Start using string_view on lexer and parser

This change fixes the memory leak when token got created.

Signed-off-by: Johnny Richard <johnny@johnnyrichard.com>

5 files changed, 23 insertions, 26 deletions

diff --git a/src/lexer.c b/src/lexer.c
index f58b0ea..9be96a8 100644
--- a/src/lexer.c
+++ b/src/lexer.c
@@ -55,8 +55,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
       lexer_drop_char(lexer);
     }
     token->kind = TOKEN_NUMBER;
-    token->value = (char *) malloc(sizeof(char) * (lexer->cur - begin + 1));
-    strncpy(token->value, lexer->src + begin, lexer->cur - begin);
+    token->value = string_view_new(lexer->src + begin, lexer->cur - begin);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = begin - lexer->bol;
@@ -70,8 +69,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
       lexer_drop_char(lexer);
     }
     token->kind = TOKEN_NAME;
-    token->value = (char *) malloc(sizeof(char) * (lexer->cur - begin + 1));
-    strncpy(token->value, lexer->src + begin, lexer->cur - begin);
+    token->value = string_view_new(lexer->src + begin, lexer->cur - begin);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = begin - lexer->bol;
@@ -80,8 +78,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
 
   if (lexer_is_not_eof(lexer) && lexer_current_char(lexer) == '(') {
     token->kind = TOKEN_OPAREN;
-    token->value = (char *) malloc(sizeof(char) + 1);
-    strcpy(token->value, "(");
+    token->value = string_view_new(lexer->src, 1);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = lexer->cur - lexer->bol;
@@ -91,8 +88,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
 
   if (lexer_is_not_eof(lexer) && lexer_current_char(lexer) == ')') {
     token->kind = TOKEN_CPAREN;
-    token->value = (char *) malloc(sizeof(char) + 1);
-    strcpy(token->value, ")");
+    token->value = string_view_new(lexer->src, 1);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = lexer->cur - lexer->bol;
@@ -102,8 +98,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
 
   if (lexer_is_not_eof(lexer) && lexer_current_char(lexer) == ':') {
     token->kind = TOKEN_COLON;
-    token->value = (char *) malloc(sizeof(char) + 1);
-    strcpy(token->value, ":");
+    token->value = string_view_new(lexer->src, 1);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = lexer->cur - lexer->bol;
@@ -113,8 +108,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
 
   if (lexer_is_not_eof(lexer) && lexer_current_char(lexer) == ';') {
     token->kind = TOKEN_SEMICOLON;
-    token->value = (char *) malloc(sizeof(char) + 1);
-    strcpy(token->value, ";");
+    token->value = string_view_new(lexer->src, 1);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = lexer->cur - lexer->bol;
@@ -124,8 +118,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
 
   if (lexer_is_not_eof(lexer) && lexer_current_char(lexer) == '{') {
     token->kind = TOKEN_OCURLY;
-    token->value = (char *) malloc(sizeof(char) + 1);
-    strcpy(token->value, "{");
+    token->value = string_view_new(lexer->src, 1);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = lexer->cur - lexer->bol;
@@ -135,8 +128,7 @@ lexer_next_token(lexer_t *lexer, token_t *token)
 
   if (lexer_is_not_eof(lexer) && lexer_current_char(lexer) == '}') {
     token->kind = TOKEN_CCURLY;
-    token->value = (char *) malloc(sizeof(char) + 1);
-    strcpy(token->value, "}");
+    token->value = string_view_new(lexer->src, 1);
     token->filepath = lexer->filepath;
     token->row = lexer->row;
     token->col = lexer->cur - lexer->bol;
diff --git a/src/lexer.h b/src/lexer.h
index 472ee10..a091fb7 100644
--- a/src/lexer.h
+++ b/src/lexer.h
@@ -20,6 +20,7 @@
 #include <stdint.h>
 #include <stdlib.h>
 #include <stdbool.h>
+#include "string_view.h"
 
 typedef enum {
     TOKEN_NAME,
@@ -35,7 +36,7 @@ typedef enum {
 
 typedef struct token_t {
   token_kind_t  kind;
-  char         *value;
+  string_view_t value;
   char         *filepath;
   uint32_t      row;
   uint32_t      col;
diff --git a/src/parser.c b/src/parser.c
index 4867201..593f75a 100644
--- a/src/parser.c
+++ b/src/parser.c
@@ -65,11 +65,11 @@ parser_parse_type(parser_t *parser)
 {
   token_t token = expected_token(parser, TOKEN_NAME);
 
-  if (strcmp(token.value, "i32") == 0) {
+  if (string_view_eq(token.value, string_view_from_str("i32"))) {
     return TYPE_I32;
   }
 
-  fprintf(stderr, "[ERROR]: expected type 'i32' but got '%s'\n", token.value); 
+  fprintf(stderr, "[ERROR]: expected type 'i32' but got '"SVFMT"'\n", SVARG(&token.value)); 
   exit(EXIT_FAILURE);
 }
 
@@ -79,9 +79,9 @@ parser_parse_return_stmt(parser_t *parser)
   expected_token(parser, TOKEN_OCURLY);
   token_t return_keyword_token = expected_token(parser, TOKEN_NAME);
 
-  if (strcmp(return_keyword_token.value, "return") != 0) {
+  if (!string_view_eq(return_keyword_token.value, string_view_from_str("return"))) {
     // TODO: Add filename:row:col prefix to expected token exceptions
-    fprintf(stderr, "[ERROR]: expected 'return' keyword but got '%s'\n", return_keyword_token.value); 
+    fprintf(stderr, "[ERROR]: expected 'return' keyword but got '"SVFMT"'\n", SVARG(&return_keyword_token.value));
     exit(EXIT_FAILURE);
   }
 
@@ -89,8 +89,11 @@ parser_parse_return_stmt(parser_t *parser)
   expected_token(parser, TOKEN_SEMICOLON);
   expected_token(parser, TOKEN_CCURLY);
 
+  char number_as_str[number_token.value.size];
+  string_view_to_str(&number_token.value, number_as_str);
+
   return (ast_return_stmt_t) {
-    .number = atoi(number_token.value)
+    .number = atoi(number_as_str)
   };
 }
 
@@ -104,7 +107,6 @@ parser_parse_function(parser_t *parser)
   type_t return_type = parser_parse_type(parser);
 
   return (ast_function_t) {
-    // FIXME: alloc memory for the ast_function.name, it's not a good idea to use the token.value
     .name = func_name_token.value,
     .return_type = return_type,
     .body = parser_parse_return_stmt(parser)
diff --git a/src/parser.h b/src/parser.h
index e9a1cb7..d44a4df 100644
--- a/src/parser.h
+++ b/src/parser.h
@@ -18,6 +18,7 @@
 #define PARSER_H
 
 #include "lexer.h"
+#include "string_view.h"
 
 typedef struct parser_t {
   lexer_t *lexer;
@@ -32,7 +33,7 @@ typedef struct ast_return_stmt_t {
 } ast_return_stmt_t;
 
 typedef struct ast_function_t {
-  char   *name;
+  string_view_t name;
   type_t  return_type;
   ast_return_stmt_t body;
 } ast_function_t;
diff --git a/src/pipac.c b/src/pipac.c
index 993d56d..6e6df9e 100644
--- a/src/pipac.c
+++ b/src/pipac.c
@@ -20,11 +20,12 @@
 
 #include "lexer.h"
 #include "parser.h"
+#include "string_view.h"
 
 void
 generate_gas_x86_64_linux(ast_function_t *func)
 {
-  if (strcmp(func->name, "main") != 0) {
+  if (!string_view_eq(func->name, string_view_from_str("main"))) {
     fprintf(stderr, "[ERROR]: no main function has been defined!\n");
     exit(EXIT_FAILURE);
   }
@@ -47,7 +48,7 @@ void
 print_tokens(lexer_t *lexer) {
   token_t token;
   for (lexer_next_token(lexer, &token); token.kind != TOKEN_EOF; lexer_next_token(lexer, &token)) {
-    printf("%s:%d:%d: [kind=%d, value='%s']\n", lexer->filepath, token.row + 1, token.col + 1, token.kind, token.value);
+    printf("%s:%d:%d: [kind=%d, value='"SVFMT"']\n", lexer->filepath, token.row + 1, token.col + 1, token.kind, SVARG(&token.value));
   }
 }