exp/template/html: make sure marshalled JSON can be parsed as JS.

This makes sure that all JS newlines are encoded in JSON.

It also moots a TODO about possibly escaping supplemental codepoints.
I served:

Content-Type: text/javascript;charset=UTF-8

var s = "%s";
document.write("<p>", s, "</p><ol>");
for (var i = 0; i < s.length; i++) {
  document.write("<li>", s.charCodeAt(i).toString(16), "</li>");
}
document.write("</l>");

where %s was replaced with bytes "\xf0\x9d\x84\x9e" to test
straight UTF-8 instead of encoding surrogates separately.

Recent Firefox, Chrome, and Safari all decoded it properly.
I have yet to try it on IE or older versions.

R=nigeltao
CC=golang-dev
https://golang.org/cl/5129042

src/pkg/exp/template/html/escape_test.go

@@ -7,6 +7,7 @@ package html
 import (
 	"bytes"
 	"fmt"
+	"json"
 	"os"
 	"strings"
 	"template"
@@ -14,11 +15,25 @@ import (
 	"testing"
 )
 
+type badMarshaler struct{}
+
+func (x *badMarshaler) MarshalJSON() ([]byte, os.Error) {
+	// Keys in valid JSON must be double quoted as must all strings.
+	return []byte("{ foo: 'not quite valid JSON' }"), nil
+}
+
+type goodMarshaler struct{}
+
+func (x *goodMarshaler) MarshalJSON() ([]byte, os.Error) {
+	return []byte(`{ "<foo>": "O'Reilly" }`), nil
+}
+
 func TestEscape(t *testing.T) {
 	var data = struct {
 		F, T    bool
 		C, G, H string
 		A, E    []string
+		B, M    json.Marshaler
 		N       int
 		Z       *int
 		W       HTML
@@ -31,6 +46,8 @@ func TestEscape(t *testing.T) {
 		A: []string{"<a>", "<b>"},
 		E: []string{},
 		N: 42,
+		B: &badMarshaler{},
+		M: &goodMarshaler{},
 		Z: nil,
 		W: HTML(`&iexcl;<b class="foo">Hello</b>, <textarea>O'World</textarea>!`),
 	}
@@ -195,6 +212,16 @@ func TestEscape(t *testing.T) {
 			"<button onclick='alert(&quot;{{.H}}&quot;)'>",
 			`<button onclick='alert(&quot;\x3cHello\x3e&quot;)'>`,
 		},
+		{
+			"badMarshaller",
+			`<button onclick='alert(1/{{.B}}in numbers)'>`,
+			`<button onclick='alert(1/ /* json: error calling MarshalJSON for type *html.badMarshaler: invalid character &#39;f&#39; looking for beginning of object key string */null in numbers)'>`,
+		},
+		{
+			"jsMarshaller",
+			`<button onclick='alert({{.M}})'>`,
+			`<button onclick='alert({&#34;&lt;foo&gt;&#34;:&#34;O&#39;Reilly&#34;})'>`,
+		},
 		{
 			"jsStrNotUnderEscaped",
 			"<button onclick='alert({{.C | urlquery}})'>",
@@ -355,8 +382,6 @@ func TestEscape(t *testing.T) {
 		},
 		{
 			"styleURLSpecialsEncoded",
-			// TODO: Find out what IE does with url(/*foo*/bar)
-			// FF, Chrome, and Safari seem to treat it as a URL.
 			`<a style="border-image: url({{"/**/'\";:// \\"}}), url(&quot;{{"/**/'\";:// \\"}}&quot;), url('{{"/**/'\";:// \\"}}'), 'http://www.example.com/?q={{"/**/'\";:// \\"}}''">`,
 			`<a style="border-image: url(/**/%27%22;://%20%5c), url(&quot;/**/%27%22;://%20%5c&quot;), url('/**/%27%22;://%20%5c'), 'http://www.example.com/?q=%2f%2a%2a%2f%27%22%3b%3a%2f%2f%20%5c''">`,
 		},

src/pkg/exp/template/html/js.go

@@ -140,19 +140,6 @@ func jsValEscaper(args ...interface{}) string {
 	// TODO: detect cycles before calling Marshal which loops infinitely on
 	// cyclic data. This may be an unnacceptable DoS risk.
 
-	// TODO: make sure that json.Marshal escapes codepoints U+2028 & U+2029
-	// so it falls within the subset of JSON which is valid JS and maybe
-	// post-process to prevent it from containing
-	// "<!--", "-->", "<![CDATA[", "]]>", or "</script"
-	// in case custom marshallers produce output containing those.
-
-	// TODO: Maybe abbreviate \u00ab to \xab to produce more compact output.
-
-	// TODO: JSON allows arbitrary unicode codepoints, but EcmaScript
-	// defines a SourceCharacter as either a UTF-16 or UCS-2 code-unit.
-	// Determine whether supplemental codepoints in UTF-8 encoded JS inside
-	// string literals are properly interpreted by major interpreters.
-
 	b, err := json.Marshal(a)
 	if err != nil {
 		// Put a space before comment so that if it is flush against
@@ -163,12 +150,50 @@ func jsValEscaper(args ...interface{}) string {
 		//          second line of error message */null
 		return fmt.Sprintf(" /* %s */null ", strings.Replace(err.String(), "*/", "* /", -1))
 	}
-	if len(b) != 0 {
-		first, _ := utf8.DecodeRune(b)
-		last, _ := utf8.DecodeLastRune(b)
-		if isJSIdentPart(first) || isJSIdentPart(last) {
-			return " " + string(b) + " "
+
+	// TODO: maybe post-process output to prevent it from containing
+	// "<!--", "-->", "<![CDATA[", "]]>", or "</script"
+	// in case custom marshallers produce output containing those.
+
+	// TODO: Maybe abbreviate \u00ab to \xab to produce more compact output.
+	if len(b) == 0 {
+		// In, `x=y/{{.}}*z` a json.Marshaler that produces "" should
+		// not cause the output `x=y/*z`.
+		return " null "
+	}
+	first, _ := utf8.DecodeRune(b)
+	last, _ := utf8.DecodeLastRune(b)
+	var buf bytes.Buffer
+	// Prevent IdentifierNames and NumericLiterals from running into
+	// keywords: in, instanceof, typeof, void
+	pad := isJSIdentPart(first) || isJSIdentPart(last)
+	if pad {
+		buf.WriteByte(' ')
+	}
+	written := 0
+	// Make sure that json.Marshal escapes codepoints U+2028 & U+2029
+	// so it falls within the subset of JSON which is valid JS.
+	for i := 0; i < len(b); {
+		rune, n := utf8.DecodeRune(b[i:])
+		repl := ""
+		if rune == 0x2028 {
+			repl = `\u2028`
+		} else if rune == 0x2029 {
+			repl = `\u2029`
 		}
+		if repl != "" {
+			buf.Write(b[written:i])
+			buf.WriteString(repl)
+			written = i + n
+		}
+		i += n
+	}
+	if buf.Len() != 0 {
+		buf.Write(b[written:])
+		if pad {
+			buf.WriteByte(' ')
+		}
+		b = buf.Bytes()
 	}
 	return string(b)
 }

src/pkg/exp/template/html/js_test.go

@@ -136,12 +136,13 @@ func TestJSValEscaper(t *testing.T) {
 		{"", `""`},
 		{"foo", `"foo"`},
 		// Newlines.
-		// {"\r\n\u2028\u2029", `"\r\n\u2028\u2029"`}, // TODO: FAILING. Maybe fix in json package.
+		{"\r\n\u2028\u2029", `"\r\n\u2028\u2029"`},
 		// "\v" == "v" on IE 6 so use "\x0b" instead.
 		{"\t\x0b", `"\u0009\u000b"`},
 		{struct{ X, Y int }{1, 2}, `{"X":1,"Y":2}`},
 		{[]interface{}{}, "[]"},
 		{[]interface{}{42, "foo", nil}, `[42,"foo",null]`},
+		{[]string{"<!--", "</script>", "-->"}, `["\u003c!--","\u003c/script\u003e","--\u003e"]`},
 		{"<!--", `"\u003c!--"`},
 		{"-->", `"--\u003e"`},
 		{"<![CDATA[", `"\u003c![CDATA["`},
@@ -331,6 +332,50 @@ func TestEscapersOnLower7AndSelectHighCodepoints(t *testing.T) {
 	}
 }
 
+func BenchmarkJSValEscaperWithNum(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		jsValEscaper(3.141592654)
+	}
+}
+
+func BenchmarkJSValEscaperWithStr(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		jsValEscaper("The <i>quick</i>,\r\n<span style='color:brown'>brown</span> fox jumps\u2028over the <canine class=\"lazy\">dog</canine>")
+	}
+}
+
+func BenchmarkJSValEscaperWithStrNoSpecials(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		jsValEscaper("The quick, brown fox jumps over the lazy dog")
+	}
+}
+
+func BenchmarkJSValEscaperWithObj(b *testing.B) {
+	o := struct {
+		S string
+		N int
+	}{
+		"The <i>quick</i>,\r\n<span style='color:brown'>brown</span> fox jumps\u2028over the <canine class=\"lazy\">dog</canine>\u2028",
+		42,
+	}
+	for i := 0; i < b.N; i++ {
+		jsValEscaper(o)
+	}
+}
+
+func BenchmarkJSValEscaperWithObjNoSpecials(b *testing.B) {
+	o := struct {
+		S string
+		N int
+	}{
+		"The quick, brown fox jumps over the lazy dog",
+		42,
+	}
+	for i := 0; i < b.N; i++ {
+		jsValEscaper(o)
+	}
+}
+
 func BenchmarkJSStrEscaperNoSpecials(b *testing.B) {
 	for i := 0; i < b.N; i++ {
 		jsStrEscaper("The quick, brown fox jumps over the lazy dog.")