add: start code gen

6 files changed, 114 insertions, 98 deletions

diff --git a/src/codegen/nasm.c b/src/codegen/nasm.c
new file mode 100644
index 0000000..0782ee1
--- /dev/null
+++ b/src/codegen/nasm.c
@@ -0,0 +1,48 @@
+#include <stdio.h>
+#include <stddef.h>
+
+#include "codegen/nasm.h"
+
+static void emit_consts(struct Program* program, FILE* out)
+{
+	for (size_t i = 0; i < program->const_count; i += 1)
+	{
+		struct ConstDecl decl = program->consts[i];
+		fprintf(out, "%.*s equ %.*s\n",
+			(int)decl.name.length, decl.name.start,
+			(int)decl.value.length, decl.value.start);
+	}
+}
+
+static void emit_data(struct Program* program, FILE* out)
+{
+	fprintf(out, "section .data\n");
+
+	for (size_t i = 0; i < program->data_count; i += 1)
+	{
+		struct DataDecl decl = program->data_decls[i];
+
+		// the value lexeme keeps its surrounding double quotes; NASM backtick
+		// strings interpret the same escapes, so re-wrap the inner content
+		fprintf(out, "%.*s: db `%.*s`\n",
+			(int)decl.name.length, decl.name.start,
+			(int)(decl.value.length - 2), decl.value.start + 1);
+		fprintf(out, ".len equ $ - %.*s\n",
+			(int)decl.name.length, decl.name.start);
+	}
+}
+
+void generate_nasm(struct Program* program, FILE* out)
+{
+	if (program->const_count > 0)
+	{
+		emit_consts(program, out);
+		fprintf(out, "\n");
+	}
+
+	emit_data(program, out);
+	fprintf(out, "\n");
+
+	fprintf(out, "section .text\n");
+	fprintf(out, "global _start\n");
+}
diff --git a/src/codegen/nasm.h b/src/codegen/nasm.h
new file mode 100644
index 0000000..b9c2c90
--- /dev/null
+++ b/src/codegen/nasm.h
@@ -0,0 +1,7 @@
+#pragma once
+
+#include <stdio.h>
+
+#include "parser/ast.h"
+
+void generate_nasm(struct Program* program, FILE* out);
diff --git a/src/main.c b/src/main.c
index 9a31f83..a908a73 100644
--- a/src/main.c
+++ b/src/main.c
@@ -1,80 +1,11 @@
 #include <stdio.h>
-#include <stddef.h>
 
 #include "io/file.h"
 #include "cli/args.h"
 #include "lexer/lexer.h"
+#include "codegen/nasm.h"
 #include "parser/parser.h"
 
-static void print_expr(struct Expr* expr)
-{
-	switch (expr->kind)
-	{
-		case EXPR_PRIMARY:
-			printf("%.*s", (int)expr->primary.token.length, expr->primary.token.start);
-			break;
-		case EXPR_BINARY:
-			print_expr(expr->binary.left);
-			printf(" %.*s ", (int)expr->binary.op.length, expr->binary.op.start);
-			print_expr(expr->binary.right);
-			break;
-		case EXPR_MEMBER:
-			print_expr(expr->member.object);
-			printf(".%.*s", (int)expr->member.member.length, expr->member.member.start);
-			break;
-	}
-}
-
-static void print_statement(struct Statement* statement, const char* indent)
-{
-	switch (statement->kind)
-	{
-		case STATEMENT_ASSIGN:
-		{
-			struct AssignStatement assign = statement->assign;
-			printf("%s%s%.*s %.*s ", indent,
-				assign.target_deref ? "^" : "",
-				(int)assign.target.length, assign.target.start,
-				(int)assign.op.length, assign.op.start);
-			print_expr(assign.value);
-			printf("\n");
-			break;
-		}
-		case STATEMENT_LABEL:
-			printf("%s%.*s:\n", indent, (int)statement->label.name.length, statement->label.name.start);
-			break;
-		case STATEMENT_GOTO:
-			printf("%sgoto %.*s\n", indent, (int)statement->jump.label.length, statement->jump.label.start);
-			break;
-		case STATEMENT_SYSCALL:
-			printf("%ssyscall\n", indent);
-			break;
-		case STATEMENT_IF:
-			printf("%sif ", indent);
-			print_expr(statement->branch.left);
-			printf(" %.*s ", (int)statement->branch.comparison.length, statement->branch.comparison.start);
-			print_expr(statement->branch.right);
-			printf("\n");
-			print_statement(statement->branch.body, "    ");
-			break;
-		case STATEMENT_CALL:
-			printf("%s%.*s(", indent, (int)statement->call.name.length, statement->call.name.start);
-			for (size_t i = 0; i < statement->call.arg_count; i += 1)
-			{
-				if (i > 0)
-					printf(", ");
-				print_expr(statement->call.args[i]);
-			}
-			printf(")\n");
-			break;
-		case STATEMENT_STACK:
-			printf("%sstack %.*s[%.*s]\n", indent,
-				(int)statement->stack.name.length, statement->stack.name.start,
-				(int)statement->stack.size.length, statement->stack.size.start);
-			break;
-	}
-}
-
 int main(int argc, char** argv)
 {
 	struct Args args;
@@ -85,6 +16,12 @@ int main(int argc, char** argv)
 	if (result == PARSE_ERROR)
 		return 1;
 
+	if (args.target != ASSEMBLER_NASM)
+	{
+		fprintf(stderr, "error: only the nasm target is supported\n");
+		return 1;
+	}
+
 	struct File source;
 	if (!read_file(args.input_path, &source))
 		return 1;
@@ -99,39 +36,24 @@ int main(int argc, char** argv)
 		return 1;
 	}
 
-	for (size_t i = 0; i < program.const_count; i += 1)
-	{
-		struct ConstDecl decl = program.consts[i];
-		printf("const %.*s = %.*s\n",
-			(int)decl.name.length, decl.name.start,
-			(int)decl.value.length, decl.value.start);
-	}
-
-	for (size_t i = 0; i < program.data_count; i += 1)
+	FILE* out = stdout;
+	if (args.output_path != NULL)
 	{
-		struct DataDecl decl = program.data_decls[i];
-		printf("data %.*s = %.*s\n",
-			(int)decl.name.length, decl.name.start,
-			(int)decl.value.length, decl.value.start);
-	}
-
-	for (size_t i = 0; i < program.proc_count; i += 1)
-	{
-		struct ProcDecl proc = program.procs[i];
-		printf("proc %.*s(", (int)proc.name.length, proc.name.start);
-		for (size_t p = 0; p < proc.param_count; p += 1)
+		out = fopen(args.output_path, "w");
+		if (out == NULL)
 		{
-			struct Param param = proc.params[p];
-			printf("%s%.*s: %.*s", p == 0 ? "" : ", ",
-				(int)param.name.length, param.name.start,
-				(int)param.reg.length, param.reg.start);
+			fprintf(stderr, "error: could not open '%s' for writing\n", args.output_path);
+			free_program(&program);
+			free_file(&source);
+			return 1;
 		}
-		printf(")\n");
-
-		for (size_t s = 0; s < proc.body_count; s += 1)
-			print_statement(&proc.body[s], "  ");
 	}
 
+	generate_nasm(&program, out);
+
+	if (out != stdout)
+		fclose(out);
+
 	free_program(&program);
 	free_file(&source);
 }
diff --git a/tests/codegen_test.c b/tests/codegen_test.c
new file mode 100644
index 0000000..64421ea
--- /dev/null
+++ b/tests/codegen_test.c
@@ -0,0 +1,37 @@
+#include <stdio.h>
+#include <string.h>
+
+#include "parser/parser.h"
+#include "codegen/nasm.h"
+#include "tests.h"
+
+static void test_generate_consts_and_data(struct TestContext* context)
+{
+	struct Lexer lexer = create_lexer("const N = 5\ndata msg = \"hi\"\n");
+	struct Program program;
+	check(context, parse_program(&lexer, &program));
+
+	FILE* out = tmpfile();
+	generate_nasm(&program, out);
+	fflush(out);
+	rewind(out);
+
+	char buffer[1024];
+	size_t read = fread(buffer, 1, sizeof(buffer) - 1, out);
+	buffer[read] = '\0';
+	fclose(out);
+
+	check(context, strstr(buffer, "N equ 5") != NULL);
+	check(context, strstr(buffer, "section .data") != NULL);
+	check(context, strstr(buffer, "msg: db `hi`") != NULL);
+	check(context, strstr(buffer, ".len equ $ - msg") != NULL);
+	check(context, strstr(buffer, "section .text") != NULL);
+	check(context, strstr(buffer, "global _start") != NULL);
+
+	free_program(&program);
+}
+
+void run_codegen_tests(struct TestContext* context)
+{
+	test_generate_consts_and_data(context);
+}
diff --git a/tests/main.c b/tests/main.c
index b69bb3e..f0caf52 100644
--- a/tests/main.c
+++ b/tests/main.c
@@ -8,6 +8,7 @@ int main(void)
 
 	run_lexer_tests(&context);
 	run_parser_tests(&context);
+	run_codegen_tests(&context);
 
 	printf("\n%d checks, %d failure(s)\n", context.checks, context.failures);
 	return context.failures == 0 ? 0 : 1;
diff --git a/tests/tests.h b/tests/tests.h
index 9136910..32ee4a6 100644
--- a/tests/tests.h
+++ b/tests/tests.h
@@ -4,3 +4,4 @@
 
 void run_lexer_tests(struct TestContext* context);
 void run_parser_tests(struct TestContext* context);
+void run_codegen_tests(struct TestContext* context);