chore(engine): Filter by unknown column (#18131)

The new engine fails to execute a query when an `arrow.Record` (batch) does not contain the column on which a label filter is used, e.g. `| foo="bar"` where `foo` is neither a label nor a structured metadata.

This change also fixed an incorrect conversion of the label filter operator `=`, which was converted to `MATCH_STR`, but requires to be `EQ`.

The new engine reflects current behaviour of a label filter `| foo=""` where `foo` is undefined.

Signed-off-by: Christian Haudum <[email protected]>

Author: chaudum

pkg/engine/executor/dataobjscan.go

@@ -406,10 +406,6 @@ func schemaFromColumns(columns []physical.ColumnExpression) (*arrow.Schema, erro
 			return nil, fmt.Errorf("invalid column expression type %T", column)
 		}
 
-		md := arrow.MetadataFrom(map[string]string{
-			types.MetadataKeyColumnType: columnExpr.Ref.Type.String(),
-		})
-
 		switch columnExpr.Ref.Type {
 		case types.ColumnTypeLabel:
 			// TODO(rfratto): Switch to dictionary encoding for labels.
@@ -424,9 +420,10 @@ func schemaFromColumns(columns []physical.ColumnExpression) (*arrow.Schema, erro
 			//
 			// We skipped dictionary encoding for now to get the initial prototype
 			// working.
+			ty, md := arrowTypeFromColumnRef(columnExpr.Ref)
 			addField(arrow.Field{
 				Name:     columnExpr.Ref.Column,
-				Type:     arrow.BinaryTypes.String,
+				Type:     ty,
 				Nullable: true,
 				Metadata: md,
 			})
@@ -436,15 +433,16 @@ func schemaFromColumns(columns []physical.ColumnExpression) (*arrow.Schema, erro
 			// has unconstrained cardinality. Using dictionary encoding would require
 			// tracking every encoded value in the record, which is likely to be too
 			// expensive.
+			ty, md := arrowTypeFromColumnRef(columnExpr.Ref)
 			addField(arrow.Field{
 				Name:     columnExpr.Ref.Column,
-				Type:     arrow.BinaryTypes.String,
+				Type:     ty,
 				Nullable: true,
 				Metadata: md,
 			})
 
 		case types.ColumnTypeBuiltin:
-			ty, md := builtinColumnType(columnExpr.Ref)
+			ty, md := arrowTypeFromColumnRef(columnExpr.Ref)
 			addField(arrow.Field{
 				Name:     columnExpr.Ref.Column,
 				Type:     ty,
@@ -472,36 +470,36 @@ func schemaFromColumns(columns []physical.ColumnExpression) (*arrow.Schema, erro
 				Name:     columnExpr.Ref.Column,
 				Type:     arrow.BinaryTypes.String,
 				Nullable: true,
-				Metadata: arrow.MetadataFrom(map[string]string{types.MetadataKeyColumnType: types.ColumnTypeLabel.String()}),
+				Metadata: datatype.ColumnMetadata(types.ColumnTypeLabel, datatype.String),
 			})
 			addField(arrow.Field{
 				Name:     columnExpr.Ref.Column,
 				Type:     arrow.BinaryTypes.String,
 				Nullable: true,
-				Metadata: arrow.MetadataFrom(map[string]string{types.MetadataKeyColumnType: types.ColumnTypeMetadata.String()}),
+				Metadata: datatype.ColumnMetadata(types.ColumnTypeMetadata, datatype.String),
 			})
 
-		case types.ColumnTypeParsed:
+		case types.ColumnTypeParsed, types.ColumnTypeGenerated:
 			return nil, fmt.Errorf("parsed column type not supported: %s", columnExpr.Ref.Type)
 		}
 	}
 
 	return arrow.NewSchema(fields, nil), nil
 }
 
-func builtinColumnType(ref types.ColumnRef) (arrow.DataType, arrow.Metadata) {
-	if ref.Type != types.ColumnTypeBuiltin {
-		panic("builtinColumnType called with a non-builtin column")
+func arrowTypeFromColumnRef(ref types.ColumnRef) (arrow.DataType, arrow.Metadata) {
+	if ref.Type == types.ColumnTypeBuiltin {
+		switch ref.Column {
+		case types.ColumnNameBuiltinTimestamp:
+			return arrow.FixedWidthTypes.Timestamp_ns, datatype.ColumnMetadataBuiltinTimestamp
+		case types.ColumnNameBuiltinMessage:
+			return arrow.BinaryTypes.String, datatype.ColumnMetadataBuiltinMessage
+		default:
+			panic(fmt.Sprintf("unsupported builtin column type %s", ref))
+		}
 	}
 
-	switch ref.Column {
-	case types.ColumnNameBuiltinTimestamp:
-		return arrow.FixedWidthTypes.Timestamp_ns, datatype.ColumnMetadataBuiltinTimestamp
-	case types.ColumnNameBuiltinMessage:
-		return arrow.BinaryTypes.String, datatype.ColumnMetadataBuiltinMessage
-	default:
-		panic(fmt.Sprintf("unsupported builtin column type %s", ref))
-	}
+	return arrow.BinaryTypes.String, datatype.ColumnMetadata(ref.Type, datatype.String)
 }
 
 // appendToBuilder appends a the provided field from record into the given

pkg/engine/executor/dataobjscan_test.go

@@ -20,16 +20,10 @@ import (
 )
 
 var (
-	labelMD    = buildMetadata(types.ColumnTypeLabel)
-	metadataMD = buildMetadata(types.ColumnTypeMetadata)
+	labelMD    = datatype.ColumnMetadata(types.ColumnTypeLabel, datatype.String)
+	metadataMD = datatype.ColumnMetadata(types.ColumnTypeMetadata, datatype.String)
 )
 
-func buildMetadata(ty types.ColumnType) arrow.Metadata {
-	return arrow.MetadataFrom(map[string]string{
-		types.MetadataKeyColumnType: ty.String(),
-	})
-}
-
 func Test_dataobjScan(t *testing.T) {
 	obj := buildDataobj(t, []logproto.Stream{
 		{

pkg/engine/executor/expressions.go

@@ -8,7 +8,6 @@ import (
 	"github.com/apache/arrow-go/v18/arrow/memory"
 
 	"github.com/grafana/loki/v3/pkg/engine/internal/datatype"
-	"github.com/grafana/loki/v3/pkg/engine/internal/errors"
 	"github.com/grafana/loki/v3/pkg/engine/internal/types"
 	"github.com/grafana/loki/v3/pkg/engine/planner/physical"
 )
@@ -46,7 +45,13 @@ func (e expressionEvaluator) eval(expr physical.Expression, input arrow.Record)
 				}, nil
 			}
 		}
-		return nil, fmt.Errorf("unknown column %s: %w", expr.Ref.String(), errors.ErrKey)
+		// A non-existent column is represented as a string scalar with zero-value.
+		// This reflects current behaviour, where a label filter `| foo=""` would match all if `foo` is not defined.
+		return &Scalar{
+			value: datatype.NewStringLiteral(""),
+			rows:  input.NumRows(),
+			ct:    types.ColumnTypeGenerated,
+		}, nil
 
 	case *physical.UnaryExpr:
 		lhr, err := e.eval(expr.Left, input)

pkg/engine/executor/expressions_test.go

@@ -10,7 +10,6 @@ import (
 	"github.com/stretchr/testify/require"
 
 	"github.com/grafana/loki/v3/pkg/engine/internal/datatype"
-	"github.com/grafana/loki/v3/pkg/engine/internal/errors"
 	"github.com/grafana/loki/v3/pkg/engine/internal/types"
 	"github.com/grafana/loki/v3/pkg/engine/planner/physical"
 )
@@ -102,7 +101,7 @@ func TestEvaluateLiteralExpression(t *testing.T) {
 func TestEvaluateColumnExpression(t *testing.T) {
 	e := expressionEvaluator{}
 
-	t.Run("invalid", func(t *testing.T) {
+	t.Run("unknown column", func(t *testing.T) {
 		colExpr := &physical.ColumnExpr{
 			Ref: types.ColumnRef{
 				Column: "does_not_exist",
@@ -112,8 +111,12 @@ func TestEvaluateColumnExpression(t *testing.T) {
 
 		n := len(words)
 		rec := batch(n, time.Now())
-		_, err := e.eval(colExpr, rec)
-		require.ErrorContains(t, err, errors.ErrKey.Error())
+		colVec, err := e.eval(colExpr, rec)
+		require.NoError(t, err)
+
+		_, ok := colVec.(*Scalar)
+		require.True(t, ok, "expected column vector to be a *Scalar, got %T", colVec)
+		require.Equal(t, arrow.STRING, colVec.Type().ArrowType().ID())
 	})
 
 	t.Run("string(message)", func(t *testing.T) {

pkg/engine/executor/filter.go

@@ -86,6 +86,7 @@ func filterBatch(batch arrow.Record, include func(int) bool) arrow.Record {
 	additions := make([]func(int), len(fields))
 
 	for i, field := range fields {
+
 		switch field.Type.ID() {
 		case arrow.BOOL:
 			builder := array.NewBooleanBuilder(mem)
@@ -127,6 +128,14 @@ func filterBatch(batch arrow.Record, include func(int) bool) arrow.Record {
 				builder.Append(src.Value(offset))
 			}
 
+		case arrow.TIMESTAMP:
+			builder := array.NewTimestampBuilder(mem, &arrow.TimestampType{Unit: arrow.Nanosecond, TimeZone: "UTC"})
+			builders[i] = builder
+			additions[i] = func(offset int) {
+				src := batch.Column(i).(*array.Timestamp)
+				builder.Append(src.Value(offset))
+			}
+
 		default:
 			panic(fmt.Sprintf("unimplemented type in filterBatch: %s", field.Type.Name()))
 		}

pkg/engine/executor/sortmerge_test.go

@@ -24,7 +24,7 @@ func TestSortMerge(t *testing.T) {
 		merge := &physical.SortMerge{
 			Column: &physical.ColumnExpr{
 				Ref: types.ColumnRef{
-					Column: "invalid",
+					Column: "not_a_timestamp_column",
 					Type:   types.ColumnTypeBuiltin,
 				},
 			},
@@ -40,7 +40,7 @@ func TestSortMerge(t *testing.T) {
 		require.NoError(t, err)
 
 		err = pipeline.Read()
-		require.ErrorContains(t, err, "key error")
+		require.ErrorContains(t, err, "column is not a timestamp column")
 	})
 
 	t.Run("ascending timestamp", func(t *testing.T) {

pkg/engine/planner/logical/planner.go

@@ -182,9 +182,9 @@ func lineColumnRef() *ColumnRef {
 func convertLabelMatchType(op labels.MatchType) types.BinaryOp {
 	switch op {
 	case labels.MatchEqual:
-		return types.BinaryOpMatchSubstr
+		return types.BinaryOpEq
 	case labels.MatchNotEqual:
-		return types.BinaryOpNotMatchSubstr
+		return types.BinaryOpNeq
 	case labels.MatchRegexp:
 		return types.BinaryOpMatchRe
 	case labels.MatchNotRegexp:

pkg/engine/planner/logical/planner_test.go

@@ -98,8 +98,8 @@ func TestConvertAST_Success(t *testing.T) {
 %7 = SELECT %5 [predicate=%6]
 %8 = LT builtin.timestamp 1970-01-01T02:00:00Z
 %9 = SELECT %7 [predicate=%8]
-%10 = MATCH_STR ambiguous.foo "bar"
-%11 = MATCH_STR ambiguous.bar "baz"
+%10 = EQ ambiguous.foo "bar"
+%11 = EQ ambiguous.bar "baz"
 %12 = OR %10 %11
 %13 = SELECT %9 [predicate=%12]
 %14 = MATCH_STR builtin.message "metric.go"