correctly advance

2 files changed, 20 insertions, 15 deletions

diff --git a/src/tokenizer.cc b/src/tokenizer.cc
index 8725e55..acd9627 100644
--- a/src/tokenizer.cc
+++ b/src/tokenizer.cc
@@ -30,8 +30,8 @@ static inline String tokenizer_make_lexeme(const Tokenizer* tokenizer,
 }
 
 static inline Token tokenizer_make_token(const Tokenizer* tokenizer,
-                                         Token_Kind kind,
-                                        usize start, usize end) {
+                                         Token_Kind kind, usize start,
+                                         usize end) {
   assert_neq(tokenizer, nullptr);
 
   String lexeme = tokenizer_make_lexeme(tokenizer, start, end);
@@ -57,7 +57,8 @@ static bool tokenizer_get_buffer(Tokenizer* tokenizer, usize* cursor) {
 }
 
 [[nodiscard]] static bool tokenizer_char(const Tokenizer* tokenizer,
-                                            usize offset, wchar* out) {
+                                         usize offset, u8* out_nobytes,
+                                         wchar* out_char) {
   const String text = tokenizer->buffer->content;
 
   assert_ste(offset, text.length);
@@ -67,7 +68,8 @@ static bool tokenizer_get_buffer(Tokenizer* tokenizer, usize* cursor) {
   u8 nobytes = utf8_nobytes(c);
   if (nobytes > 1) panic("no support for multi-byte chars: %c:%d", c, nobytes);
 
-  *out = c;
+  *out_nobytes = nobytes;
+  *out_char = (wchar)c;
   return true;
 }
 
@@ -76,13 +78,16 @@ static void tokenizer_advance(usize* offset, usize nbytes) {
   *offset += nbytes;
 }
 
-static Token tokenizer_lex_identifier(Tokenizer* tokenizer, usize start, usize *offset) {
+static Token tokenizer_lex_identifier(Tokenizer* tokenizer, usize start,
+                                      usize* offset) {
   assert_neq(tokenizer, nullptr);
 
   wchar c;
-  while (tokenizer_char(tokenizer, *offset, &c)) {
+  u8 nobytes;
+
+  while (tokenizer_char(tokenizer, *offset, &nobytes, &c)) {
     if (!utf8_is_alnum(c) && c != '_') break;
-    tokenizer_advance(offset, 1);
+    tokenizer_advance(offset, nobytes);
   }
 
   return tokenizer_make_token(tokenizer, Token_Kind_Identifier, start, *offset);
@@ -93,12 +98,15 @@ bool tokenizer_next(Tokenizer* tokenizer, Token* out) {
   assert_neq(out, nullptr);
 
   usize cursor;
-  if(unlikely(!tokenizer_get_buffer(tokenizer, &cursor))) return false;
+  if (unlikely(!tokenizer_get_buffer(tokenizer, &cursor))) return false;
   usize advance = cursor;
 
   wchar c;
-  (void)tokenizer_char(tokenizer, advance, &c);
-  
+  u8 nobytes;
+
+  (void)tokenizer_char(tokenizer, advance, &nobytes, &c);
+  tokenizer_advance(&advance, nobytes);
+
   if (utf8_is_identifier(c)) {
     *out = tokenizer_lex_identifier(tokenizer, cursor, &advance);
     goto out;
@@ -106,9 +114,8 @@ bool tokenizer_next(Tokenizer* tokenizer, Token* out) {
 
   switch (c) {
     default:
-      tokenizer_advance(&advance, 1);
       *out = tokenizer_make_token(tokenizer, Token_Kind_Invalid_Char, cursor,
-                           advance);
+                                  advance);
   }
 
 out:
diff --git a/src/utf8.cc b/src/utf8.cc
index e2da09d..ace1ec0 100644
--- a/src/utf8.cc
+++ b/src/utf8.cc
@@ -26,9 +26,7 @@ inline bool utf8_is_identifier(wchar c) {
   return ('a' <= c && c <= 'z') || ('A' <= c && c <= 'Z');
 }
 
-inline bool utf8_is_number(wchar c) {
-  return ('0' <= c && c <= '9');
-}
+inline bool utf8_is_number(wchar c) { return ('0' <= c && c <= '9'); }
 
 inline bool utf8_is_alnum(wchar c) {
   return utf8_is_identifier(c) || utf8_is_number(c);