parser: Handle self-closing elements and void elements

- Add `selfClosing` field to `nodeElement` struct to track self-closing tags
- Update `match` function to accept multiple token types
- Check for self-closing tags in `parseElement` and set `selfClosing` field
- Return early for void elements in `parseElement`
- Refactor `parseChildren` to handle both start and self-closing tags
- Add `isVoidElement` function to check if a tag is a void element
- Update tests to cover self-closing and void elements

🌟 These changes improve the HTML parser to properly handle self-closing
elements and void elements according to the HTML spec.

Fixes #126

ast.go

@@ -177,6 +177,7 @@ type nodeElement struct {
 	startTagNodes []node
 	children      []node
 	pos           span
+	selfClosing   bool
 }
 
 func (e nodeElement) Pos() span { return e.pos }

parser.go

@@ -420,23 +420,35 @@ func newTag(tagname []byte, attrs []*attr) tag {
 	return tag{name: string(tagname), attrs: attrs}
 }
 
-func (p *htmlParser) match(typ html.TokenType) bool {
-	return p.toktyp == typ
+func (p *htmlParser) match(types ...html.TokenType) bool {
+	for _, typ := range types {
+		if p.toktyp == typ {
+			return true
+		}
+	}
+	return false
 }
 
 func (p *htmlParser) parseElement() node {
 	var result *nodeElement
 
-	// FIXME(paulsmith): handle self-closing elements
-	if !p.match(html.StartTagToken) {
-		p.errorf("expected an HTML element start tag, got %s", p.toktyp)
+	if !p.match(html.StartTagToken, html.SelfClosingTagToken) {
+		p.errorf("expected an HTML element start or self-closing tag, got %s", p.toktyp)
 	}
 
 	result = new(nodeElement)
 	result.tag = newTag(p.tagname, p.attrs)
 	result.pos.start = p.parser.offset - len(p.raw)
 	result.pos.end = p.parser.offset
+	if p.match(html.SelfClosingTagToken) {
+		result.selfClosing = true
+	}
 	result.startTagNodes = p.parseStartTag()
+
+	if isVoidElement(result.tag.name) {
+		return result
+	}
+
 	p.advance()
 
 	result.children = p.parseChildren()
@@ -462,31 +474,28 @@ func (p *htmlParser) parseChildren() []node {
 	var elemStack []*nodeElement
 loop:
 	for {
-		switch p.toktyp {
+		switch tt := p.toktyp; tt {
 		case html.ErrorToken:
 			if p.err == io.EOF {
 				break loop
 			} else {
 				p.errorf("HTML tokenizer: %w", p.err)
 			}
-		case html.SelfClosingTagToken:
-			elem := new(nodeElement)
-			elem.tag = newTag(p.tagname, p.attrs)
-			elem.pos.start = p.parser.offset - len(p.raw)
-			elem.pos.end = p.parser.offset
-			elem.startTagNodes = p.parseStartTag()
-			p.advance()
-			result = append(result, elem)
-		case html.StartTagToken:
+		case html.StartTagToken, html.SelfClosingTagToken:
 			elem := new(nodeElement)
 			elem.tag = newTag(p.tagname, p.attrs)
 			elem.pos.start = p.parser.offset - len(p.raw)
 			elem.pos.end = p.parser.offset
 			elem.startTagNodes = p.parseStartTag()
+			if tt == html.SelfClosingTagToken {
+				elem.selfClosing = true
+			}
 			p.advance()
-			elem.children = p.parseChildren()
+			if !isVoidElement(elem.tag.name) {
+				elem.children = p.parseChildren()
+				elemStack = append(elemStack, elem)
+			}
 			result = append(result, elem)
-			elemStack = append(elemStack, elem)
 		case html.EndTagToken:
 			if len(elemStack) == 0 {
 				return result
@@ -534,6 +543,25 @@ loop:
 	return result
 }
 
+func isVoidElement(tagname string) bool {
+	// Per spec: https://html.spec.whatwg.org/multipage/syntax.html#void-elements
+	voidElements := map[string]bool{"area": true,
+		"base":   true,
+		"br":     true,
+		"col":    true,
+		"embed":  true,
+		"hr":     true,
+		"img":    true,
+		"input":  true,
+		"link":   true,
+		"meta":   true,
+		"source": true,
+		"track":  true,
+		"wbr":    true,
+	}
+	return voidElements[tagname]
+}
+
 type Optional[T any] struct {
 	value *T
 }

parser_test.go

@@ -405,6 +405,27 @@ func TestParse(t *testing.T) {
 				},
 			},
 		},
+		{
+			`^if true { <br/> }`,
+			&syntaxTree{
+				nodes: []node{
+					&nodeIf{
+						cond: &nodeGoStrExpr{expr: "true", pos: span{start: 4, end: 8}},
+						then: &nodeBlock{
+							nodes: []node{
+								&nodeLiteral{str: " ", pos: span{10, 11}},
+								&nodeElement{
+									tag:           tag{name: "br"},
+									startTagNodes: []node{&nodeLiteral{str: "<br/>", pos: span{start: 11, end: 16}}},
+									pos:           span{11, 16},
+									selfClosing:   true,
+								},
+							},
+						},
+					},
+				},
+			},
+		},
 	}
 	opts := cmp.AllowUnexported(unexported...)
 	for _, test := range tests {
@@ -450,7 +471,7 @@ func TestParseSyntaxErrors(t *testing.T) {
 		{
 			`^if true {
 	<illegal />
-}`, 2, 13,
+}`, 3, 2,
 		},
 		// FIXME(paulsmith): add more syntax errors
 	}