feat: add sematic analysis

3 files changed, 88 insertions, 0 deletions

diff --git a/src/main.c b/src/main.c
index a908a73..90b3c82 100644
--- a/src/main.c
+++ b/src/main.c
@@ -2,6 +2,7 @@
 
 #include "io/file.h"
 #include "cli/args.h"
+#include "sema/sema.h"
 #include "lexer/lexer.h"
 #include "codegen/nasm.h"
 #include "parser/parser.h"
@@ -36,6 +37,13 @@ int main(int argc, char** argv)
 		return 1;
 	}
 
+	if (!analyze_program(&program))
+	{
+		free_program(&program);
+		free_file(&source);
+		return 1;
+	}
+
 	FILE* out = stdout;
 	if (args.output_path != NULL)
 	{
diff --git a/src/sema/sema.c b/src/sema/sema.c
new file mode 100644
index 0000000..99ec4f1
--- /dev/null
+++ b/src/sema/sema.c
@@ -0,0 +1,73 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "sema/sema.h"
+
+static bool names_equal(struct Token a, struct Token b)
+{
+	return a.length == b.length && memcmp(a.start, b.start, a.length) == 0;
+}
+
+static bool check_duplicate_names(struct Program* program)
+{
+	size_t count = program->const_count + program->data_count + program->proc_count;
+	if (count == 0)
+		return true;
+
+	struct Token* names = malloc(count * sizeof(struct Token));
+	size_t n = 0;
+	for (size_t i = 0; i < program->const_count; i += 1)
+	{
+		names[n] = program->consts[i].name;
+		n += 1;
+	}
+	for (size_t i = 0; i < program->data_count; i += 1)
+	{
+		names[n] = program->data_decls[i].name;
+		n += 1;
+	}
+	for (size_t i = 0; i < program->proc_count; i += 1)
+	{
+		names[n] = program->procs[i].name;
+		n += 1;
+	}
+
+	bool ok = true;
+	for (size_t i = 0; i < count; i += 1)
+		for (size_t j = 0; j < i; j += 1)
+			if (names_equal(names[i], names[j]))
+			{
+				fprintf(stderr, "error: line %u: '%.*s' is already defined\n",
+					names[i].line, (int)names[i].length, names[i].start);
+				ok = false;
+			}
+
+	free(names);
+	return ok;
+}
+
+static bool check_entry_point(struct Program* program)
+{
+	for (size_t i = 0; i < program->proc_count; i += 1)
+	{
+		struct Token name = program->procs[i].name;
+		if (name.length == 4 && memcmp(name.start, "main", 4) == 0)
+			return true;
+	}
+
+	fprintf(stderr, "error: no 'main' procedure defined\n");
+	return false;
+}
+
+bool analyze_program(struct Program* program)
+{
+	bool ok = true;
+
+	if (!check_duplicate_names(program))
+		ok = false;
+	if (!check_entry_point(program))
+		ok = false;
+
+	return ok;
+}
diff --git a/src/sema/sema.h b/src/sema/sema.h
new file mode 100644
index 0000000..e015f41
--- /dev/null
+++ b/src/sema/sema.h
@@ -0,0 +1,7 @@
+#pragma once
+
+#include <stdbool.h>
+
+#include "parser/ast.h"
+
+bool analyze_program(struct Program* program);