cJSON_Parse{,WithOpts}: Skip UTF-8 (Byte Order Marks)

cJSON.c

@@ -958,6 +958,22 @@ static parse_buffer *buffer_skip_whitespace(parse_buffer * const buffer)
     return buffer;
 }
 
+/* skip the UTF-8 BOM (byte order mark) if it is at the beginning of a buffer */
+static parse_buffer *skip_utf8_bom(parse_buffer * const buffer)
+{
+    if ((buffer == NULL) || (buffer->content == NULL) || (buffer->offset != 0))
+    {
+        return NULL;
+    }
+
+    if (can_access_at_index(buffer, 4) && (strncmp((const char*)buffer_at_offset(buffer), "\xEF\xBB\xBF", 3) == 0))
+    {
+        buffer->offset += 3;
+    }
+
+    return buffer;
+}
+
 /* Parse an object - create a new root, and populate. */
 CJSON_PUBLIC(cJSON *) cJSON_ParseWithOpts(const char *value, const char **return_parse_end, cJSON_bool require_null_terminated)
 {
@@ -984,7 +1000,7 @@ CJSON_PUBLIC(cJSON *) cJSON_ParseWithOpts(const char *value, const char **return
         goto fail;
     }
 
-    if (!parse_value(item, buffer_skip_whitespace(&buffer)))
+    if (!parse_value(item, buffer_skip_whitespace(skip_utf8_bom(&buffer))))
     {
         /* parse failure. ep is set. */
         goto fail;
@@ -1222,7 +1238,6 @@ static cJSON_bool parse_value(cJSON * const item, parse_buffer * const input_buf
         return parse_object(item, input_buffer);
     }
 
-
     return false;
 }
 

tests/misc_tests.c

@@ -410,6 +410,30 @@ static void cjson_functions_shouldnt_crash_with_null_pointers(void)
     cJSON_Delete(item);
 }
 
+static void skip_utf8_bom_should_skip_bom(void)
+{
+    const unsigned char string[] = "\xEF\xBB\xBF{}";
+    parse_buffer buffer = { 0, 0, 0, 0, { 0, 0, 0 } };
+    buffer.content = string;
+    buffer.length = sizeof(string);
+    buffer.hooks = global_hooks;
+
+    TEST_ASSERT_TRUE(skip_utf8_bom(&buffer) == &buffer);
+    TEST_ASSERT_EQUAL_UINT(3U, (unsigned int)buffer.offset);
+}
+
+static void skip_utf8_bom_should_not_skip_bom_if_not_at_beginning(void)
+{
+    const unsigned char string[] = " \xEF\xBB\xBF{}";
+    parse_buffer buffer = { 0, 0, 0, 0, { 0, 0, 0 } };
+    buffer.content = string;
+    buffer.length = sizeof(string);
+    buffer.hooks = global_hooks;
+    buffer.offset = 1;
+
+    TEST_ASSERT_NULL(skip_utf8_bom(&buffer));
+}
+
 int main(void)
 {
     UNITY_BEGIN();
@@ -425,6 +449,8 @@ int main(void)
     RUN_TEST(cjson_replace_item_via_pointer_should_replace_items);
     RUN_TEST(cjson_replace_item_in_object_should_preserve_name);
     RUN_TEST(cjson_functions_shouldnt_crash_with_null_pointers);
+    RUN_TEST(skip_utf8_bom_should_skip_bom);
+    RUN_TEST(skip_utf8_bom_should_not_skip_bom_if_not_at_beginning);
 
     return UNITY_END();
 }

tests/parse_with_opts.c

@@ -69,6 +69,22 @@ static void parse_with_opts_should_return_parse_end(void)
     cJSON_Delete(item);
 }
 
+static void parse_with_opts_should_parse_utf8_bom(void)
+{
+    cJSON *with_bom = NULL;
+    cJSON *without_bom = NULL;
+
+    with_bom = cJSON_ParseWithOpts("\xEF\xBB\xBF{}", NULL, true);
+    TEST_ASSERT_NOT_NULL(with_bom);
+    without_bom = cJSON_ParseWithOpts("{}", NULL, true);
+    TEST_ASSERT_NOT_NULL(with_bom);
+
+    TEST_ASSERT_TRUE(cJSON_Compare(with_bom, without_bom, true));
+
+    cJSON_Delete(with_bom);
+    cJSON_Delete(without_bom);
+}
+
 int main(void)
 {
     UNITY_BEGIN();
@@ -77,6 +93,7 @@ int main(void)
     RUN_TEST(parse_with_opts_should_handle_empty_strings);
     RUN_TEST(parse_with_opts_should_require_null_if_requested);
     RUN_TEST(parse_with_opts_should_return_parse_end);
+    RUN_TEST(parse_with_opts_should_parse_utf8_bom);
 
     return UNITY_END();
 }