Implement quoted identifier parsing (#2715)

Part of annotations and WebAssembly 3.0.

Author: zherczeg

include/wabt/wast-lexer.h

@@ -89,7 +89,7 @@ class WastLexer {
     return ReadReservedChars() == ReservedChars::None;
   }
   void ReadSign();
-  Token GetStringToken();
+  Token GetStringToken(TokenType);
   Token GetNumberToken(TokenType);
   Token GetHexNumberToken(TokenType);
   Token GetInfToken();

include/wabt/wast-parser.h

@@ -132,6 +132,9 @@ class WastParser {
   // Returns true if the next two tokens are form reference type - (ref $t)
   bool PeekMatchRefType();
 
+  // Returns true if the next token represents a var.
+  bool PeekMatchVar();
+
   // Returns true if the next token's type is equal to the parameter. If so,
   // then the token is consumed.
   bool Match(TokenType);
@@ -159,17 +162,18 @@ class WastParser {
   // synchronized.
   Result Synchronize(SynchronizeFunc);
 
-  bool ParseBindVarOpt(std::string* name);
+  Result ParseVarText(Token& token, std::string* out_text);
+  Result ParseBindVarOpt(std::string* name);
   Result ParseVar(Var* out_var);
-  bool ParseVarOpt(Var* out_var, Var default_var = Var());
+  Result ParseVarOpt(Var* out_var, Var default_var = Var());
   Result ParseOffsetExpr(ExprList* out_expr_list);
   bool ParseOffsetExprOpt(ExprList* out_expr_list);
   Result ParseTextList(std::vector<uint8_t>* out_data);
   bool ParseTextListOpt(std::vector<uint8_t>* out_data);
   Result ParseVarList(VarVector* out_var_list);
   bool ParseElemExprOpt(ExprList* out_elem_expr);
   bool ParseElemExprListOpt(ExprListVector* out_list);
-  bool ParseElemExprVarListOpt(ExprListVector* out_list);
+  Result ParseElemExprVarListOpt(ExprListVector* out_list);
   Result ParseRefDeclaration(Var* out_type);
   Result ParseValueType(Var* out_type);
   Result ParseValueTypeList(

src/wast-lexer.cc

@@ -111,7 +111,7 @@ Token WastLexer::GetToken() {
         continue;
 
       case '"':
-        return GetStringToken();
+        return GetStringToken(TokenType::Text);
 
       case '+':
       case '-':
@@ -158,6 +158,10 @@ Token WastLexer::GetToken() {
         return GetNumberToken(TokenType::Nat);
 
       case '$':
+        ReadChar();
+        if (PeekChar() == '"') {
+          return GetStringToken(TokenType::Var);
+        }
         return GetIdChars();  // Initial $ is idchar, so this produces id token
 
       case 'a':
@@ -319,7 +323,7 @@ void WastLexer::ReadWhitespace() {
   }
 }
 
-Token WastLexer::GetStringToken() {
+Token WastLexer::GetStringToken(TokenType token_type) {
   const char* saved_token_start = token_start_;
   bool has_error = false;
   bool in_string = true;
@@ -437,7 +441,7 @@ Token WastLexer::GetStringToken() {
     return Token(GetLocation(), TokenType::Invalid);
   }
 
-  return TextToken(TokenType::Text);
+  return TextToken(token_type);
 }
 
 // static
@@ -497,7 +501,7 @@ WastLexer::ReservedChars WastLexer::ReadReservedChars() {
         ret = ReservedChars::Id;
       }
     } else if (peek == '"') {
-      GetStringToken();
+      GetStringToken(TokenType::Text);
       ret = ReservedChars::Some;
     } else {
       break;
@@ -602,7 +606,7 @@ Token WastLexer::GetNameEqNumToken(std::string_view name,
 }
 
 Token WastLexer::GetIdChars() {
-  if (ReadReservedChars() == ReservedChars::Id) {
+  if (ReadReservedChars() != ReservedChars::Some) {
     return TextToken(TokenType::Var);
   }