update bluemonday to v1.0.5 to fix https://github.com/microcosm-cc/bluemonday/issues/111

backend/vendor/github.com/microcosm-cc/bluemonday/sanitize.go

@@ -229,7 +229,7 @@ func (p *Policy) sanitize(r io.Reader) *bytes.Buffer {
 
 		case html.StartTagToken:
 
-			mostRecentlyStartedToken = strings.ToLower(token.Data)
+			mostRecentlyStartedToken = normaliseElementName(token.Data)
 
 			aps, ok := p.elsAndAttrs[token.Data]
 			if !ok {
@@ -272,7 +272,7 @@ func (p *Policy) sanitize(r io.Reader) *bytes.Buffer {
 
 		case html.EndTagToken:
 
-			if mostRecentlyStartedToken == strings.ToLower(token.Data) {
+			if mostRecentlyStartedToken == normaliseElementName(token.Data) {
 				mostRecentlyStartedToken = ""
 			}
 
@@ -350,11 +350,11 @@ func (p *Policy) sanitize(r io.Reader) *bytes.Buffer {
 
 			if !skipElementContent {
 				switch mostRecentlyStartedToken {
-				case "script":
+				case `script`:
 					// not encouraged, but if a policy allows JavaScript we
 					// should not HTML escape it as that would break the output
 					buff.WriteString(token.Data)
-				case "style":
+				case `style`:
 					// not encouraged, but if a policy allows CSS styles we
 					// should not HTML escape it as that would break the output
 					buff.WriteString(token.Data)
@@ -887,3 +887,23 @@ func (p *Policy) matchRegex(elementName string) (map[string]attrPolicy, bool) {
 	}
 	return aps, matched
 }
+
+
+// normaliseElementName takes a HTML element like <script> which is user input
+// and returns a lower case version of it that is immune to UTF-8 to ASCII
+// conversion tricks (like the use of upper case cyrillic i scrİpt which a
+// strings.ToLower would convert to script). Instead this func will preserve
+// all non-ASCII as their escaped equivalent, i.e. \u0130 which reveals the
+// characters when lower cased
+func normaliseElementName(str string) string {
+	// that useful QuoteToASCII put quote marks at the start and end
+	// so those are trimmed off
+	return strings.TrimSuffix(
+		strings.TrimPrefix(
+			strings.ToLower(
+				strconv.QuoteToASCII(str),
+			),
+			`"`),
+		`"`,
+	)
+}