Fix UTF8 related issues in virtio filer

MercuryWorkshop · Aug 28, 2024 · eb6b7c6 · eb6b7c6
1 parent 0858e59
commit eb6b7c6
Show file tree

Hide file tree

Showing 2 changed files with 27 additions and 23 deletions.
diff --git a/lib/marshall.js b/lib/marshall.js
@@ -48,14 +48,20 @@ marshall.Marshall = function(typelist, input, struct, offset) {
                 struct[offset++] = 0; // set the length later
                 struct[offset++] = 0;
                 size += 2;
-                for(var j of item) {
-                    var utf8 = UnicodeToUTF8Stream(j.charCodeAt(0));
-                    utf8.forEach( function(c) {
-                        struct[offset++] = c;
-                        size += 1;
-                        length++;
-                    });
-                }
+
+                var stringBytes = texten.encode(item);
+                size += stringBytes.byteLength;
+                length += stringBytes.byteLength;
+                struct.set(stringBytes, offset);
+                offset += stringBytes.byteLength;
+                // for(var j of item) {
+                //     var utf8 = UnicodeToUTF8Stream(j.charCodeAt(0));
+                //     utf8.forEach( function(c) {
+                //         struct[offset++] = c;
+                //         size += 1;
+                //         length++;
+                //     });
+                // }
                 struct[lengthoffset+0] = length & 0xFF;
                 struct[lengthoffset+1] = (length >> 8) & 0xFF;
                 break;
@@ -104,14 +110,16 @@ marshall.Unmarshall = function(typelist, struct, state) {
             case "s":
                 var len = struct[offset++];
                 len += struct[offset++] << 8;
-                var str = "";
-                var utf8converter = new UTF8StreamToUnicode();
-                for(var j=0; j < len; j++) {
-                    var c = utf8converter.Put(struct[offset++]);
-                    if(c === -1) continue;
-                    str += String.fromCharCode(c);
-                }
-                output.push(str);
+                // var str = "";
+                // var utf8converter = new UTF8StreamToUnicode();
+
+                // for(var j=0; j < len; j++) {
+                //     var c = utf8converter.Put(struct[offset++]);
+                //     if(c === -1) continue;ofc 
+                //     str += String.fromCharCode(c);
+                // }
+                var stringBytes = struct.slice(offset, offset + len);
+                output.push(textde.decode(stringBytes));
                 break;
             case "Q":
                 state.offset = offset;

diff --git a/lib/utf8.js b/lib/utf8.js
@@ -51,13 +51,9 @@ function UnicodeToUTF8Stream(key)
         if(key < 0x80)  return [key];
         if(key < 0x800) return [0xC0|((key>>6)&0x1F), 0x80|(key&0x3F)];
 }
-
+const texten = new TextEncoder();
+const textde = new TextDecoder();
 UTF8.UTF8Length = function(s)
 {
-    var length = 0;
-    for(var i=0; i<s.length; i++) {
-        var c = s.charCodeAt(i);
-        length += c<128?1:2;
-    }
-    return length;
+    return texten.encode(s).length;
 };