Better strategy for restoring multicast address

This commit is contained in:
Sergey Lyubka 2022-01-07 18:40:48 +00:00
parent 6725d61b3f
commit 97f95e17db

View File

@ -6,17 +6,25 @@
static const char *s_ssdp_url = "udp://239.255.255.250:1900";
static void fn(struct mg_connection *c, int ev, void *ev_data, void *fn_data) {
if (ev == MG_EV_OPEN) c->is_hexdumping = 1;
LOG(LL_DEBUG, ("%p got event: %d %p %p", c, ev, ev_data, fn_data));
if (ev == MG_EV_OPEN) {
c->is_hexdumping = 1;
} else if (ev == MG_EV_RESOLVE) {
// c->peer gets populated with multicast address. Store it in c->label
memcpy(c->label, &c->peer, sizeof(c->peer));
} else if (ev == MG_EV_READ) {
// Each response to the SSDP socket will change c->peer.
// We can now do mg_printf(c, "haha"); to respond back to the remote side.
// But in our case, we should restore the multicast address in order
// to have next search to go to the multicast address
memcpy(&c->peer, c->label, sizeof(c->peer));
}
}
static void tfn(void *param) {
struct mg_connection *c = param;
if (c == NULL) return;
LOG(LL_INFO, ("Sending M-SEARCH"));
// Each response to the SSDP socket will change c->peer, so re-initialise
// c->peer to the multicast address before each search request
mg_resolve(c, s_ssdp_url);
mg_printf(c, "%s",
"M-SEARCH * HTTP/1.1\r\n"
"HOST: 239.255.255.250:1900\r\n"