On Mon, Jun 22, 2009 at 09:17:40PM +0800, Peng Jing wrote: > Here are my questions: > 1). > I have read part of Pidgin's source code, but I'm still no so clear about the > implementation of the command processing mechanism. Could anyone tell me how > command processing is implemented in Pidgin?
Basically there's a hashtable that registers callbacks, and when we receive a command we tokenize it and then look up the handler, then we just call the handler. > 2). > I plan to employ some multi-threading: one thread keeps listenning to the > incoming commands from the server, put them in a cache, and the other thread > scans the cache and pick out those complete commands and enqueue them. But > this may involve some thread synchronization problems which are a little > difficult and tedious to deal with. > Does anyone have some good ideas? Yes, don't use threads. -khc _______________________________________________ Support mailing list Support@pidgin.im http://pidgin.im/cgi-bin/mailman/listinfo/support You can unsubscribe from the list's website above.
