From 6ca377762516888b0488d60c971e0660ae6035f4 Mon Sep 17 00:00:00 2001
From: Camil Staps
Date: Fri, 14 Oct 2016 22:17:01 +0200
Subject: token_list using an array for memory efficiency

---
 interpreter/lex.c | 92 ++++++++++++++++++++++++++++---------------------------
 1 file changed, 47 insertions(+), 45 deletions(-)

(limited to 'interpreter/lex.c')

diff --git a/interpreter/lex.c b/interpreter/lex.c
index 5032bfb..497ee8e 100644
--- a/interpreter/lex.c
+++ b/interpreter/lex.c
@@ -34,87 +34,89 @@ unsigned char lex_name_length(char *input) {
 }
 
 struct token_list *lex(struct token_list *list, char *input) {
-	struct token_list *first_list;
 	bool create_new_token;
 
 	while (*input && is_space_char(*input)) input++;
-	if (*input == 0) {
+	if (*input == 0)
 		return list;
-	}
 	
-	if (list) {
-		first_list = list;
-		while (list->rest) list = list->rest;
-		list->rest = my_calloc(1, sizeof(struct token_list));
-		list = list->rest;
-	} else {
-		first_list = list = my_calloc(1, sizeof(struct token_list));
+	if (!list) {
+		list = my_calloc(1,
+				sizeof(struct token_list) +
+				INITIAL_TOKEN_LIST_SIZE * sizeof(struct token) + 1);
+		list->length = INITIAL_TOKEN_LIST_SIZE;
+		list->index = 0;
 	}
 
-	create_new_token = 1;
+	create_new_token = true;
 
 	while (*input) {
-		list->elem.var = NULL;
+		if (list->index >= list->length) {
+			list = my_realloc(list,
+					sizeof(struct token_list) +
+					2 * list->length * sizeof(struct token) + 1);
+			list->length *= 2;
+		}
+
+		list->elems[list->index].var = NULL;
 
 		switch (*input) {
-			case ';': list->elem.kind = TOKEN_SEMICOLON; break;
-			case ':': list->elem.kind = TOKEN_COLON;     break;
-			case '(': list->elem.kind = TOKEN_OPEN_P;    break;
-			case ')': list->elem.kind = TOKEN_CLOSE_P;   break;
-			case '[': list->elem.kind = TOKEN_OPEN_SQ;   break;
-			case ']': list->elem.kind = TOKEN_CLOSE_SQ;  break;
-			case '=': list->elem.kind = TOKEN_EQUALS;    break;
-			case ',': list->elem.kind = TOKEN_COMMA;     break;
+			case ';': list->elems[list->index].kind = TOKEN_SEMICOLON; break;
+			case ':': list->elems[list->index].kind = TOKEN_COLON;     break;
+			case '(': list->elems[list->index].kind = TOKEN_OPEN_P;    break;
+			case ')': list->elems[list->index].kind = TOKEN_CLOSE_P;   break;
+			case '[': list->elems[list->index].kind = TOKEN_OPEN_SQ;   break;
+			case ']': list->elems[list->index].kind = TOKEN_CLOSE_SQ;  break;
+			case '=': list->elems[list->index].kind = TOKEN_EQUALS;    break;
+			case ',': list->elems[list->index].kind = TOKEN_COMMA;     break;
 			default:
 				if (input[0] == '/' && input[1] == '/') {
 					while (input && input[0] != '\n') input++;
+					create_new_token = false;
 					break;
-				}
-				if (input[0] == 'c' && input[1] == 'o' && input[2] == 'd' &&
-						input[3] == 'e' && is_space_char(input[4])) {
-					list->elem.kind = TOKEN_CODE;
+				} else if (input[0] == 'c' && input[1] == 'o' &&
+						input[2] == 'd' && input[3] == 'e' &&
+						is_space_char(input[4])) {
+					list->elems[list->index].kind = TOKEN_CODE;
 					input += 4;
 					break;
-				}
-				if (input[0] == 'i' && input[1] == 'm' && input[2] == 'p' &&
-						input[3] == 'o' && input[4] == 'r' &&
-						input[5] == 't' && is_space_char(input[6])) {
-					list->elem.kind = TOKEN_IMPORT;
+				} else if (input[0] == 'i' && input[1] == 'm' &&
+						input[2] == 'p' && input[3] == 'o' &&
+						input[4] == 'r' && input[5] == 't' &&
+						is_space_char(input[6])) {
+					list->elems[list->index].kind = TOKEN_IMPORT;
 					input += 6;
 					break;
-				}
-				if (is_int_char(*input)) {
+				} else if (is_int_char(*input)) {
 					char *s;
 					unsigned char len = lex_int_length(input);
 					s = my_calloc(1, len + 1);
-					list->elem.kind = TOKEN_INT;
-					list->elem.var = my_calloc(1, sizeof(int));
+					list->elems[list->index].kind = TOKEN_INT;
+					list->elems[list->index].var = my_calloc(1, sizeof(int));
 					strncpy(s, input, len);
-					*((int*) list->elem.var) = atoi(s);
+					*((int*) list->elems[list->index].var) = atoi(s);
 					my_free(s);
 					input += len - 1;
 				} else if (is_name_char(*input)) {
 					unsigned char len = lex_name_length(input);
-					list->elem.kind = TOKEN_NAME;
-					list->elem.var = my_calloc(1, len + 1);
-					strncpy(list->elem.var, input, len);
+					list->elems[list->index].kind = TOKEN_NAME;
+					list->elems[list->index].var = my_calloc(1, len + 1);
+					strncpy(list->elems[list->index].var, input, len);
 					input += len - 1;
 				} else if (is_space_char(*input)) {
-					create_new_token = 0;
+					create_new_token = false;
 				} else {
-					free_token_list(first_list);
-					my_free(first_list);
+					free_token_list(list);
+					my_free(list);
 					return NULL;
 				}
 		}
 
 		do input++; while (*input && is_space_char(*input));
 
-		if (*input && create_new_token) {
-			list->rest = my_calloc(1, sizeof(struct token_list));
-			list = list->rest;
-		}
+		if (create_new_token)
+			list->index++;
 	}
 
-	return first_list;
+	return list;
 }
-- 
cgit v1.2.3