Disable EQUS expansion for raw symbols

Rangi42 · Feb 5, 2025 · 2cf28a7 · 2cf28a7
1 parent d9d381c
commit 2cf28a7
Show file tree

Hide file tree

Showing 3 changed files with 85 additions and 15 deletions.
diff --git a/man/rgbasm.5 b/man/rgbasm.5
@@ -1104,7 +1104,9 @@ Additionally, label names can contain up to a single dot
 .Ql \&. ,
 which may not be the first character.
 .Pp
-A symbol cannot have the same name as a reserved keyword, unless it is prefixed by a hash
+A symbol cannot have the same name as a reserved keyword, unless it is a
+.Dq raw symbol
+prefixed by a hash
 .Sq # .
 For example,
 .Ql #load
@@ -1387,6 +1389,8 @@ This expansion is disabled in a few contexts:
 and
 .Ql MACRO name
 will not expand string constants in their names.
+Expansion is also disabled for raw string constant symbols (string constant symbols prefixed by a hash
+.Sq # ) .
 .Bd -literal -offset indent
 DEF COUNTREG EQUS "[hl+]"
     ld a, COUNTREG

diff --git a/src/asm/lexer.cpp b/src/asm/lexer.cpp
@@ -1910,8 +1910,8 @@ static Token yylex_NORMAL() {
 				// `token` is either an `ID` or a `LOCAL_ID`, and both have a `std::string` value.
 				assume(token.value.holds<std::string>());
 
-				// Local symbols cannot be string expansions
-				if (token.type == T_(ID) && lexerState->expandStrings) {
+				// Raw symbols and local symbols cannot be string expansions
+				if (!raw && token.type == T_(ID) && lexerState->expandStrings) {
 					// Attempt string expansion
 					Symbol const *sym = sym_FindExactSymbol(token.value.get<std::string>());
 

diff --git a/src/asm/parser.y b/src/asm/parser.y
@@ -342,6 +342,7 @@
 
 // Strings
 %type <std::string> string
+%type <std::string> string_literal
 %type <std::string> strcat_args
 // Strings used for identifiers
 %type <std::string> def_id
@@ -1226,10 +1227,20 @@ print_expr:
 	relocexpr_no_str {
 		printf("$%" PRIX32, $1.getConstVal());
 	}
-	| string {
+	| string_literal {
 		// Allow printing NUL characters
 		fwrite($1.data(), 1, $1.length(), stdout);
 	}
+	| scoped_anon_id {
+		if (Symbol *sym = sym_FindScopedSymbol($1); sym && sym->type == SYM_EQUS) {
+			std::shared_ptr<std::string> str = sym->getEqus();
+			fwrite(str->data(), 1, str->length(), stdout);
+		} else {
+			Expression expr;
+			expr.makeSymbol($1);
+			printf("$%" PRIX32, expr.getConstVal());
+		}
+	}
 ;
 
 bit_const:
@@ -1252,10 +1263,21 @@ constlist_8bit_entry:
 		$1.checkNBit(8);
 		sect_RelByte($1, 0);
 	}
-	| string {
+	| string_literal {
 		std::vector<int32_t> output = charmap_Convert($1);
 		sect_ByteString(output);
 	}
+	| scoped_anon_id {
+		if (Symbol *sym = sym_FindScopedSymbol($1); sym && sym->type == SYM_EQUS) {
+			std::vector<int32_t> output = charmap_Convert(*sym->getEqus());
+			sect_ByteString(output);
+		} else {
+			Expression expr;
+			expr.makeSymbol($1);
+			expr.checkNBit(8);
+			sect_RelByte(expr, 0);
+		}
+	}
 ;
 
 constlist_16bit:
@@ -1268,10 +1290,21 @@ constlist_16bit_entry:
 		$1.checkNBit(16);
 		sect_RelWord($1, 0);
 	}
-	| string {
+	| string_literal {
 		std::vector<int32_t> output = charmap_Convert($1);
 		sect_WordString(output);
 	}
+	| scoped_anon_id {
+		if (Symbol *sym = sym_FindScopedSymbol($1); sym && sym->type == SYM_EQUS) {
+			std::vector<int32_t> output = charmap_Convert(*sym->getEqus());
+			sect_WordString(output);
+		} else {
+			Expression expr;
+			expr.makeSymbol($1);
+			expr.checkNBit(16);
+			sect_RelWord(expr, 0);
+		}
+	}
 ;
 
 constlist_32bit:
@@ -1283,10 +1316,20 @@ constlist_32bit_entry:
 	relocexpr_no_str {
 		sect_RelLong($1, 0);
 	}
-	| string {
+	| string_literal {
 		std::vector<int32_t> output = charmap_Convert($1);
 		sect_LongString(output);
 	}
+	| scoped_anon_id {
+		if (Symbol *sym = sym_FindScopedSymbol($1); sym && sym->type == SYM_EQUS) {
+			std::vector<int32_t> output = charmap_Convert(*sym->getEqus());
+			sect_LongString(output);
+		} else {
+			Expression expr;
+			expr.makeSymbol($1);
+			sect_RelLong(expr, 0);
+		}
+	}
 ;
 
 reloc_8bit:
@@ -1318,17 +1361,17 @@ relocexpr:
 	relocexpr_no_str {
 		$$ = std::move($1);
 	}
-	| string {
+	| string_literal {
 		std::vector<int32_t> output = charmap_Convert($1);
 		$$.makeNumber(strToNum(output));
 	}
+	| scoped_anon_id {
+		$$.makeSymbol($1);
+	}
 ;
 
 relocexpr_no_str:
-	scoped_anon_id {
-		$$.makeSymbol($1);
-	}
-	| NUMBER {
+	NUMBER {
 		$$.makeNumber($1);
 	}
 	| OP_LOGICNOT relocexpr %prec NEG {
@@ -1422,7 +1465,7 @@ relocexpr_no_str:
 		// '@' is also an ID; it is handled here
 		$$.makeBankSymbol($3);
 	}
-	| OP_BANK LPAREN string RPAREN {
+	| OP_BANK LPAREN string_literal RPAREN {
 		$$.makeBankSection($3);
 	}
 	| OP_SIZEOF LPAREN string RPAREN {
@@ -1543,7 +1586,7 @@ precision_arg:
 	}
 ;
 
-string:
+string_literal:
 	STRING {
 		$$ = std::move($1);
 	}
@@ -1606,6 +1649,19 @@ string:
 	}
 ;
 
+string:
+	string_literal {
+		$$ = std::move($1);
+	}
+	| scoped_anon_id {
+		if (Symbol *sym = sym_FindScopedSymbol($1); sym && sym->type == SYM_EQUS) {
+			$$ = *sym->getEqus();
+		} else {
+			::error("Symbol \"%s\" is not a string symbol\n", $1.c_str());
+		}
+	}
+;
+
 strcat_args:
 	string {
 		$$ = std::move($1);
@@ -1630,10 +1686,20 @@ strfmt_va_args:
 		$$ = std::move($1);
 		$$.args.push_back(static_cast<uint32_t>($3.getConstVal()));
 	}
-	| strfmt_va_args COMMA string {
+	| strfmt_va_args COMMA string_literal {
 		$$ = std::move($1);
 		$$.args.push_back(std::move($3));
 	}
+	| strfmt_va_args COMMA scoped_anon_id {
+		$$ = std::move($1);
+		if (Symbol *sym = sym_FindScopedSymbol($3); sym && sym->type == SYM_EQUS) {
+			$$.args.push_back(*sym->getEqus());
+		} else {
+			Expression expr;
+			expr.makeSymbol($3);
+			$$.args.push_back(static_cast<uint32_t>(expr.getConstVal()));
+		}
+	}
 ;
 
 section: