Let's say we have a simple python model class.
class Library:
def __init__():
self.library_name = None
self.books = []
We would use that by:
new_library = Library()
new_library.library_name = "Delhi International Library"
This would be an empty library with no books. Is this safe? I know you shouldn't use mutable types in the method signature due to when it is evaluated.
The idiomatic way to use a mutable type for a default value is to use a None default value:
class Library:
def __init__(self, name, books=None):
self.library_name = name
self.books = [] if books is None else books
new_library = Library("Delhi International Library")
Here name is a required param because IMHO a library should have a name, but you could also use a default if you later will have to create a library with no parameters.
And books is optional. If it is not given, you will get en empty library, and if you pass it on first time, it will not become the default because None is not mutable...
you can use:
class Library:
def __init__(self, library_name, books=None):
self.library_name = library_name
self.books = books or []
and then instantiate like:
new_library = Library("Delhi International Library")
self.book didn't save a referent to it. Use books=None, then set self.books = [] if books is None else book (or an equivalent)Library and doesn't want that object to make changes to its existing list, the caller passes a copy as an argument, rather than assuming Library will make its own copy.
new_library.library_name = "Delhi International Library"would be fine, I'm guessing.nameis a typo?self.books = []as part of the init but not as an argument is OK.Librarywill then have its own (initially empty) list of books, rather than sharing a reference to a single central book list.