Hi Kalle,
some comments to your patch...
+/*
+�*�Handle�delayed�reply�to�HTTP�sendsms�client,�if�any
+�*/
+static�void�delayed_http_reply(Msg�*msg)
+{
...
+����uuid_unparse(msg->ack.id,�id);
+����os�=�octstr_create(id);
+����debug("sms.http",�0,�"Got�ACK�(%d)�of�%s",�msg->ack.nack,�octstr_get_cstr(os));
+����client�=�dict_remove(client_dict,�os);
+����if�(client�==�NULL)�{
+��������debug("sms.http",�0,�"No�client�-�multi-send�or�ACK�to�pull-reply");
+��������octstr_destroy(os);
+��������return;
+����}
...
Destroy 'os' immediately after dict_remove instead of in if{} and later at
the end of function, it's not used later... then it's easier to undestand
that it not used anymore.
+static�void�store_uuid(Msg�*msg,�Octstr�**stored_uuid)
+{
...
How about 'static Octstr *store_uuid(Msg *msg)'? I don't see any sense to
put Octstr** as function params...
...
+����dict_destroy(client_dict);�/*�XXX�is�this�right�location?�*/
+����http_destroy_headers(sendsms_reply_hdrs);�/*�ditto�*/
...
why not test if it's a right location todo it instead of writing such
comment? IMO such comment is useless.
Otherwise patch looks good although not tested yet.
Kalle Marjola wrote:
> (patch attached)
>
> Why am I using this FC2 when all it causes is problems..
>
>
--
Thanks,
Alex
