The following is the test code.

class C1:
	myurl = []	
	def test(self):
		url = [5,6,7]
		self.myurl.extend(url)		
def testv():
	c = C1()
	c.test()
	print(c.myurl)
	
i = 0
while i<10 :
	testv()
	i = i+1

The output is :

[5, 6, 7]
[5, 6, 7, 5, 6, 7]
[5, 6, 7, 5, 6, 7, 5, 6, 7]
[5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7]
[5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7]
[5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7]
[5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7]
[5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6,
7]
[5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6,
7, 5, 6, 7]
[5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6, 7, 5, 6,
7, 5, 6, 7, 5, 6, 7]

The myurl of class C1 is not set to [] when a new object is created.
All objects use the same memory.

AFAIK, This is expected behavior. myurl is a class attribute if you
want it to be different per instance you should re-initialize it in the
__init__ method. See below.

>>> class C1(object):
...     def __init__(self):
...             self.myurl = []
...     def test(self):
...             self.myurl.extend([5,6,7])
...
[44085 refs]
>>> def testv():
...     c = C1()
...     c.test()
...     print(c.myurl)
...
[44108 refs]
>>> for i in range(10):
...     testv()
...
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[5, 6, 7]
[44119 refs]

Thanks a lot!
Kozyarchuk.

I have thought the self.myurl should be the object variable, not class
variable.
The class variable really is confusing.

Why do they like that?