html: handle breakout tags in foreign content.

Also recognize that, in the latest version of the HTML5 spec, foreign content is not an insertion mode, but a separate concern. Pass tests10.dat, test 13: <!DOCTYPE html><body><table><caption><svg><g>foo</g><g>bar</g><p>baz</table><p>quux | <!DOCTYPE html> | <html> | <head> | <body> | <table> | <caption> | <svg svg> | <svg g> | "foo" | <svg g> | "bar" | <p> | "baz" | <p> | "quux" Also pass tests through test 15: <!DOCTYPE html><body><table><colgroup><svg><g>foo</g><g>bar</g><p>baz</table><p>quux R=andybalholm CC=golang-dev https://golang.org/cl/5494078

html: handle breakout tags in foreign content.
fe28d1aa · Nigel Tao · b0eb68ad · fe28d1aa · fe28d1aa
Commit fe28d1aa authored Dec 20, 2011 by Nigel Tao
Hide whitespace changes
Inline Side-by-side

Showing with 31 additions and 12 deletions

parse.go src/pkg/html/parse.go +30 -11

parse_test.go src/pkg/html/parse_test.go +1 -1

No files found.
--- a/src/pkg/html/parse.go
+++ b/src/pkg/html/parse.go
@@ -319,10 +319,7 @@ func (p *parser) resetInsertionMode() {
 		case "html":
 			p.im = beforeHeadIM
 		default:
-			if p.top().Namespace == "" {
-				continue
-			}
-			p.im = inForeignContentIM
+			continue
 		}
 		return
 	}
@@ -814,7 +811,6 @@ func inBodyIM(p *parser) bool {
 			// TODO: adjust foreign attributes.
 			p.addElement(p.tok.Data, p.tok.Attr)
 			p.top().Namespace = namespace
-			p.im = inForeignContentIM
 			return true
 		case "caption", "col", "colgroup", "frame", "head", "tbody", "td", "tfoot", "th", "thead", "tr":
 			// Ignore the token.
@@ -1590,7 +1586,7 @@ func afterAfterFramesetIM(p *parser) bool {
 }

 // Section 12.2.5.5.
-func inForeignContentIM(p *parser) bool {
+func parseForeignContent(p *parser) bool {
 	switch p.tok.Type {
 	case TextToken:
 		// TODO: HTML integration points.
@@ -1610,7 +1606,14 @@ func inForeignContentIM(p *parser) bool {
 		})
 	case StartTagToken:
 		if breakout[p.tok.Data] {
-			// TODO.
+			for i := len(p.oe) - 1; i >= 0; i-- {
+				// TODO: HTML, MathML integration points.
+				if p.oe[i].Namespace == "" {
+					p.oe = p.oe[:i+1]
+					break
+				}
+			}
+			return false
 		}
 		switch p.top().Namespace {
 		case "mathml":
@@ -1626,15 +1629,13 @@ func inForeignContentIM(p *parser) bool {
 	case EndTagToken:
 		for i := len(p.oe) - 1; i >= 0; i-- {
 			if p.oe[i].Namespace == "" {
-				inBodyIM(p)
-				break
+				return p.im(p)
 			}
 			if strings.EqualFold(p.oe[i].Data, p.tok.Data) {
 				p.oe = p.oe[:i]
 				break
 			}
 		}
-		p.resetInsertionMode()
 		return true
 	default:
 		// Ignore the token.
@@ -1642,6 +1643,20 @@ func inForeignContentIM(p *parser) bool {
 	return true
 }

+// Section 12.2.5.
+func (p *parser) inForeignContent() bool {
+	if len(p.oe) == 0 {
+		return false
+	}
+	n := p.oe[len(p.oe)-1]
+	if n.Namespace == "" {
+		return false
+	}
+	// TODO: MathML, HTML integration points.
+	// TODO: MathML's annotation-xml combining with SVG's svg.
+	return true
+}
+
 func (p *parser) parse() error {
 	// Iterate until EOF. Any other error will cause an early return.
 	consumed := true
@@ -1654,7 +1669,11 @@ func (p *parser) parse() error {
 				return err
 			}
 		}
-		consumed = p.im(p)
+		if p.inForeignContent() {
+			consumed = parseForeignContent(p)
+		} else {
+			consumed = p.im(p)
+		}
 	}
 	// Loop until the final token (the ErrorToken signifying EOF) is consumed.
 	for {

--- a/src/pkg/html/parse_test.go
+++ b/src/pkg/html/parse_test.go
@@ -173,7 +173,7 @@ func TestParser(t *testing.T) {
 		{"tests4.dat", -1},
 		{"tests5.dat", -1},
 		{"tests6.dat", 45},
-		{"tests10.dat", 13},
+		{"tests10.dat", 16},
 	}
 	for _, tf := range testFiles {
 		f, err := os.Open("testdata/webkit/" + tf.filename)