Use emscripten's utf8tojs string function

4 files changed, 47 insertions, 8 deletions

diff --git a/src/lobject.js b/src/lobject.js
index 739db50..556da08 100644
--- a/src/lobject.js
+++ b/src/lobject.js
@@ -111,7 +111,44 @@ class TValue {
     }
 
     jsstring() {
-        return this.ttisstring() ? String.fromCharCode(...this.value) : null;
+        //return this.ttisstring() ? String.fromCharCode(...this.value) : null;
+        let u0, u1, u2, u3, u4, u5;
+        let idx = 0;
+
+        var str = '';
+        while (1) {
+            // For UTF8 byte structure, see http://en.wikipedia.org/wiki/UTF-8#Description and https://www.ietf.org/rfc/rfc2279.txt and https://tools.ietf.org/html/rfc3629
+            u0 = this.value[idx++];
+            if (!u0) return str;
+            if (!(u0 & 0x80)) { str += String.fromCharCode(u0); continue; }
+            u1 = this.value[idx++] & 63;
+            if ((u0 & 0xE0) == 0xC0) { str += String.fromCharCode(((u0 & 31) << 6) | u1); continue; }
+            u2 = this.value[idx++] & 63;
+            if ((u0 & 0xF0) == 0xE0) {
+                u0 = ((u0 & 15) << 12) | (u1 << 6) | u2;
+            } else {
+                u3 = this.value[idx++] & 63;
+                if ((u0 & 0xF8) == 0xF0) {
+                    u0 = ((u0 & 7) << 18) | (u1 << 12) | (u2 << 6) | u3;
+                } else {
+                    u4 = this.value[idx++] & 63;
+                    if ((u0 & 0xFC) == 0xF8) {
+                        u0 = ((u0 & 3) << 24) | (u1 << 18) | (u2 << 12) | (u3 << 6) | u4;
+                    } else {
+                        u5 = this.value[idx++] & 63;
+                        u0 = ((u0 & 1) << 30) | (u1 << 24) | (u2 << 18) | (u3 << 12) | (u4 << 6) | u5;
+                    }
+                }
+            }
+            if (u0 < 0x10000) {
+                str += String.fromCharCode(u0);
+            } else {
+                var ch = u0 - 0x10000;
+                str += String.fromCharCode(0xD800 | (ch >> 10), 0xDC00 | (ch & 0x3FF));
+            }
+        }
+
+        return str;
     }
 
 }
diff --git a/src/ltm.js b/src/ltm.js
index 6377c93..e48ff23 100644
--- a/src/ltm.js
+++ b/src/ltm.js
@@ -89,7 +89,7 @@ const ttypename = function(t) {
 const luaT_init = function(L) {
     L.l_G.tmname = [];
     for (let event in TMS) {
-        let name = lua.to_luastring(TMS[event], TMS[event].length);
+        let name = lua.to_luastring(TMS[event]);
         L.l_G.tmname.push(L.l_G.intern(name)); // Strings are already interned by JS
     }
 };
diff --git a/src/lua.js b/src/lua.js
index 08bad58..be6cc1e 100644
--- a/src/lua.js
+++ b/src/lua.js
@@ -133,7 +133,7 @@ class lua_Debug {
 }
 
 const to_luastring = function(str, maxBytesToWrite) {
-    maxBytesToWrite = maxBytesToWrite !== undefined ? maxBytesToWrite : str.length + 1;
+    maxBytesToWrite = maxBytesToWrite !== undefined ? maxBytesToWrite : Number.MAX_SAFE_INTEGER;
     let outU8Array = [];
 
     if (!(maxBytesToWrite > 0)) // Parameter maxBytesToWrite is not optional. Negative values, 0, null, undefined and false each don't write out any bytes.
diff --git a/src/lutf8lib.js b/src/lutf8lib.js
index bc43d54..db5f7ad 100644
--- a/src/lutf8lib.js
+++ b/src/lutf8lib.js
@@ -8,7 +8,8 @@ const lauxlib = require('./lauxlib.js');
 
 
 const iscont = function(p) {
-    return p & 0xC0 === 0x80;
+    let c = p & 0xC0;
+    return c === 0x80;
 };
 
 /* translate a relative string position: negative means back from end */
@@ -24,6 +25,7 @@ const u_posrelat = function(pos, len) {
 */
 const byteoffset = function(L) {
     let s = lauxlib.luaL_checkstring(L, 1);
+    s = L.stack[lapi.index2addr_(L, 1)].value;
     let n = lauxlib.luaL_checkinteger(L, 2);
     let posi = n >= 0 ? 1 : s.length + 1;
     posi = u_posrelat(lauxlib.luaL_optinteger(L, 3, posi), s.length);
@@ -31,16 +33,16 @@ const byteoffset = function(L) {
 
     if (n === 0) {
         /* find beginning of current byte sequence */
-        while (posi > 0 && iscont(s.slice(posi))) posi--;
+        while (posi > 0 && iscont(s[posi])) posi--;
     } else {
-        if (iscont(s.slice(posi)))
+        if (iscont(s[posi]))
             lauxlib.luaL_error(L, "initial position is a continuation byte");
 
         if (n < 0) {
             while (n < 0 && posi > 0) {  /* move back */
                 do {  /* find beginning of previous character */
                     posi--;
-                } while (posi > 0 && iscont(s.slice(posi)));
+                } while (posi > 0 && iscont(s[posi]));
                 n++;
             }
         } else {
@@ -48,7 +50,7 @@ const byteoffset = function(L) {
             while (n > 0 && posi < s.length) {
                 do {  /* find beginning of next character */
                     posi++;
-                } while (iscont(s.slice(posi)));  /* (cannot pass final '\0') */
+                } while (iscont(s[posi]));  /* (cannot pass final '\0') */
                 n--;
             }
         }