 select distinct b,c from distinct_in_order order by c desc;
 
+select '-- distinct with constants columns';
+-- { echoOn }
+select distinct 1 as x, 2 as y from distinct_in_order;
+select distinct 1 as x, 2 as y from distinct_in_order order by x;
+select distinct 1 as x, 2 as y from distinct_in_order order by x, y;
+select distinct a, 1 as x from distinct_in_order order by x;
+select distinct a, 1 as x, 2 as y from distinct_in_order order by a;
+select distinct a, b, 1 as x, 2 as y from distinct_in_order order by a;
+select distinct x, y from (select 1 as x, 2 as y from distinct_in_order order by x) order by y;
+select distinct a, b, x, y from (select a, b, 1 as x, 2 as y from distinct_in_order order by a) order by b;
+-- { echoOff }
+
 drop table if exists distinct_in_order sync;
 
 select '-- check that distinct in order has the same result as ordinary distinct';