diff --git a/.github/workflows/go-test.yml b/.github/workflows/go-test.yml
index ea97f05..e912800 100644
--- a/.github/workflows/go-test.yml
+++ b/.github/workflows/go-test.yml
@@ -17,3 +17,9 @@ jobs:
 
     - name: Test
       run: go test -v ./...
+
+    - name: Upload coverage reports to Codecov
+      uses: codecov/codecov-action@v5
+      with:
+        token: ${{ secrets.CODECOV_TOKEN }}
+        slug: emicklei/pgtalk
diff --git a/README.md b/README.md
index 56aa81d..2e67570 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,9 @@
 # pgtalk
 
+[![Go](https://github.com/emicklei/pgtalk/actions/workflows/go-test.yml/badge.svg)](https://github.com/emicklei/pgtalk/actions/workflows/go-test.yml)
+[![Go Report Card](https://goreportcard.com/badge/github.com/emicklei/pgtalk)](https://goreportcard.com/report/github.com/emicklei/pgtalk)
 [![GoDoc](https://pkg.go.dev/badge/github.com/emicklei/pgtalk)](https://pkg.go.dev/github.com/emicklei/pgtalk)
+[![codecov](https://codecov.io/gh/emicklei/pgtalk/branch/master/graph/badge.svg)](https://codecov.io/gh/emicklei/pgtalk)
 
 More type safe SQL query building and execution using Go code generated (pgtalk-gen) from PostgreSQL table definitions.
 After code generation, you get a Go type for each table or view with functions to create a QuerySet or MutationSet value.
diff --git a/iterator.go b/iterator.go
index 4288027..fc58e1a 100644
--- a/iterator.go
+++ b/iterator.go
@@ -6,11 +6,11 @@ import (
 )
 
 type resultIterator[T any] struct {
-	queryError error
-	commandTag pgconn.CommandTag
-	rows       pgx.Rows
-	selectors  []ColumnAccessor
-	params     []any
+	queryError       error
+	commandTag       pgconn.CommandTag
+	rows             pgx.Rows
+	orderedSelectors []ColumnAccessor
+	params           []any
 }
 
 // Close closes the rows, making the connection ready for use again. It is safe
@@ -55,15 +55,9 @@ func (i *resultIterator[T]) HasNext() bool {
 
 func (i *resultIterator[T]) Next() (*T, error) {
 	entity := new(T)
-	list := i.rows.FieldDescriptions()
-	// order of list is not the same as selectors?
 	toScan := []any{}
-	for _, each := range list {
-		for _, other := range i.selectors {
-			if other.Column().columnName == each.Name {
-				toScan = append(toScan, other.FieldValueToScan(entity))
-			}
-		}
+	for _, each := range i.orderedSelectors {
+		toScan = append(toScan, each.FieldValueToScan(entity))
 	}
 	if err := i.rows.Scan(toScan...); err != nil {
 		return nil, err
diff --git a/iterator_test.go b/iterator_test.go
new file mode 100644
index 0000000..e6c4ed4
--- /dev/null
+++ b/iterator_test.go
@@ -0,0 +1,134 @@
+package pgtalk
+
+import (
+	"errors"
+	"testing"
+
+	"github.com/jackc/pgx/v5"
+	"github.com/jackc/pgx/v5/pgconn"
+)
+
+func TestResultIterator_Next(t *testing.T) {
+	t.Run("happy flow", func(t *testing.T) {
+		rows := &mockRows{
+			nextResult:  true,
+			scanError:   nil,
+			closeCalled: false,
+		}
+		selectors := []ColumnAccessor{
+			mockColumnAccessor{
+				fieldValueToScan: func(entity any) any {
+					return &entity.(*testEntity).ID
+				},
+			},
+		}
+		i := &resultIterator[testEntity]{
+			rows:             rows,
+			orderedSelectors: selectors,
+		}
+		if !i.HasNext() {
+			t.Fatal("expected next")
+		}
+		entity, err := i.Next()
+		if err != nil {
+			t.Fatal(err)
+		}
+		if entity.ID != "test-id" {
+			t.Errorf("expected test-id, got %s", entity.ID)
+		}
+		if i.HasNext() {
+			t.Fatal("unexpected next")
+		}
+		if !rows.closeCalled {
+			t.Error("expected close to be called")
+		}
+	})
+}
+
+func TestResultIterator_Err(t *testing.T) {
+	t.Run("query error", func(t *testing.T) {
+		i := &resultIterator[testEntity]{
+			queryError: errors.New("query error"),
+		}
+		if err := i.Err(); err == nil || err.Error() != "query error" {
+			t.Errorf("expected query error, got %v", err)
+		}
+	})
+	t.Run("rows error", func(t *testing.T) {
+		rows := &mockRows{
+			err: errors.New("rows error"),
+		}
+		i := &resultIterator[testEntity]{
+			rows: rows,
+		}
+		if err := i.Err(); err == nil || err.Error() != "rows error" {
+			t.Errorf("expected rows error, got %v", err)
+		}
+	})
+}
+
+func TestResultIterator_GetParams(t *testing.T) {
+	params := []any{"param1", 2}
+	i := &resultIterator[testEntity]{
+		params: params,
+	}
+	p := i.GetParams()
+	if len(p) != 2 {
+		t.Errorf("expected 2 params, got %d", len(p))
+	}
+	if p[1] != "param1" {
+		t.Errorf("expected param1, got %v", p[1])
+	}
+	if p[2] != 2 {
+		t.Errorf("expected 2, got %v", p[2])
+	}
+}
+
+// mockColumnAccessor is a mock for the ColumnAccessor interface
+type mockColumnAccessor struct {
+	fieldValueToScan func(entity any) any
+}
+
+func (m mockColumnAccessor) SQLOn(w WriteContext)                {}
+func (m mockColumnAccessor) Name() string                        { return "" }
+func (m mockColumnAccessor) ValueToInsert() any                  { return nil }
+func (m mockColumnAccessor) Column() ColumnInfo                  { return ColumnInfo{} }
+func (m mockColumnAccessor) FieldValueToScan(entity any) any     { return m.fieldValueToScan(entity) }
+func (m mockColumnAccessor) AppendScannable(list []any) []any    { return list }
+func (m mockColumnAccessor) Get(values map[string]any) any       { return nil }
+func (m mockColumnAccessor) SetSource(parameterIndex int) string { return "" }
+
+// mockRows is a mock for the pgx.Rows interface
+type mockRows struct {
+	nextResult  bool
+	scanError   error
+	closeCalled bool
+	err         error
+}
+
+func (m *mockRows) Close()                                       { m.closeCalled = true }
+func (m *mockRows) Err() error                                   { return m.err }
+func (m *mockRows) CommandTag() pgconn.CommandTag                { return pgconn.CommandTag{} }
+func (m *mockRows) FieldDescriptions() []pgconn.FieldDescription { return nil }
+func (m *mockRows) Next() bool                                   { return m.nextResult }
+func (m *mockRows) Scan(dest ...any) error {
+	if m.scanError != nil {
+		return m.scanError
+	}
+	// simulate scanning a value
+	if len(dest) > 0 {
+		if id, ok := dest[0].(*string); ok {
+			*id = "test-id"
+		}
+	}
+	m.nextResult = false // only one row
+	return nil
+}
+func (m *mockRows) RawValues() [][]byte    { return nil }
+func (m *mockRows) Conn() *pgx.Conn        { return nil }
+func (m *mockRows) Values() ([]any, error) { return nil, nil }
+
+// testEntity is a simple struct for testing
+type testEntity struct {
+	ID string
+}
diff --git a/mutationset.go b/mutationset.go
index 967374b..4aba9e1 100644
--- a/mutationset.go
+++ b/mutationset.go
@@ -103,10 +103,31 @@ func (m MutationSet[T]) Exec(ctx context.Context, conn querier, parameters ...*Q
 		return &resultIterator[T]{queryError: err, commandTag: ct, params: params}
 	}
 	rows, err := conn.Query(ctx, query, params...)
-	if err == nil && !m.canProduceResults() {
+	if err != nil {
+		return &resultIterator[T]{queryError: err}
+	}
+	if !m.canProduceResults() {
 		rows.Close()
 	}
-	return &resultIterator[T]{queryError: err, rows: rows, selectors: m.returning, params: params}
+	// order the selectors once
+	fds := rows.FieldDescriptions()
+	ordered := make([]ColumnAccessor, len(fds))
+
+	// create a map for faster lookup
+	selectorMap := make(map[string]ColumnAccessor, len(m.returning))
+	for _, sel := range m.returning {
+		selectorMap[sel.Column().columnName] = sel
+	}
+
+	for i, fd := range fds {
+		sel, ok := selectorMap[fd.Name]
+		if !ok {
+			// this should not happen
+			return &resultIterator[T]{queryError: fmt.Errorf("selector not found for column %s", fd.Name)}
+		}
+		ordered[i] = sel
+	}
+	return &resultIterator[T]{rows: rows, orderedSelectors: ordered, params: params}
 }
 
 // valuesToInsert returns the parameters values for the mutation query.
diff --git a/queryset.go b/queryset.go
index 85b89f3..2f4e7b4 100644
--- a/queryset.go
+++ b/queryset.go
@@ -131,11 +131,32 @@ func (q QuerySet[T]) Exists() unaryExpression {
 func (d QuerySet[T]) Iterate(ctx context.Context, conn querier, parameters ...*QueryParameter) (*resultIterator[T], error) {
 	params := argumentValues(parameters)
 	rows, err := conn.Query(ctx, SQL(d), params...)
+	if err != nil {
+		return nil, err
+	}
+	// order the selectors once
+	fds := rows.FieldDescriptions()
+	ordered := make([]ColumnAccessor, len(fds))
+
+	// create a map for faster lookup
+	selectorMap := make(map[string]ColumnAccessor, len(d.selectors))
+	for _, sel := range d.selectors {
+		selectorMap[sel.Column().columnName] = sel
+	}
+
+	for i, fd := range fds {
+		sel, ok := selectorMap[fd.Name]
+		if !ok {
+			return nil, fmt.Errorf("no selector found for field '%s'", fd.Name)
+		}
+		ordered[i] = sel
+	}
+
 	return &resultIterator[T]{
-		queryError: err,
-		rows:       rows,
-		selectors:  d.selectors,
-		params:     params,
+		queryError:       err,
+		rows:             rows,
+		orderedSelectors: ordered,
+		params:           params,
 	}, err
 }
 
diff --git a/sql_writing.go b/sql_writing.go
index 7aba5e9..0d7dd98 100644
--- a/sql_writing.go
+++ b/sql_writing.go
@@ -49,6 +49,32 @@ func (w wc) TableAlias(tableName, defaultAlias string) string {
 	return defaultAlias
 }
 
+// onelineWriter is a writer that replaces newlines and tabs with spaces,
+// and collapses multiple spaces into a single space.
+type onelineWriter struct {
+	b                *strings.Builder
+	lastCharWasSpace bool
+}
+
+func newOnelineWriter(b *strings.Builder) *onelineWriter {
+	return &onelineWriter{b: b, lastCharWasSpace: true}
+}
+
+func (w *onelineWriter) Write(p []byte) (n int, err error) {
+	for _, b := range p {
+		if b == '\n' || b == '\t' || b == ' ' {
+			if !w.lastCharWasSpace {
+				w.b.WriteByte(' ')
+				w.lastCharWasSpace = true
+			}
+		} else {
+			w.b.WriteByte(b)
+			w.lastCharWasSpace = false
+		}
+	}
+	return len(p), nil
+}
+
 // IndentedSQL returns source with tabs and lines trying to have a formatted view.
 func IndentedSQL(some SQLWriter) string {
 	buf := new(bytes.Buffer)
@@ -58,6 +84,8 @@ func IndentedSQL(some SQLWriter) string {
 
 // SQL returns source as a oneliner without tabs or line ends.
 func SQL(some SQLWriter) string {
-	src := IndentedSQL(some)
-	return strings.ReplaceAll(strings.ReplaceAll(strings.ReplaceAll(src, "\t", " "), "\n", " "), "  ", " ")
+	var b strings.Builder
+	w := newOnelineWriter(&b)
+	some.SQLOn(NewWriteContext(w))
+	return strings.TrimRight(b.String(), " ")
 }