如何拆分字符串列表并使用递归返回元组？

Question

One function that I am having trouble with is split_list, where I have to split a list into three tuples, with each containing a strlist, through recursion. The first tuple with strings must start with a vowel, the second tuple with strings must start with a consonant, and the third tuple with strings should not start with an alpha character.

class Node:
    def __init__(self, value, rest):
        self.value = value
        self.rest = rest
    def __eq__(self, other):
        return ((type(other) == Node)
          and self.value == other.value
          and self.rest == other.rest
        )
    def __repr__(self):
        return ("Node({!r}, {!r})".format(self.value, self.rest))

# a StrList is one of
# - None, or
# - Node(string, StrList)

def split_list(strlist):
    if strlist is None:
       return (None, None, None)
    res = split_list(strlist.rest)
    if strlist.value or res(strlist.value) == 'AEIOUaeiou':
       return (Node(strlist.value, res[0]), res[1], res[2])
    if strlist.value or res(strlist.value) != 'AEIOUaeiou':
       return (res[0], Node(strlist.value, res[1]), res[2])
    if strlist.value.isalpha() or res(strlist.value) == False:
       return (res[0], res[1], Node(strlist.value, res[2]))

This is what I have currently, but the main problem is that I am not getting the correct output when I run my unit tests. The problem I have is AssertionError.

Examples:

strlist = Node("xyz", Node("Abc", Node("49ers", None)))
self.assertEqual(split_list(strlist),(Node('Abc', None), Node('xyz', None), Node('49ers', None)))

strlist = Node("Yellow", Node("abc", Node("$7.25", Node("lime", Node("42", Node("Ethan", None))))))
self.assertEqual(split_list(strlist),(Node('abc', Node("Ethan", None)), Node('Yellow', Node("lime", None)), Node('$7.25', Node("42", None))))

Output:

AssertionError: Tuples differ: (Node('Yellow', Node('abc', Node('$7.25', Node('[50 chars]None) != (Node('abc', Node('Ethan', None)), Node('Yellow'[50 chars]ne)))

Can somebody figure out the problem? I would be thankful.

Answer 1

You should use recursion only for the traversal part. The recursive function should not contain 3 specific and distinct process conditions.

What you could do is write generic a node filter function that will build a new node chain and use it to assemble your tuple. Without the specific conditions, this function would be very simple and easy to test:

def filter(self,condition): # on the Node class
    nextFound = self.rest.filter(condition) if self.rest else None
    return Node(self.value,nextFound) if condition(self) else nextFound 

Then you can build your tuple using this filter function with different parameters:

def split_list(strlist):
    return ( strlist.filter(lambda n:n.value[:1].upper() in "AEIOU"), 
             strlist.filter(lambda n:n.value[:1].upper() in "BCDFGHJKLMNPQRSTVWXYZ"), 
             strlist.filter(lambda n:not n.value[:1].isalpha()) )

[EDIT] if you must do it all in a single function, you could combine the two (unwieldy as it may be):

def split_list(strlist):
    nextFound = split_list(strlist.rest) if strlist.rest else (None,None,None)
    return ( Node(strlist.value,nextFound[0]) if strlist.value[:1].upper() in "AEIOU" else nextFound[0], 
             Node(strlist.value,nextFound[1]) if strlist.value[:1].upper() in "BCDFGHJKLMNPQRSTVXYZ" else nextFound[1], 
             Node(strlist.value,nextFound[2]) if not strlist.value[:1].isalpha() else nextFound[2] )

if this for a homework, I sincerely hope your teacher will promote separation of concerns and not require questionable coding practices