Catch all OSC Messages / RegExpr on OSC Formatted strings
Hi there,
I'm trying to catch all incoming OSC Messages with no success.
First wildcards like /* and /? don't work on the OSCDecode Node, but are stated to work in the OSC standard.
Second I'm trying to RegExpr on the #bunlde string of the OSCEncoder, with no success.
The bundle message with all characters looks like this
- bundle#0#0#0#0#0#0#0#0#1#0#0#0$/hello/what#0,s#0#0soever#32but#32its#32ok#0#0#0
Now if I RegExpr on it with this expression
\/(.*?)\#
I get 'hello/what returned, which is the OSC address.
But! This doesn't work when the output of OSCEncoder is connected directly to the RegExpr Node, only when the string is stored in a StringIOBox (and the connection to OSCEncoder cut). I assume there is some reformatting from UTF to ASCII going on, which the RegExpr isn't capable of. Is there a way of doing this manually?
THX
Eno
oi eno,
the problem with RegExpr might be, that the # in the osc messages are not that ascii sybol but rather 'escapes', e.g. '#0' is the ascii symbol for 0;
osc syntax is structured in 4 (and sometimes 8) bytes. might be easier for manual parsing, if you first split the message accordingly
e.g. for bundled messages get rid of the 16 characters in the beginning (#bundle and timetag), what you have left are the osc messages with an integer in the beginning (padded with #0 to be 4 bytes length), which tells you the lenght of each message....
might be easier to solve in a plugin because of that byte-length hassle
Check out https://github.com/elliotwoods/VVVV.Nodes.OSC for an alternative workflow for osc in vvvv (it just wraps vvvv's own c# osc api mostly)
Its designed to be more high level / cleaner to use than oscencode, etc
Uses global channels and a 'Sift' node (which specifically you want to separate out and not use)
anonymous user login
Shoutbox
~4min ago
~44min ago
~55min ago
~2h ago
~3h ago
~3h ago
~5h ago
~5h ago
~8h ago