I am trying to check if a dictionary is empty but it doesn’t behave properly. It just skips it and displays ONLINE without anything except of display the message. Any ideas why ?

 def isEmpty(self, dictionary):
   for element in dictionary:
     if element:
       return True
     return False

 def onMessage(self, socket, message):
  if self.isEmpty(self.users) == False:
     socket.send("Nobody is online, please use REGISTER command" \
                 " in order to register into the server")
  else:
     socket.send("ONLINE " + ' ' .join(self.users.keys())) 

Empty dictionaries evaluate to False in Python:

>>> dct = {}
>>> bool(dct)
False
>>> not dct
True
>>>

Thus, your isEmpty function is unnecessary. All you need to do is:

def onMessage(self, socket, message):
    if not self.users:
        socket.send("Nobody is online, please use REGISTER command" \
                    " in order to register into the server")
    else:
        socket.send("ONLINE " + ' ' .join(self.users.keys()))

Here are three ways you can check if dict is empty. I prefer using the first way only though. The other two ways are way too wordy.

test_dict = {}

if not test_dict:
    print "Dict is Empty"


if not bool(test_dict):
    print "Dict is Empty"


if len(test_dict) == 0:
    print "Dict is Empty"

dict = {}
print(len(dict.keys()))

if length is zero means that dict is empty

Simple ways to check an empty dict are below:

        a= {}

    1. if a == {}:
           print ('empty dict')
    2. if not a:
           print ('empty dict')

Although method 1st is more strict as when a = None, method 1 will provide correct result but method 2 will give an incorrect result.

A dictionary can be automatically cast to boolean which evaluates to False for empty dictionary and True for non-empty dictionary.

if myDictionary: non_empty_clause()
else: empty_clause()

If this looks too idiomatic, you can also test len(myDictionary) for zero, or set(myDictionary.keys()) for an empty set, or simply test for equality with {}.

The isEmpty function is not only unnecessary but also your implementation has multiple issues that I can spot prima-facie.

1. The return False statement is indented one level too deep. It should be outside the for loop and at the same level as the for statement. As a result, your code will process only one, arbitrarily selected key, if a key exists. If a key does not exist, the function will return None, which will be cast to boolean False. Ouch! All the empty dictionaries will be classified as false-nagatives.
2. If the dictionary is not empty, then the code will process only one key and return its value cast to boolean. You cannot even assume that the same key is evaluated each time you call it. So there will be false positives.
3. Let us say you correct the indentation of the return False statement and bring it outside the for loop. Then what you get is the boolean OR of all the keys, or False if the dictionary empty. Still you will have false positives and false negatives. Do the correction and test against the following dictionary for an evidence.

myDictionary={0:'zero', '':'Empty string', None:'None value', False:'Boolean False value', ():'Empty tuple'}

You can also use get(). Initially I believed it to only check if key existed.

>>> d = { 'a':1, 'b':2, 'c':{}}
>>> bool(d.get('c'))
False
>>> d['c']['e']=1
>>> bool(d.get('c'))
True

What I like with get is that it does not trigger an exception, so it makes it easy to traverse large structures.

Why not use equality test?

def is_empty(my_dict):
    """
    Print true if given dictionary is empty
    """
    if my_dict == {}:
        print("Dict is empty !")

use ‘any’

dict = {}

if any(dict) :

     # true
     # dictionary is not empty 

else :

     # false 
     # dictionary is empty