Rx3/drmr2
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Move to using atom instead of event. So far just pass midi via atom.

Browse Source
This commit is contained in:
Nick Lanham
2012-03-22 14:15:00 +01:00
parent 1a624dd489
commit 990caaf510
4 changed files with 63 additions and 57 deletions
Download Patch File Download Diff File Expand all files Collapse all files

80
drmr.c
View File

@@ -81,17 +81,16 @@ instantiate(const LV2_Descriptor* descriptor,
// Map midi uri
while(*features) {
if (!strcmp((*features)->URI, LV2_URI_MAP_URI)) {
drmr->map = (LV2_URI_Map_Feature *)((*features)->data);
drmr->uris.midi_event = drmr->map->uri_to_id
(drmr->map->callback_data,
"http://lv2plug.in/ns/ext/event",
"http://lv2plug.in/ns/ext/midi#MidiEvent");
if (!strcmp((*features)->URI, LV2_URID_URI "#map")) {
drmr->map = (LV2_URID_Map *)((*features)->data);
drmr->uris.midi_event =
drmr->map->map(drmr->map->handle,
"http://lv2plug.in/ns/ext/midi#MidiEvent");
}
features++;
}
if (!drmr->map) {
fprintf(stderr, "LV2 host does not support uri-map.\n");
fprintf(stderr, "LV2 host does not support urid#map.\n");
free(drmr);
return 0;
}
@@ -126,8 +125,8 @@ connect_port(LV2_Handle instance,
DrMr* drmr = (DrMr*)instance;
DrMrPortIndex port_index = (DrMrPortIndex)port;
switch (port_index) {
case DRMR_MIDI:
drmr->midi_port = (LV2_Event_Buffer*)data;
case DRMR_CONTROL:
drmr->control_port = (LV2_Atom_Sequence*)data;
break;
case DRMR_LEFT:
drmr->left = (float*)data;
@@ -191,42 +190,37 @@ static void run(LV2_Handle instance, uint32_t n_samples) {
if (kitInt != drmr->curKit) // requested a new kit
pthread_cond_signal(&drmr->load_cond);
LV2_Event_Iterator eit;
if (drmr->midi_port && lv2_event_begin(&eit,drmr->midi_port)) { // if we have any events
LV2_Event *cur_ev;
uint8_t* data;
while (lv2_event_is_valid(&eit)) {
cur_ev = lv2_event_get(&eit,&data);
if (cur_ev->type == drmr->uris.midi_event) {
//int channel = *data & 15;
switch ((*data) >> 4) {
case 8: // ignore note-offs for now, should probably be a setting
//if (drmr->cur_samp) drmr->cur_samp->active = 0;
break;
case 9: {
uint8_t nn = data[1];
nn-=baseNote;
// need to mutex this to avoid getting the samples array
// changed after the check that the midi-note is valid
pthread_mutex_lock(&drmr->load_mutex);
if (nn >= 0 && nn < drmr->num_samples) {
if (drmr->samples[nn].layer_count > 0) {
layer_to_sample(drmr->samples+nn,*(drmr->gains[nn]));
if (drmr->samples[nn].limit == 0)
fprintf(stderr,"Failed to find layer at: %i for %f\n",nn,*drmr->gains[nn]);
}
drmr->samples[nn].active = 1;
drmr->samples[nn].offset = 0;
LV2_SEQUENCE_FOREACH(drmr->control_port, i) {
LV2_Atom_Event* const ev = lv2_sequence_iter_get(i);
if (ev->body.type == drmr->uris.midi_event) {
uint8_t* const data = (uint8_t* const)(ev + 1);
//int channel = *data & 15;
switch ((*data) >> 4) {
case 8: // ignore note-offs for now, should probably be a setting
//if (drmr->cur_samp) drmr->cur_samp->active = 0;
break;
case 9: {
uint8_t nn = data[1];
nn-=baseNote;
// need to mutex this to avoid getting the samples array
// changed after the check that the midi-note is valid
pthread_mutex_lock(&drmr->load_mutex);
if (nn >= 0 && nn < drmr->num_samples) {
if (drmr->samples[nn].layer_count > 0) {
layer_to_sample(drmr->samples+nn,*(drmr->gains[nn]));
if (drmr->samples[nn].limit == 0)
fprintf(stderr,"Failed to find layer at: %i for %f\n",nn,*drmr->gains[nn]);
}
pthread_mutex_unlock(&drmr->load_mutex);
break;
drmr->samples[nn].active = 1;
drmr->samples[nn].offset = 0;
}
default:
printf("Unhandeled status: %i\n",(*data)>>4);
}
} else printf("unrecognized event\n");
lv2_event_increment(&eit);
}
pthread_mutex_unlock(&drmr->load_mutex);
break;
}
default:
printf("Unhandeled status: %i\n",(*data)>>4);
}
} else printf("unrecognized event\n");
}
for(i = 0;i<n_samples;i++) {