reencode: Rethink the string backslash encoder

1 files changed, 53 insertions, 28 deletions

diff --git a/misc.go b/misc.go
index a567cc7..4d8b136 100644
--- a/misc.go
+++ b/misc.go
@@ -68,34 +68,42 @@ func writeRune(w io.Writer, c rune) (int, error) {
 
 // JSON string encoding ////////////////////////////////////////////////////////
 
-func UnicodeEscapeJSSafe(c rune, _ bool) bool {
+type BackslashEscapeMode uint8
+
+const (
+	BackslashEscapeNone = BackslashEscapeMode(iota)
+	BackslashEscapeShort
+	BackslashEscapeUnicode
+)
+
+func EscapeJSSafe(c rune, _ BackslashEscapeMode) BackslashEscapeMode {
 	// JSON is notionally a JS subset, but that's not actually
 	// true.
 	//
 	// http://timelessrepo.com/json-isnt-a-javascript-subset
 	switch c {
 	case '\u2028', '\u2029':
-		return true
+		return BackslashEscapeUnicode
 	default:
-		return false
+		return BackslashEscapeNone
 	}
 }
 
-func UnicodeEscapeHTMLSafe(c rune, wasEscaped bool) bool {
+func EscapeHTMLSafe(c rune, wasEscaped BackslashEscapeMode) BackslashEscapeMode {
 	switch c {
 	case '&', '<', '>':
-		return true
+		return BackslashEscapeUnicode
 	default:
-		return UnicodeEscapeJSSafe(c, wasEscaped)
+		return EscapeJSSafe(c, wasEscaped)
 	}
 }
 
-func UnicodeEscapeDefault(c rune, wasEscaped bool) bool {
+func EscapeDefault(c rune, wasEscaped BackslashEscapeMode) BackslashEscapeMode {
 	switch c {
 	case '\b', '\f', utf8.RuneError:
-		return true
+		return BackslashEscapeUnicode
 	default:
-		return UnicodeEscapeHTMLSafe(c, wasEscaped)
+		return EscapeHTMLSafe(c, wasEscaped)
 	}
 }
 
@@ -114,31 +122,48 @@ func writeStringShortEscape(w io.Writer, c byte) (int, error) {
 	buf := [2]byte{'\\', c}
 	return w.Write(buf[:])
 }
-func writeStringChar(w io.Writer, c rune, wasEscaped bool, escaper func(rune, bool) bool) (int, error) {
+func writeStringChar(w io.Writer, c rune, wasEscaped BackslashEscapeMode, escaper func(rune, BackslashEscapeMode) BackslashEscapeMode) (int, error) {
 	if escaper == nil {
-		escaper = UnicodeEscapeDefault
+		escaper = EscapeDefault
 	}
-	switch {
-	case c <= 0xFFFF && escaper(c, wasEscaped):
-		return writeStringUnicodeEscape(w, c)
-	case c == '"' || c == '\\':
-		return writeStringShortEscape(w, byte(c))
-	case c < 0x0020:
+	switch escaper(c, wasEscaped) {
+	case BackslashEscapeNone:
+		switch {
+		case c < 0x0020:
+			switch c {
+			case '\b':
+				return writeStringShortEscape(w, 'b')
+			case '\f':
+				return writeStringShortEscape(w, 'f')
+			case '\n':
+				return writeStringShortEscape(w, 'n')
+			case '\r':
+				return writeStringShortEscape(w, 'r')
+			case '\t':
+				return writeStringShortEscape(w, 't')
+			default:
+				return writeStringUnicodeEscape(w, c)
+			}
+		case c == '"' || c == '\\':
+			return writeStringShortEscape(w, byte(c))
+		default:
+			return writeRune(w, c)
+		}
+	case BackslashEscapeShort:
 		switch c {
-		case '\b':
-			return writeStringShortEscape(w, 'b')
-		case '\f':
-			return writeStringShortEscape(w, 'f')
-		case '\n':
-			return writeStringShortEscape(w, 'n')
-		case '\r':
-			return writeStringShortEscape(w, 'r')
-		case '\t':
-			return writeStringShortEscape(w, 't')
+		case '"', '\\', '/', '\b', '\f', '\n', '\r', '\t':
+			return writeStringShortEscape(w, byte(c))
+		default:
+			return writeRune(w, c)
+		}
+	case BackslashEscapeUnicode:
+		switch {
+		case c > 0xFFFF:
+			return writeRune(w, c)
 		default:
 			return writeStringUnicodeEscape(w, c)
 		}
 	default:
-		return writeRune(w, c)
+		panic("escaper returned an invalid escape mode")
 	}
 }