The way of determining modes here is error prone.

[llmII]

go-ircevent/irc_nicktrack.go

Lines 37 to 48 in e71f5fe

	for _, modenick := range strings.Split(e.Message(), " ") {
	nickandmode := mode_split.FindStringSubmatch(modenick)
	u := User{}
	if len(nickandmode) == 3 {
	if nickandmode[1] == "@" {
	u.Mode = "+o" // Ooof should be mode struct?
	} else if nickandmode[1] == "+" {
	u.Mode = "+v" // Ooof should be mode struct?
	} else if nickandmode[1] == "%" {
	u.Mode = "+h"
	}
	irc.Channels[channelName].Users[nickandmode[2]] = u

We need to get the mode characters from RPL_ISUPPORT. This would stop us from needing to do things like the following:

https://github.com/llmII/go-discord-irc/blob/485c29f816b8e5007fa80a47e5ecf21034acd98e/bridge/irc_listener.go#L61-L74

[llmII]

Should remove need for mode_split regex (just loop through the list of modes on first character and set an index for nick if matched). Parse out ISUPPORT for nick length and prefix modes. These are sort of pointers at "how to fix".