doc/parser_grammar.txt

@@ -1,13 +1,13 @@
 <program>   ::= <statement>*
-<statement> ::= <label> | <directive> | <instruction>
+<statement> ::= <label> | <directive> | <instruction> | <newline>
 
 <label> ::= <identifier> <colon>
 
-<directive> ::= <dot> <section_directive>
+<directive> ::= <dot> <section_directive> <newline>
 
 <section_directive> ::= "section" <identifier>
 
-<instruction> ::= <identifier> <operands>
+<instruction> ::= <identifier> <operands> <newline>
 
 <operands> ::= <operand> ( <comma> <operand> )*
 

src/ast.c

@@ -157,6 +157,8 @@ const char *ast_node_id_to_cstr(node_id_t id) {
         return "NODE_ASTERISK";
     case NODE_DOT:
         return "NODE_DOT";
+    case NODE_NEWLINE:
+        return "NODE_NEWLINE";
     }
     assert(!"Unreachable, weird node id" && id);
     __builtin_unreachable();
@@ -172,7 +174,8 @@ static void ast_node_print_internal(ast_node_t *node, int indent) {
     }
     printf("%s", ast_node_id_to_cstr(node->id));
 
-    if (node->token_entry && node->token_entry->token.value) {
+    if (node->token_entry && node->token_entry->token.value &&
+        node->id != NODE_NEWLINE) {
         printf(" \"%s\"", node->token_entry->token.value);
     }
     printf("\n");

src/ast.h

@@ -50,6 +50,7 @@ typedef enum node_id {
     NODE_MINUS,
     NODE_ASTERISK,
     NODE_DOT,
+    NODE_NEWLINE,
 } node_id_t;
 
 typedef struct ast_node ast_node_t;

src/parser/parser.c

@@ -120,18 +120,20 @@ parse_result_t parse_section_directive(tokenlist_entry_t *current) {
 }
 
 parse_result_t parse_directive(tokenlist_entry_t *current) {
-    parser_t parsers[] = {parse_dot, parse_section_directive, nullptr};
+    parser_t parsers[] = {parse_dot, parse_section_directive, parse_newline,
+                          nullptr};
     return parse_consecutive(current, NODE_DIRECTIVE, parsers);
 }
 
 parse_result_t parse_instruction(tokenlist_entry_t *current) {
-    parser_t parsers[] = {parse_identifier, parse_operands, nullptr};
+    parser_t parsers[] = {parse_identifier, parse_operands, parse_newline,
+                          nullptr};
     return parse_consecutive(current, NODE_INSTRUCTION, parsers);
 }
 
 parse_result_t parse_statement(tokenlist_entry_t *current) {
     parser_t parsers[] = {parse_label, parse_directive, parse_instruction,
-                          nullptr};
+                          parse_newline, nullptr};
     return parse_any(current, parsers);
 }
 

src/parser/primitives.c

@@ -62,6 +62,10 @@ parse_result_t parse_dot(tokenlist_entry_t *current) {
     return parse_token(current, TOKEN_DOT, NODE_DOT, nullptr);
 }
 
+parse_result_t parse_newline(tokenlist_entry_t *current) {
+    return parse_token(current, TOKEN_NEWLINE, NODE_NEWLINE, nullptr);
+}
+
 parse_result_t parse_label_reference(tokenlist_entry_t *current) {
     return parse_token(current, TOKEN_IDENTIFIER, NODE_LABEL_REFERENCE,
                        nullptr);

src/parser/primitives.h

@@ -18,6 +18,7 @@ parse_result_t parse_plus(tokenlist_entry_t *current);
 parse_result_t parse_minus(tokenlist_entry_t *current);
 parse_result_t parse_asterisk(tokenlist_entry_t *current);
 parse_result_t parse_dot(tokenlist_entry_t *current);
+parse_result_t parse_newline(tokenlist_entry_t *current);
 parse_result_t parse_label_reference(tokenlist_entry_t *current);
 
 /* These are "primitives" with a different name and some extra validation on top

src/tokenlist.c

@@ -86,7 +86,6 @@ bool is_trivia(tokenlist_entry_t *trivia) {
     switch (trivia->token.id) {
     case TOKEN_WHITESPACE:
     case TOKEN_COMMENT:
-    case TOKEN_NEWLINE:
         return true;
     default:
         return false;