explicit username in IRC backend

[aadis]

I had another issue on trying to use the IRC backend for flowdock. In flowdock the nickname and the username are different. Here is my diff:

--- irc.py      2014-06-22 04:17:54.153544441 +0530
+++ /home/idbot/.virtualenvs/idbot27/lib/python2.7/site-packages/errbot/backends/irc.py 2014-04-19 14:01:08.425979216 +0530
 class IRCConnection(SingleServerIRCBot):
-    def __init__(self, callback, nickname, server, port=6667, ssl=False, password=None):
+    def __init__(self, callback, nickname, server, port=6667, ssl=False, password=None, username=None):
         self.use_ssl = ssl
         self.callback = callback
-        super().__init__([(server, port, password)], nickname, nickname)
+        super().__init__([(server, port, password)], nickname, username)

     def connect(self, *args, **kwargs):
         if self.use_ssl:
 class IRCBackend(ErrBot):
-    def __init__(self, nickname, server, port=6667, password=None, ssl=False):
+    def __init__(self, nickname, server, port=6667, password=None, ssl=False, username=None):
         super(IRCBackend, self).__init__()
-        self.conn = IRCConnection(self, nickname, server, port, ssl, password)
+        self.conn = IRCConnection(self, nickname, server, port, ssl, password, username)

This should be improved by setting username = nickname when username is None. If you want I can send in a pull request for this change.

[gbin]

Same thing, if you could open a proper pull request, it would be awesome. thx

[aadis]

Sent pull request (for both changes). I've tested the changes on my end, and they seem to work.

For the irc backend, the following config now works:

# The identity, or credentials, used to connect to a server
BOT_IDENTITY = {
    'nickname' : 'foobar',
    'username': 'foobar@example.com',
    'password' : 'foobar@example.com foobarpassword',
    'server' : 'irc.flowdock.com',
    'port': 6697, 
    'ssl': True,
}

(BTW, this config works for flowdock). Perhaps this can be documented somewhere?

[sijis]

Looks like this was fixed in PR #256