parser: create a block node type

Since the next step to pipa programming language is about having control
flow statements we could benefit ourselves by having a block node to
control scope. Now, functions has a block node, instead of an vector as
body. As you can see through the ast-dump:

  FunctionDecl name='main'
  └─ body:
    └─ Block
      └─ ReturnStmt
        └─ Literal type=i32 value='69'

This same node kind can be used for parsing if, for and while blocks.

I could use ast_block_t as body for functions but instead, I opted to
use an ast_node_t. This brings the flexibility to, in the future, having
another function body kinds, such as arrow functions if we want to:

  fn add(a: i32, b: i32): i32 => a + b;

Signed-off-by: Carlos Maniero <carlos@maniero.me>

6 files changed, 87 insertions, 24 deletions

diff --git a/src/ast.c b/src/ast.c
index 4fd5a3f..1200e30 100644
--- a/src/ast.c
+++ b/src/ast.c
@@ -45,7 +45,10 @@ ast_node_destroy(ast_node_t *node)
 {
   switch (node->kind) {
     case AST_FUNCTION_DECLARATION:
-      ast_node_destroy_vector(node->data.function.body);
+      ast_node_destroy(node->data.function.body);
+      break;
+    case AST_BLOCK:
+      ast_node_destroy_vector(node->data.block.body);
       break;
     case AST_BINARY_OPERATION:
       ast_node_destroy(node->data.binary_operation.left);
@@ -82,7 +85,7 @@ ast_node_new_return_stmt(ast_node_t *argument)
 }
 
 ast_node_t *
-ast_node_new_function_declaration(string_view_t function_name, type_t return_type, vector_t *body)
+ast_node_new_function_declaration(string_view_t function_name, type_t return_type, ast_node_t *body)
 {
   ast_node_t *node = ast_node_new();
 
@@ -101,6 +104,23 @@ ast_node_new_function_declaration(string_view_t function_name, type_t return_typ
 }
 
 ast_node_t *
+ast_node_new_block(vector_t *body)
+{
+  ast_node_t *node = ast_node_new();
+
+  *node = (ast_node_t){
+    .kind = AST_BLOCK,
+    .data = {
+      .block = {
+        .body = body,
+      }
+    },
+  };
+
+  return node;
+}
+
+ast_node_t *
 ast_node_new_literal_integer(uint32_t number)
 {
   ast_node_t *node = ast_node_new();
diff --git a/src/ast.h b/src/ast.h
index f95aee2..1637173 100644
--- a/src/ast.h
+++ b/src/ast.h
@@ -29,6 +29,11 @@ typedef enum
 
 typedef struct ast_node_t ast_node_t;
 
+typedef struct ast_block_t
+{
+  vector_t *body;
+} ast_block_t;
+
 typedef struct ast_return_stmt_t
 {
   ast_node_t *argument;
@@ -48,7 +53,7 @@ typedef struct ast_function_declaration_t
 {
   ast_identifier_t identifier;
   type_t return_type;
-  vector_t *body;
+  ast_node_t *body;
 } ast_function_declaration_t;
 
 typedef enum ast_binary_operation_kind_t
@@ -109,6 +114,7 @@ typedef struct ast_variable_assignment_t
 typedef enum
 {
   AST_BINARY_OPERATION,
+  AST_BLOCK,
   AST_FUNCTION_DECLARATION,
   AST_LITERAL,
   AST_RETURN_STMT,
@@ -123,6 +129,7 @@ typedef union
   ast_binary_operation_t binary_operation;
   ast_function_declaration_t function;
   ast_literal_t literal;
+  ast_block_t block;
   ast_return_stmt_t return_stmt;
   ast_variable_declaration_t variable_declaration;
   ast_variable_assignment_t variable_assignment;
@@ -155,7 +162,7 @@ ast_node_new_binary_operation(ast_binary_operation_kind_t kind,
                               type_t result_type);
 
 ast_node_t *
-ast_node_new_function_declaration(string_view_t function_name, type_t return_type, vector_t *body);
+ast_node_new_function_declaration(string_view_t function_name, type_t return_type, ast_node_t *body);
 
 ast_node_t *
 ast_node_new_return_stmt(ast_node_t *argument);
@@ -167,6 +174,9 @@ ast_node_t *
 ast_node_new_literal_integer(uint32_t number);
 
 ast_node_t *
+ast_node_new_block(vector_t *body);
+
+ast_node_t *
 ast_node_new_literal_bool(bool boolean);
 
 ast_node_t *
diff --git a/src/ast_pretty_printer.c b/src/ast_pretty_printer.c
index 95f5993..718d22a 100644
--- a/src/ast_pretty_printer.c
+++ b/src/ast_pretty_printer.c
@@ -99,13 +99,33 @@ ast_pretty_printer_print_ast(ast_pretty_printer_t *printer, ast_node_t *ast)
       ast_pretty_printer_printf(printer, "FunctionDecl name='" SVFMT "'\n", SVARG(&function.identifier.name));
 
       ast_pretty_printer_add_indentation(printer);
-      for (size_t i = 0; i < function.body->size; ++i) {
-        if (i + 1 >= function.body->size) {
-          ast_pretty_printer_set_lst_children(printer);
+      {
+        ast_pretty_printer_set_lst_children(printer);
+        ast_pretty_printer_printf(printer, "body:\n");
+
+        ast_pretty_printer_add_indentation(printer);
+        {
+          ast_pretty_printer_print_ast(printer, ast->data.function.body);
+          ast_pretty_printer_rm_indentation(printer);
+        }
+
+        ast_pretty_printer_rm_indentation(printer);
+      }
+      break;
+    }
+    case AST_BLOCK: {
+      ast_pretty_printer_set_lst_children(printer);
+      ast_pretty_printer_printf(printer, "Block\n");
+      ast_pretty_printer_add_indentation(printer);
+      {
+        for (size_t i = 0; i < ast->data.block.body->size; ++i) {
+          if (i + 1 >= ast->data.block.body->size) {
+            ast_pretty_printer_set_lst_children(printer);
+          }
+          ast_pretty_printer_print_ast(printer, vector_at(ast->data.block.body, i));
         }
-        ast_pretty_printer_print_ast(printer, vector_at(function.body, i));
+        ast_pretty_printer_rm_indentation(printer);
       }
-      ast_pretty_printer_rm_indentation(printer);
       break;
     }
     case AST_LITERAL: {
diff --git a/src/gas_assembly_generator.c b/src/gas_assembly_generator.c
index d9feedb..0cce205 100644
--- a/src/gas_assembly_generator.c
+++ b/src/gas_assembly_generator.c
@@ -42,6 +42,9 @@ gas_assembly_generator_compile_variable_assignment(gas_assembly_generator_t *gen
                                                    ast_variable_assignment_t *variable_assignment);
 
 static void
+gas_assembly_generator_compile_block(gas_assembly_generator_t *gen, ast_block_t *block);
+
+static void
 gas_assembly_generator_compile_variable(gas_assembly_generator_t *gen, ast_variable_t *variable);
 
 static void
@@ -132,6 +135,9 @@ gas_assembly_generator_compile(gas_assembly_generator_t *gen, ast_node_t *ast)
     case AST_VARIABLE:
       gas_assembly_generator_compile_variable(gen, &ast->data.variable);
       break;
+    case AST_BLOCK:
+      gas_assembly_generator_compile_block(gen, &ast->data.block);
+      break;
     case AST_UNKOWN_NODE:
       assert(false && "unreachable");
   }
@@ -153,9 +159,7 @@ gas_assembly_generator_compile_function(gas_assembly_generator_t *gen, ast_funct
   fprintf(gen->stream, "  push %%rbp\n");
   fprintf(gen->stream, "  mov %%rsp, %%rbp\n");
 
-  for (size_t i = 0; i < func->body->size; i++) {
-    gas_assembly_generator_compile(gen, vector_at(func->body, i));
-  }
+  gas_assembly_generator_compile(gen, func->body);
 
   fprintf(gen->stream, "  pop %%rbp\n");
 }
@@ -258,6 +262,14 @@ gas_assembly_generator_compile_literal(gas_assembly_generator_t *gen, ast_litera
 }
 
 static void
+gas_assembly_generator_compile_block(gas_assembly_generator_t *gen, ast_block_t *block)
+{
+  for (size_t i = 0; i < block->body->size; i++) {
+    gas_assembly_generator_compile(gen, vector_at(block->body, i));
+  }
+}
+
+static void
 gas_assembly_generator_binary_operation(gas_assembly_generator_t *gen, ast_binary_operation_t *binary_operation)
 {
   gas_assembly_generator_compile(gen, binary_operation->right);
diff --git a/src/parser.c b/src/parser.c
index 2817fa0..848bb4a 100644
--- a/src/parser.c
+++ b/src/parser.c
@@ -516,9 +516,11 @@ parser_error_report_unexpected_token(parser_t *parser)
 }
 
 static bool
-parser_ensure_function_return_statement(parser_t *parser, vector_t *body, token_t *function_token)
+parser_ensure_function_return_statement(parser_t *parser, ast_node_t *block, token_t *function_token)
 {
-  ast_node_t *latest_node = vector_at(body, body->size - 1);
+  assert(block->kind == AST_BLOCK);
+
+  ast_node_t *latest_node = vector_at(block->data.block.body, block->data.block.body->size - 1);
 
   if (latest_node->kind != AST_RETURN_STMT) {
     parser_error_t error;
@@ -527,14 +529,13 @@ parser_ensure_function_return_statement(parser_t *parser, vector_t *body, token_
     sprintf(error.message, "expected 'return' keyword.");
 
     parser->errors[parser->errors_len++] = error;
-    vector_destroy(body);
     return false;
   }
 
   return true;
 }
 
-static vector_t *
+static ast_node_t *
 parser_parse_block_declarations(parser_t *parser, type_t result_type)
 {
   if (!drop_expected_token(parser, TOKEN_OCURLY)) {
@@ -602,7 +603,7 @@ parser_parse_block_declarations(parser_t *parser, type_t result_type)
     return NULL;
   }
 
-  return body;
+  return ast_node_new_block(body);
 }
 
 static bool
@@ -638,14 +639,14 @@ parser_parse_function_declaration(parser_t *parser)
     return NULL;
   }
 
-  vector_t *body = parser_parse_block_declarations(parser, return_type);
+  ast_node_t *body = parser_parse_block_declarations(parser, return_type);
 
   if (body == NULL) {
     return NULL;
   }
 
   if (!parser_ensure_function_return_statement(parser, body, &func_name_token)) {
-    vector_destroy(body);
+    ast_node_destroy(body);
     return NULL;
   }
 
diff --git a/test/parser_test.c b/test/parser_test.c
index 830fed7..2d8f0bb 100644
--- a/test/parser_test.c
+++ b/test/parser_test.c
@@ -76,7 +76,7 @@ test_parse_function(const MunitParameter params[], void *user_data_or_fixture)
   assert_string_equal("main", actual);
   assert_int(AST_FUNCTION_DECLARATION, ==, ast_function->kind);
 
-  ast_node_t *ast_return = vector_at(ast_function->data.function.body, 0);
+  ast_node_t *ast_return = vector_at(ast_function->data.function.body->data.block.body, 0);
 
   assert_int(AST_RETURN_STMT, ==, ast_return->kind);
 
@@ -110,12 +110,12 @@ test_parse_variable_definition(const MunitParameter params[], void *user_data_or
   assert_string_equal("main", actual);
   assert_int(AST_FUNCTION_DECLARATION, ==, ast_function->kind);
 
-  ast_node_t *ast_variable = vector_at(ast_function->data.function.body, 0);
+  ast_node_t *ast_variable = vector_at(ast_function->data.function.body->data.block.body, 0);
 
   assert_int(AST_VARIABLE_DECLARATION, ==, ast_variable->kind);
   assert_string_view_equal("variable", ast_variable->data.variable_declaration.identifier.name);
 
-  ast_node_t *ast_return = vector_at(ast_function->data.function.body, 1);
+  ast_node_t *ast_return = vector_at(ast_function->data.function.body->data.block.body, 1);
 
   assert_int(AST_RETURN_STMT, ==, ast_return->kind);
 
@@ -146,7 +146,7 @@ test_parse_boolean(const MunitParameter params[], void *user_data_or_fixture)
   assert_string_view_equal("my_bool_fn", ast_function->data.function.identifier.name);
   assert_int(AST_FUNCTION_DECLARATION, ==, ast_function->kind);
 
-  ast_node_t *ast_variable = vector_at(ast_function->data.function.body, 0);
+  ast_node_t *ast_variable = vector_at(ast_function->data.function.body->data.block.body, 0);
 
   assert_int(AST_VARIABLE_DECLARATION, ==, ast_variable->kind);
 
@@ -156,7 +156,7 @@ test_parse_boolean(const MunitParameter params[], void *user_data_or_fixture)
   assert_int(AST_LITERAL_BOOL, ==, ast_variable->data.variable_declaration.value->data.literal.kind);
   assert_true(ast_variable->data.variable_declaration.value->data.literal.value.boolean);
 
-  ast_node_t *ast_return = vector_at(ast_function->data.function.body, 1);
+  ast_node_t *ast_return = vector_at(ast_function->data.function.body->data.block.body, 1);
 
   assert_int(AST_RETURN_STMT, ==, ast_return->kind);