Recursion, with "some_dict={}" function parameter.

[MvGulik]

I'm not getting some python(3.5.2) behaviour.

I got a function def get_links_titles( links_list, titles_dict={}, r=0 ):. (Key-part: titles_dict={})

In it I recall(recursion) it with titles_del = get_links_titles( case_dict['redirlinks'], {}, r=r+1 ) it works as intended. The recursive case start with an empty titles_dict case.

But when I don't use the , {} part, and use titles_del = get_links_titles( case_dict['redirlinks'], r=r+1 )
The titles_dict in the recursion case is not empty, but contains the data titles_dict had before the recursive call. (???).

Ergo: What Python behaviour rule(s) I'm I missing to make sense of that last case.

---

Mmm, seems that titles_dict={} as function parameter is not a good idea based on some similar down the line unsuspected hiccup.
So ditched it in the def-part and moved it inside the function. (not needed in def + wip function. re-investigate later.)

[fishhook]

Dictionaries are objects that are always passed as links to the original. Never do the following:

def get_links_titles( links_list, titles_dict={}, r=0 ):
     pass

in this case, you will have only one object titles_dict through all of the function calls. Do like this:

def get_links_titles(links_list, titles_dict=None, r=0):
     titles_dict = titles_dict or {}

[ThomasL]

The same is with lists as default arguments
https://nikos7am.com/posts/mutable-default-arguments/

[MvGulik]

Aha. Its starting to make more sense to me now.

Definitely a subject I need to explore some more (todo). Additional Link definitely going to help with that.
(I think I used this ones to my benefit, but without actually realising why or how it worked)

Thanks for putting me on the right track. :-)