diff --git a/toota-palooza/cli.py b/toota-palooza/cli.py
index 8c9ca2e..e12b95e 100644
--- a/toota-palooza/cli.py
+++ b/toota-palooza/cli.py
@@ -18,94 +18,108 @@ from pprint import pprint
 # import toota-palooza
 
 
-def mastodonInit():
-    """Connect to the Mastodon instance based on .env files"""
+class Server(Mastodon):
+    dotenv_loaded = False
 
-    server = Mastodon(
-        client_id = '.toota-palooza.env',
-        api_base_url = os.getenv('MD_HOST')
-    )
+    def __init__(self):
+        self.maybe_load_dotenv()
 
-    load_dotenv()
-    server.log_in(
-        os.getenv('MD_USER'),
-        os.getenv('MD_PASS'),
-        to_file = '.toota-palooza-usercred.env'
-    )
-    return(server)
+        self.client_id = '.toota-palooza.env'
+        self.hostname  = os.getenv('MD_HOST')
+        self.username  = os.getenv('MD_USER')
+        self.passwd    = os.getenv('MD_PASS')
+        self.to_file   = '.toota-palooza-usercred.env'
+
+        super().__init__(client_id=self.client_id, api_base_url=self.hostname)
+
+        self.log_in(
+            self.username,
+            self.passwd,
+            to_file=self.to_file
+        )
+
+    def post(self, message):
+        self.status_post(message, visibility='public')
+
+    @classmethod
+    def maybe_load_dotenv(cls):
+        if not cls.dotenv_loaded:
+            load_dotenv()
+            cls.dotenv_loaded = True
 
 
-def checkPublicTimeline( server):
+def check_public_timeline(server):
     """Do one run. Connect to the database, connect to the server, get the public timeline.
     Look at users and check to see if any are potential impersonators. Then exit."""
 
-    # Here's the idea: pick a chunkSize. Ask the server for that many toots off the public
+    # Here's the idea: pick a chunk_size. Ask the server for that many toots off the public
     # timeline. As long as the server gives us as many as we asked for, keep trying.
     # as soon as we get less than we asked for, quit.
     #
     # XXX Not sure about rate-limiting
     #
-    chunkSize    = 20
-    maxPosts     = 1000
-    timelineList = []
-    useridList   = {}
-    total        = 0
-    calls        = 0
-    while( total < maxPosts ):
-        timelineList = server.timeline(timeline='public', since_id=latestId, limit=chunkSize)
-        calls = calls + 1
-        for post in timelineList:
+    chunk_size    = 20
+    max_posts     = 1000
+    timeline_list = []
+    userid_list   = {}
+    total         = 0
+    calls         = 0
+    domain_def    = server.hostname.split('/')[2]
+    while total < max_posts:
+        timeline_list = server.timeline(timeline='public', limit=chunk_size)
+        calls += 1
+        for post in timeline_list:
             userid = post.account.acct
-            username=userid.split('@')[0]
+            name_and_domain = userid.split('@', 1)
+            username = name_and_domain[0]
 
-            try:
-                domain=userid.split('@')[1]
-            except IndexError:
+            if len(name_and_domain) == 2:
+                domain = name_and_domain[1]
+            else:
                 # if there is no domain, then it's a local account
-                domain=os.getenv('MD_HOST').split('/')[2]
-            useridList[userid] = (username,domain,post.account.display_name,post.account.bot,post.url)
-            latestId=post.id
+                domain = domain_def
+            userid_list[userid] = (username, domain,
+                post.account.display_name, post.account.bot, post.url)
 
-        if( len(timelineList) < 1):
+        if len(timeline_list) == 0:
             # We got fewer than we asked for. Drop out of the loop.
-            total = total + len(timelineList)
             break
-        else:
-            # record how many we did, and go again.
-            total = total + len(timelineList)
-            timelineList = []
+        # record how many we did, and go again.
+        total += len(timeline_list)
+        timeline_list.clear()
 
     # Ok, we got them all, time to insert
-    server.status_post( "{} calls to get {} posts, {} users processed".format(calls, total, len(useridList)), 
-        visibility='public' )
+    return (f'{calls} calls to get {total} posts,'
+        +f' {len(userid_list)} users processed')
 
 def daemon_main():
     """Run from a command line."""
-    server = mastodonInit()
 
-    while(True):
+    while True:
         # do a thing
         time.sleep(600)
     
-def once():
+def once(server):
     """Run from a command line."""
-    server = mastodonInit()
+    message = check_public_timeline(server)
+    server.post(message)
     return 0
 
 def main():
     parser = argparse.ArgumentParser(
-        description='Check for suspicious impersonators.'
-    )
+        description='Check for suspicious impersonators.')
     parser.add_argument( '-d', '--debug', action='store_true',
         help='Enable debugging messages.')
     parser.add_argument( '-o', '--once', action='store_true',
         help='Run once and exit. Default is to run as a daemon.')
     args = parser.parse_args()
 
-    if( args.once ):
-        exit(once())
-    else:
-        daemon_main()
+    server = Server()
+
+    if args.once:
+        return once(server)
+
+    daemon_main(server)
 
 if __name__ == '__main__':
-    exit(once())
\ No newline at end of file
+    sys.exit(main())
\ No newline at end of file