feat: Add EmitSelfClosingTag field to emit EndEvent for self-closing tags

Author: orisano

compat.go

@@ -39,9 +39,6 @@ import (
 // StartElement converts a byte slice to an xml.StartElement.
 func StartElement(b []byte) (xml.StartElement, error) {
 	name, b := Name(b)
-	if len(b) > 0 && b[len(b)-1] == '/' {
-		b = b[:len(b)-1]
-	}
 	e := xml.StartElement{
 		Name: xmlName(name),
 	}
@@ -64,7 +61,7 @@ func StartElement(b []byte) (xml.StartElement, error) {
 func EndElement(b []byte) xml.EndElement {
 	name, _ := Name(b)
 	return xml.EndElement{
-		Name: xmlName(name[1:]),
+		Name: xmlName(name),
 	}
 }
 

example_test.go

@@ -148,3 +148,39 @@ func ExampleToken() {
 	// EndElement element
 	// EndElement root
 }
+
+func ExampleReader_EmitSelfClosingTag() {
+	xmlData := `<root><element>Value</element><selfclosing/></root>`
+	reader := strings.NewReader(xmlData)
+
+	r := gosax.NewReader(reader)
+	r.EmitSelfClosingTag = true
+	for {
+		e, err := r.Event()
+		if err != nil {
+			log.Fatal(err)
+		}
+		if e.Type() == gosax.EventEOF {
+			break
+		}
+		switch e.Type() {
+		case gosax.EventStart:
+			name, _ := gosax.Name(e.Bytes)
+			fmt.Println("EventStart", string(name))
+		case gosax.EventEnd:
+			name, _ := gosax.Name(e.Bytes)
+			fmt.Println("EventEnd", string(name))
+		case gosax.EventText:
+			fmt.Println("EventText", string(e.Bytes))
+		default:
+		}
+	}
+	// Output:
+	// EventStart root
+	// EventStart element
+	// EventText Value
+	// EventEnd element
+	// EventStart selfclosing
+	// EventEnd selfclosing
+	// EventEnd root
+}

gosax.go

@@ -62,6 +62,9 @@ func (e Event) Type() uint8 {
 type Reader struct {
 	reader byteReader
 	state  func(*Reader) (Event, error)
+
+	EmitSelfClosingTag bool
+	selfClosingEnd     int
 }
 
 func NewReader(r io.Reader) *Reader {
@@ -282,7 +285,12 @@ func (r *Reader) stateInsideMarkup() (Event, error) {
 					}
 					if p >= 0 {
 						if ch == '>' {
-							r.reader.offset += offset + p + 1
+							if r.EmitSelfClosingTag && w[offset+p-1] == '/' {
+								r.selfClosingEnd = offset + p
+								r.state = (*Reader).stateSelfClosingTag
+							} else {
+								r.reader.offset += offset + p + 1
+							}
 							return Event{
 								Bytes: w[:offset+p+1],
 								value: EventStart,
@@ -312,6 +320,16 @@ func (r *Reader) stateInsideMarkup() (Event, error) {
 	}
 }
 
+func (r *Reader) stateSelfClosingTag() (Event, error) {
+	r.state = (*Reader).stateInsideText
+	w := r.reader.window()
+	r.reader.offset += r.selfClosingEnd + 1
+	return Event{
+		Bytes: w[:r.selfClosingEnd+1],
+		value: EventEnd,
+	}, nil
+}
+
 func (r *Reader) stateDone() (Event, error) {
 	return Event{
 		value: EventEOF,
@@ -346,9 +364,15 @@ func Name(b []byte) ([]byte, []byte) {
 	if len(b) > 1 && b[0] == '<' {
 		b = b[1:]
 	}
+	if len(b) > 1 && b[0] == '/' {
+		b = b[1:]
+	}
 	if len(b) > 1 && b[len(b)-1] == '>' {
 		b = b[:len(b)-1]
 	}
+	if len(b) > 1 && b[len(b)-1] == '/' {
+		b = b[:len(b)-1]
+	}
 	for i, c := range b {
 		if whitespace[c] {
 			return b[:i], b[i+1:]