Retry rtm.start if it fails

If you're connecting to multiple teams, especially if one of them is
large, the call to rtm.start may time out. By retrying it, there's a
greater chance that it works (especially since when it tries again, the
requests to rtm.start for the other teams have probably finished).

1 files changed, 7 insertions, 0 deletions

diff --git a/wee_slack.py b/wee_slack.py
index 5c75431..a0fcb3c 100644
--- a/wee_slack.py
+++ b/wee_slack.py
@@ -442,6 +442,13 @@ class EventRouter(object):
         elif return_code != -1:
             self.reply_buffer.pop(request_metadata.response_id, None)
             self.delete_context(data)
+            if request_metadata.request == 'rtm.start':
+                w.prnt('', ('Failed connecting to slack team with token starting with {}, ' +
+                        'retrying. If this persists, try increasing slack_timeout.')
+                        .format(request_metadata.token[:15]))
+                dbg('rtm.start failed with return_code {}. stack:\n{}'
+                        .format(return_code, ''.join(traceback.format_stack())), level=5)
+                self.receive(request_metadata)
         else:
             if request_metadata.response_id not in self.reply_buffer:
                 self.reply_buffer[request_metadata.response_id] = StringIO()