def test_for_non_operators(self):
self.assertFalse(sy.isOperator("1"))
self.assertFalse(sy.isOperator("("))
self.assertFalse(sy.isOperator(")"))
self.assertFalse(sy.isOperator("["))
self.assertFalse(sy.isOperator("]"))
self.assertFalse(sy.isOperator("a"))
self.assertFalse(sy.isOperator("x"))
self.assertFalse(sy.isOperator("d"))
self.assertFalse(sy.isOperator("p"))
self.assertFalse(sy.isOperator("m"))
self.assertFalse(sy.isOperator("|"))
self.assertFalse(sy.isOperator("~"))
self.assertFalse(sy.isOperator("\\"))
def test_wrong_input_type(self):
with self.assertRaises(TypeError):
sy.isOperator(1)
def test_for_operators(self):
self.assertTrue(sy.isOperator("+"))
self.assertTrue(sy.isOperator("-"))
self.assertTrue(sy.isOperator("*"))
self.assertTrue(sy.isOperator("/"))
def test_empty_char(self):
with self.assertRaises(ValueError):
sy.isOperator('')
def test_a_string(self):
with self.assertRaises(ValueError):
sy.isOperator('21')
def test_non_operator(self):
isOperator = sy.isOperator('0')
self.assertFalse(isOperator)
def test_an_operator(self):
isOperator = sy.isOperator('-')
self.assertTrue(isOperator)
def search3(self,query):
"search for many query with boolean expression"
root = self.inverted_index.root
query_list = query.split()
expression = shunting_yard.infixToRPN(query_list)
expression.reverse()
stack = []
plist_num=0
plist_hashlist={}
#[2012/08/22 added] to initial variables to find score
slist_num=0
slist_hashlist={}
# just find that term if doing only single word query.
if len(expression) ==0:
return []
elif len(expression) <= 1:
word = expression.pop()
answer_last= self.inverted_index.lookup(root,word)
# get node to find tf-idf
node_answer_last = self.inverted_index.lookupNode(root,word)
if(node_answer_last==None):
return []
tf_idf_list = node_answer_last.get_tf_idf_list(self.inverted_index.docCount,self.inverted_index.highest_tc)
ranked_answer = Search_System.rank_answer(answer_last,tf_idf_list)
return ranked_answer
# in case of > 1 word query
while len(expression)!=0:
token = expression.pop()
if shunting_yard.isOperator(token):
word1 = stack.pop()
if( not word1 in plist_hashlist ):
answer_list1 = self.inverted_index.lookup(root,word1)
# [2012/08/22 added] to get score of word
node1 = self.inverted_index.lookupNode(root,word1)
if(node1!=None):
score_list1 = node1.get_tf_idf_list(self.inverted_index.docCount,self.inverted_index.highest_tc)
else: score_list1 = []
else:
answer_list1 = plist_hashlist[word1]
# [2012/08/22 added] to get score of word
score_list1 = slist_hashlist[word1]
word2 = stack.pop()
if( not word2 in plist_hashlist ):
answer_list2 = self.inverted_index.lookup(root,word2)
# [2012/08/22 added] to get score of word
node2 = self.inverted_index.lookupNode(root,word2)
if(node2!=None):
score_list2 = node2.get_tf_idf_list(self.inverted_index.docCount,self.inverted_index.highest_tc)
else: score_list2 = []
else:
answer_list2 = plist_hashlist[word2]
# [2012/08/22 added] to get score of word
score_list2 = slist_hashlist[word2]
if(token=="&&"):
answer_last = self.intersect(answer_list1,answer_list2)
# [2012/08/22 added] to get score of word
score_list_last = self.intersect_score(answer_list1,answer_list2,score_list1,score_list2)
elif(token=="||"):
answer_last = self.union(answer_list1,answer_list2)
# [2012/08/22 added] to get score of word
score_list_last = self.union_score(answer_list1,answer_list2,score_list1,score_list2)
else:
print("Wrong Operator: program will exit")
exit()
stack.append("plist"+str(plist_num))
plist_hashlist["plist"+str(plist_num)] = answer_last
plist_num+=1
#[2012/08/22 added] to store new score list in hash table
slist_hashlist["plist"+str(slist_num)] = score_list_last
slist_num+=1
else:
stack.append(token)
# get all the last answers
#[2012/08/22 added] to store new score list in hash table
last_word = stack.pop()
answer = plist_hashlist[last_word];
#[2012/08/22 added] to get answer score list in hash table
s_answer = slist_hashlist[last_word]
print(answer)
print(s_answer)
ranked_answer = Search_System.rank_answer(answer,s_answer)
return ranked_answer