On 23.10.2012 12:36, Patrik Flykt wrote:
Hi,
On Tue, 2012-10-23 at 10:42 +0200, Daniel Wagner wrote:
+static int parse_context(DBusMessage *msg, unsigned char **context,
+ unsigned int *size)
+{
+ DBusMessageIter iter, array;
+ unsigned char *ctx;
+ int i = 0;
+
+ dbus_message_iter_init(msg, &iter);
+ dbus_message_iter_recurse(&iter, &array);
+ while (dbus_message_iter_get_arg_type(&array) == DBUS_TYPE_BYTE) {
+ i++;
+
+ dbus_message_iter_next(&array);
+ }
+
+ *size = i;
+ if (i == 0)
+ return 0;
+
+ ctx = g_try_malloc((gsize) *size);
+ if (ctx == NULL)
+ return -ENOMEM;
+
+ i = 0;
+ dbus_message_iter_init(msg, &iter);
+ dbus_message_iter_recurse(&iter, &array);
+ while (dbus_message_iter_get_arg_type(&array) == DBUS_TYPE_BYTE) {
+ dbus_message_iter_get_basic(&array, &ctx[i]);
+
So it's an array of bytes including a NULL byte as it seems.
Ok.
+ i++;
+ dbus_message_iter_next(&array);
+ }
+
+ *context = ctx;
+
+ return 0;
+}
+
Perhaps the parsed char ** could be returned here?
What do you mean with 'parsed char **'? parse_context(ctx)?
Are we ever going to be interested in the return value of
parse_context() as it is always zero, or could we just 'return ctx;'
instead?
If the only error is ENOMEM then we could just return the allocated
buffer. The callee can then figure if ENOMEM or valid happened.
_______________________________________________
connman mailing list
connman@connman.net
http://lists.connman.net/listinfo/connman
