networkx和列表的迭代-我无法理解的一段代码

Question

I'm new to the python language and have been working with the networkx package. Basically I have a list of customers and producers and want a function that retrieves the list of the current list based on these types.

Here is the relevant code for the function that retrieves customers:

def customers_iter(self, data=False):

    """ Return an iterator over all customers.

        If the network is changed during iteration, the iterator becomes
        invalid.

        Parameters
        -----------

        data -  if True, return a list of (name, attributes) pairs, such
                that attributes == net.node[name]. Otherwise,
                only a list of customer names is returned. Default is
                False.
    """

    if data:
        return (n for n in self.nodes_iter(data=True) 
                    if self.node[n[0]]["type"] == "customer")
    else:
        return (n for n in self.nodes_iter() 
                    if self.node[n]["type"] == "customer")

My question is specifically regarding the if- and else statement. What is the point if first checking the first node n[0]? Doesn't the statement in the else-section define exactly the same thing?

Regards, jazy

Answer 1

According to the doc comment:

data -  if True, return a list of (name, attributes) pairs, such
        that attributes == net.node[name]. Otherwise,
        only a list of customer names is returned. Default is
        False.

Assuming, that data parameter has the same meaning in the self.node_iter function, the first branch ( if ) should filter pairs of (name, attributes) , however the second one ( else ) should filter only name s.

Assuming, also, that self.node is a dictionary-like structure for holding associative pairs name -> node , it is easy to see, that in the first branch we have to retrieve the name from the tuple (which is n[0] ), while in the second branch we could just use n as the node name.