Clean up text handling a bit:

- Rename IO.write -> IO.print.
- Make IO.write not print a newline.
- Support \u Unicode escapes in strings.

src/wren_compiler.c

@@ -514,10 +514,71 @@ static void readName(Parser* parser, TokenType type)
   makeToken(parser, type);
 }
 
-// Adds [c] to the current string literal being tokenized.
-static void addStringChar(Parser* parser, char c)
+// Adds [c] to the current string literal being tokenized. If [c] is outside of
+// ASCII range, it will emit the UTF-8 encoded byte sequence for it.
+static void addStringChar(Parser* parser, uint32_t c)
 {
-  wrenByteBufferWrite(parser->vm, &parser->string, c);
+  ByteBuffer* buffer = &parser->string;
+
+  if (c <= 0x7f)
+  {
+    // Single byte (i.e. fits in ASCII).
+    wrenByteBufferWrite(parser->vm, buffer, c);
+  }
+  else if (c <= 0x7ff)
+  {
+    // Two byte sequence: 110xxxxx	 10xxxxxx.
+    wrenByteBufferWrite(parser->vm, buffer, 0xc0 | ((c & 0x7c0) >> 6));
+    wrenByteBufferWrite(parser->vm, buffer, 0x80 | (c & 0x3f));
+  }
+  else if (c <= 0xffff)
+  {
+    // Three byte sequence: 1110xxxx	 10xxxxxx 10xxxxxx.
+    wrenByteBufferWrite(parser->vm, buffer, 0xe0 | ((c & 0xf000) >> 12));
+    wrenByteBufferWrite(parser->vm, buffer, 0x80 | ((c & 0xfc0) >> 6));
+    wrenByteBufferWrite(parser->vm, buffer, 0x80 | (c & 0x3f));
+  }
+  else if (c <= 0x10ffff)
+  {
+    // Four byte sequence: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx.
+    wrenByteBufferWrite(parser->vm, buffer, 0xf0 | ((c & 0x1c0000) >> 18));
+    wrenByteBufferWrite(parser->vm, buffer, 0x80 | ((c & 0x3f000) >> 12));
+    wrenByteBufferWrite(parser->vm, buffer, 0x80 | ((c & 0xfc0) >> 6));
+    wrenByteBufferWrite(parser->vm, buffer, 0x80 | (c & 0x3f));
+  }
+  else
+  {
+    // Invalid Unicode value. See: http://tools.ietf.org/html/rfc3629
+    // TODO: Error.
+  }
+}
+
+// Reads the next character, which should be a hex digit (0-9, a-f, or A-F) and
+// returns its numeric value. If the character isn't a hex digit, returns -1.
+static int readHexDigit(Parser* parser)
+{
+  char c = nextChar(parser);
+  if (c >= '0' && c <= '9') return c - '0';
+  if (c >= 'a' && c <= 'f') return c - 'a' + 10;
+  if (c >= 'A' && c <= 'F') return c - 'A' + 10;
+  return -1;
+}
+
+// Reads a four hex digit Unicode escape sequence in a string literal.
+static void readUnicodeEscape(Parser* parser)
+{
+  int value = 0;
+  for (int i = 0; i < 4; i++)
+  {
+    char digit = readHexDigit(parser);
+    // TODO: Signal error.
+    if (digit == -1) return;
+
+    value = (value * 16) | digit;
+  }
+
+  // TODO: Handle encoding!
+  addStringChar(parser, value);
 }
 
 // Finishes lexing a string literal.
@@ -538,6 +599,7 @@ static void readString(Parser* parser)
         case '\\': addStringChar(parser, '\\'); break;
         case 'n':  addStringChar(parser, '\n'); break;
         case 't':  addStringChar(parser, '\t'); break;
+        case 'u':  readUnicodeEscape(parser); break;
         default:
           // TODO: Emit error token.
           break;

src/wren_core.c

@@ -43,8 +43,14 @@
 // make_corelib. Do not edit here.
 const char* coreLibSource =
 "class IO {\n"
+"  static print(obj) {\n"
+"    IO.writeString_(obj.toString)\n"
+"    IO.writeString_(\"\n\")\n"
+"    return obj\n"
+"  }\n"
+"\n"
 "  static write(obj) {\n"
-"    IO.write__native__(obj.toString)\n"
+"    IO.writeString_(obj.toString)\n"
 "    return obj\n"
 "  }\n"
 "}\n"
@@ -471,9 +477,9 @@ DEF_NATIVE(string_subscript)
 
 DEF_NATIVE(io_writeString)
 {
+  if (!validateString(vm, args, 1, "Argument")) return PRIM_ERROR;
   wrenPrintValue(args[1]);
-  printf("\n");
-  RETURN_VAL(args[1]);
+  RETURN_NULL;
 }
 
 DEF_NATIVE(os_clock)
@@ -621,5 +627,5 @@ void wrenInitializeCore(WrenVM* vm)
   NATIVE(vm->numClass, "!= ", num_bangeq);
 
   ObjClass* ioClass = AS_CLASS(findGlobal(vm, "IO"));
-  NATIVE(ioClass->metaclass, "write__native__ ", io_writeString);
+  NATIVE(ioClass->metaclass, "writeString_ ", io_writeString);
 }