Add support for auto-closing tags

etree.go

@@ -52,6 +52,12 @@ type ReadSettings struct {
 
 	// Entity to be passed to standard xml.Decoder. Default: nil.
 	Entity map[string]string
+
+	// When Permissive is true, AutoClose indicates a set of elements to
+	// consider closed immediately after they are opened, regardless of
+	// whether an end element is present. Commonly set to xml.HTMLAutoClose.
+	// Default: nil.
+	AutoClose []string
 }
 
 // newReadSettings creates a default ReadSettings record.
@@ -796,6 +802,27 @@ func (e *Element) RemoveChildAt(index int) Token {
 	return t
 }
 
+// autoClose analyzes the stack's top element and the current token to decide
+// whether the top element should be closed.
+func (e *Element) autoClose(stack *stack, t xml.Token, tags []string) {
+	if stack.empty() {
+		return
+	}
+
+	top := stack.peek().(*Element)
+
+	for _, tag := range tags {
+		if strings.EqualFold(tag, top.FullTag()) {
+			if e, ok := t.(xml.EndElement); !ok ||
+				!strings.EqualFold(e.Name.Space, top.Space) ||
+				!strings.EqualFold(e.Name.Local, top.Tag) {
+				stack.pop()
+			}
+			break
+		}
+	}
+}
+
 // ReadFrom reads XML from the reader 'ri' and stores the result as a new
 // child of this element.
 func (e *Element) readFrom(ri io.Reader, settings ReadSettings) (n int64, err error) {
@@ -822,6 +849,10 @@ func (e *Element) readFrom(ri io.Reader, settings ReadSettings) (n int64, err er
 
 		t, err := dec.RawToken()
 
+		if settings.Permissive && settings.AutoClose != nil {
+			e.autoClose(&stack, t, settings.AutoClose)
+		}
+
 		switch {
 		case err == io.EOF:
 			if len(stack.data) != 1 {

etree_test.go

@@ -348,6 +348,45 @@ func TestDocumentReadHTMLEntities(t *testing.T) {
 	}
 }
 
+func TestDocumentReadHTMLAutoClose(t *testing.T) {
+	cases := []struct {
+		name  string
+		input string
+		want  string
+	}{
+		{"empty", ``, ``},
+		{"oneSelfClosing", `<br>`, `<br/>`},
+		{"twoSelfClosingAndText", `<br>some text<br>`, `<br/>some text<br/>`},
+		{
+			name: "largerExample",
+			input: `<img src="cover.jpg">
+<hr>
+Author: Charles Dickens<br>
+Book: Great Expectations<br>`,
+			want: `<img src="cover.jpg"/>
+<hr/>
+Author: Charles Dickens<br/>
+Book: Great Expectations<br/>`},
+	}
+
+	for _, c := range cases {
+		t.Run(c.name, func(t *testing.T) {
+			doc := NewDocument()
+			doc.ReadSettings.Permissive = true
+			doc.ReadSettings.AutoClose = xml.HTMLAutoClose
+			err := doc.ReadFromString(c.input)
+			if err != nil {
+				t.Fatal("etree: ReadFromString() error = ", err)
+			}
+			s, err := doc.WriteToString()
+			if err != nil {
+				t.Fatal("etree: WriteToString() error = ", err)
+			}
+			checkStrEq(t, s, c.want)
+		})
+	}
+}
+
 func TestEscapeCodes(t *testing.T) {
 	cases := []struct {
 		input         string