class TestSkipListCreation(unittest.TestCase):
def setUp(self):
data = range(0, 20)
self.lst = SkipList()
for i in data:
self.lst.append(i)
def test_listCreation(self):
rt = self.lst.root
for i in range(0, 20):
self.assertEqual(rt.val, i)
rt = rt.next
def test_skipCreation(self):
self.lst.create_skips()
# TODO::test that the *right* skip pointers were created, i.e.
# test the targets for the skips!
lngth = self.lst.default_skip_length()
nextSkip = 0
nd = self.lst.root
for i in range(0, 20):
if i == nextSkip and i + lngth < 20:
self.assertNotEqual(len(nd.pointers),0)
nextSkip += lngth
else:
self.assertIsNone(nd.pointers)
nd = nd.next
def index_word(word, docId):
"""
indexes the docId for the given word, creating the skiplist if necessary.
"""
global current_line
if word not in dictionary:
dictionary[word] = current_line
lst = SkipList()
lst.append(docId)
postings.insert(current_line, lst)
current_line += 1
else:
postings[dictionary[word]].append(docId)
def merge_two_list(self, la, lb, op):
"""
FOr NOT, la is irrelevant.TODO::fix? low priority
For OR, AND, order doesn't matter.
"""
lst = SkipList()
nodea = la.root
nodeb = lb.root
if op is Operation.OR:
# SLOW OR MERGE
# ===================================
# while nodea != None and nodeb != None:
# if nodea.val < nodeb.val:
# if lst.last:
# if nodea.val != lst.last.val:
# lst.append(nodea.val)
# else:
# lst.append(nodea.val)
# nodea = nodea.next
# elif nodea.val > nodeb.val:
# if lst.last:
# if nodeb.val != lst.last.val:
# lst.append(nodeb.val)
# else:
# lst.append(nodeb.val)
# nodeb = nodeb.next
# else:
# lst.append(nodea.val)
# nodea = nodea.next
# nodeb = nodeb.next
# while nodea != None:
# if lst.last:
# if nodea.val != lst.last.val:
# lst.append(nodea.val)
# else:
# lst.append(nodea.val)
# nodea = nodea.next
# while nodeb != None:
# if lst.last:
# if nodeb.val != lst.last.val:
# lst.append(nodeb.val)
# else:
# lst.append(nodeb.val)
# nodeb = nodeb.next
# lst.create_skips()
# return lst
# FAST OR MERGE
# ======================================
lst = la.get_list() + lb.get_list()
lst = {}.fromkeys(lst).keys()
lst.sort(key=lambda x: int(x))
lst = SkipList(lst)
lst.create_skips()
return lst
elif op is Operation.AND:
while nodea != None and nodeb != None:
if nodea.val < nodeb.val:
if nodea.pointers != None:
jmp = False
for target in nodea.pointers:
if target.val <= nodeb.val:
nodea = target
jmp = True
if not jmp:
nodea = nodea.next
else:
nodea = nodea.next
elif nodea.val > nodeb.val:
if nodeb.pointers != None:
jmp = False
for target in nodeb.pointers:
if target.val <= nodea.val:
nodeb = target
jmp = True
if not jmp:
nodeb = nodeb.next
else:
nodeb = nodeb.next
else:
lst.append(nodea.val)
nodea = nodea.next
nodeb = nodeb.next
lst.create_skips()
return lst
elif op is Operation.NOT:
# SLOW NOT MERGE
# =====================================
# universal_set = self.UNIVERSAL_SET
# na, nb = universal_set.root, lb.root
# while na != None and nb != None:
# if na.val < nb.val:
# lst.append(na.val)
# na = na.next
# elif na.val > nb.val:
# nb = nb.next
# elif na.val == nb.val:
# na = na.next
# while na != None:
# lst.append(na.val)
# na = na.next
# lst.create_skips()
# return lst
# FAST NOT MERGE
# ======================================
lsta = set(self.search_term("UNIVERSAL_SET").get_list())
lstb = set(lb.get_list())
results = list(lsta - lstb)
results.sort(key=lambda x: int(x))
return SkipList(results)
def get_skipList(self, length):
lst = SkipList()
data = sorted([random.randint(0, length*4) for i in range(0, length)])
for i in range(0, length):
lst.append(data[i]) # TODO:swap out with SkipList(data)
return lst