1 files changed, 60 insertions, 0 deletions
diff --git a/source/util/Bytes.mc b/source/util/Bytes.mc
index 1008fd2..a32475a 100644
--- a/source/util/Bytes.mc
+++ b/source/util/Bytes.mc
@@ -47,6 +47,66 @@ module Bytes {
         }) as ByteArray;
     }
 
+    // Decode UTF-8 bytes to a string. Empty input gives "". Malformed UTF-8
+    // or a NUL byte gives null (callers treat that as damaged data). The
+    // bytes are checked first because convertEncodedString stops the app
+    // with an uncatchable system error on invalid UTF-8.
+    function toUtf8(data as ByteArray) as String? {
+        if (data.size() == 0) {
+            return "";
+        }
+        if (!isValidUtf8(data)) {
+            return null;
+        }
+        return StringUtil.convertEncodedString(data, {
+            :fromRepresentation => StringUtil.REPRESENTATION_BYTE_ARRAY,
+            :toRepresentation => StringUtil.REPRESENTATION_STRING_PLAIN_TEXT,
+            :encoding => StringUtil.CHAR_ENCODING_UTF8
+        }) as String;
+    }
+
+    // True for well-formed UTF-8 (RFC 3629) without NUL bytes: no overlong
+    // forms, no surrogates (U+D800-DFFF), nothing above U+10FFFF.
+    function isValidUtf8(data as ByteArray) as Boolean {
+        var i = 0;
+        var n = data.size();
+        while (i < n) {
+            var b = data[i];
+            var more = 0;
+            var min = 0;
+            var cp = 0;
+            if (b == 0) {
+                return false;
+            } else if (b < 0x80) {
+                i++;
+                continue;
+            } else if (b >= 0xC2 && b <= 0xDF) {
+                more = 1; min = 0x80; cp = b & 0x1F;
+            } else if (b >= 0xE0 && b <= 0xEF) {
+                more = 2; min = 0x800; cp = b & 0x0F;
+            } else if (b >= 0xF0 && b <= 0xF4) {
+                more = 3; min = 0x10000; cp = b & 0x07;
+            } else {
+                return false;
+            }
+            if (i + more >= n) {
+                return false;
+            }
+            for (var k = 1; k <= more; k++) {
+                var c = data[i + k];
+                if ((c & 0xC0) != 0x80) {
+                    return false;
+                }
+                cp = (cp << 6) | (c & 0x3F);
+            }
+            if (cp < min || cp > 0x10FFFF || (cp >= 0xD800 && cp <= 0xDFFF)) {
+                return false;
+            }
+            i += more + 1;
+        }
+        return true;
+    }
+
     // A new array holding a followed by b; neither input is changed.
     function concat(a as ByteArray, b as ByteArray) as ByteArray {
         var out = new [0]b;