__eq__ (equality) can in some cases return false for two identical graphs

[Sappique]

In some cases the private _edges of a graph can gain an empty dict as a entry. This leads to incorrect behavior when comparing graphs.
This bug is very similar to #40.

To reproduce

Python 3.10.0 (tags/v3.10.0:b494f59, Oct  4 2021, 19:00:18) [MSC v.1929 64 bit (AMD64)] on win32
>>> from graph import Graph
>>> g1 = Graph(from_list=[(1,2),(2,3)])
>>> g2 = Graph(from_list=[(1,2),(2,3)])
>>> g1.edge(3,1)
>>> g1 == g2
False
>>> g1._edges
defaultdict(<class 'dict'>, {1: {2: 1}, 2: {3: 1}, 3: {}})
>>> g2._edges
defaultdict(<class 'dict'>, {1: {2: 1}, 2: {3: 1}})

Cause

The equality function of graphs compares the private variable _edges of both graphs and if they are not equal, the function returns false.
The variable _edges can in some cases be accessed with keys that don't exist, since it is a defaultdict(dict) in those cases a empty dict is created for that key. If that happens with only one of two identical graphs, they are no longer equal.
The accessing of _edges with not existing keys and thus the creation of a new entry can occur when edge(node1, node2) is called with a node1 without outgoing edges or sometimes when is_connected(n1, n2) is called depending on the graph. Those are the two examples I found, there could be more.

Other problems

Just like with #40 this also causes strange behavior when copying graphs, because copy() uses the public edge getter edges() instead of the private variable _edges.

Python 3.10.0 (tags/v3.10.0:b494f59, Oct  4 2021, 19:00:18) [MSC v.1929 64 bit (AMD64)] on win32
>>> from graph import Graph
>>> g1 = Graph(from_list=[(1,2),(2,3)])
>>> g1.edge(3,1)
>>> g2 = g1.copy()
>>> g1 == g2
False
>>> g1._edges
defaultdict(<class 'dict'>, {1: {2: 1}, 2: {3: 1}, 3: {}})
>>> g2._edges  
defaultdict(<class 'dict'>, {1: {2: 1}, 2: {3: 1}})

Additional information

I'm using version 2023.7.6 of graph-theory with Python 3.10 on windows.

[root-11]

Thank you so much for the write-up @Sappique. This is a role-model bug report. Excellent work!

I have added your tests in a9be334 and added the bugfix in 58d78ac .

The release 2023.7.7 is now on pypi.org thanks to you.