that's kinda cool

1 files changed, 109 insertions, 21 deletions

diff --git a/src/main.c b/src/main.c
index 1bcb71c..666f847 100644
--- a/src/main.c
+++ b/src/main.c
@@ -1,6 +1,24 @@
+#include <ctype.h>
 #include <stdarg.h>
+#include <stdbool.h>
 #include <stdio.h>
 #include <stdlib.h>
+#include <string.h>
+
+typedef enum {
+	TOK_PUNCT,
+	TOK_NUM,
+	TOK_EOF,
+} TokenType;
+
+typedef struct Token Token;
+struct Token {
+	TokenType type;
+	Token *next;
+	int val;
+	char *loc;
+	int len;
+};
 
 static void
 error(const char *fmt, ...) {
@@ -13,38 +31,108 @@ error(const char *fmt, ...) {
 	exit(1);
 }
 
+static void
+expect(const Token *tok, TokenType type)
+{
+	if (tok->type != type)
+		error("expected a %u", type);
+}
+
+static bool
+equal(const Token *tok, const char *op)
+{
+	return memcmp(tok->loc, op, tok->len) == 0 && op[tok->len] == '\0';
+}
+
+static Token *
+skip(Token *tok, char *s)
+{
+	if (!equal(tok, s))
+		error("expected '%s'", s);
+	return tok->next;
+}
+
+static int
+get_number(Token *tok)
+{
+	expect(tok, TOK_NUM);
+	return tok->val;
+}
+
+static Token *
+new_token(TokenType type, char *start, char *end)
+{
+	Token *tok = calloc(1, sizeof(Token));
+	if (tok == NULL)
+		error("calloc failed");
+	tok->type = type;
+	tok->loc = start;
+	tok->len = end - start;
+	return tok;
+}
+
+static Token *
+tokenize(char *p)
+{
+	Token head = {0};
+	Token *cur = &head;
+
+	while (*p != '\0') {
+		if (isspace(*p)) {
+			p++;
+			continue;
+		}
+
+		if (isdigit(*p)) {
+			char *q = p;
+			const unsigned long val = strtoul(p, &p, 10);
+			cur = cur->next = new_token(TOK_NUM, q, p);
+			cur->val = val;
+			continue;
+		}
+
+		if (strchr("+-/*", *p) != NULL) {
+			cur = cur->next = new_token(TOK_PUNCT, p, p + 1);
+			p += 1;
+			continue;
+		}
+
+		error("invalid token");
+	}
+	cur->next = new_token(TOK_EOF, p, p);
+
+	return head.next;
+}
+
 int
 main(int argc, char **argv)
 {
 	if (argc != 2)
 		error("usage: %s code", argv[0]);
 
-	char *p = argv[1];
+	Token *tok = tokenize(argv[1]);
 
+	/* first token is a number */
 	printf("main:\n");
-	printf("\tLIT %04lu\n", strtoul(p, &p, 10));
+	printf("\tLIT %04d\n", get_number(tok));
+	tok = tok->next;
 
-	while (*p != '\0') {
-		const char *op;
-		switch (*p) {
-		case '+':
-			op = "ADD";
-			break;
-		case '-':
-			op = "SUB";
-			break;
-		case '*':
-			op = "MUL";
-			break;
-		case '/':
-			op = "DIV";
-			break;
-		default:
-			error("unexpected character: '%c'", *p);
+	/* followed by punct then another number */
+	while (tok->type != TOK_EOF) {
+		const char *punct;
+		expect(tok, TOK_PUNCT);
+		switch (tok->loc[0]) {
+		case '+': punct = "ADD"; break;
+		case '-': punct = "SUB"; break;
+		case '*': punct = "MUL"; break;
+		case '/': punct = "DIV"; break;
 		}
-		p++;
-		printf("\tLIT %04lu %s\n", strtoul(p, &p, 10), op);
+		tok = tok->next;
+
+		printf("\tLIT %04d %s\n", get_number(tok), punct);
+		tok = tok->next;
 	}
+	expect(tok, TOK_EOF);
 
 	printf("\tRET\n");