def __init__(self, s):
s = s.split(':-')
left = s[1].split('&')
right = s[0]
self.left = []
self.right = Relation(right)
for i in left:
temp = Relation(i)
self.left.append(temp)
class Infer:
#Prolog
#VD: husband(Person,Wife):-male(Person);married(Person,Wife)
#left -> right
def __init__(self, s):
s = s.split(':-')
left = s[1].split('&')
right = s[0]
self.left = []
self.right = Relation(right)
for i in left:
temp = Relation(i)
self.left.append(temp)
def generate(self, KB):
#Chua cac gia tri của cac bien ben trai co the thay the vao trong infer
temp = []
#Chua ten cac bien ben trai
temp1 = []
for i in self.left:
temp1 += i.getDatas()
#Luu nhung quan he tuong tu voi i
temp2 = []
for j in KB:
if i.isCompatible(j):
temp2.append(j.getDatas())
temp.append(temp2)
#Intertools.product(temp) tao mang ket hop cac mang trong temp lai voi nhau
#VD: temp = [[1,2],[3,4]]
#Intertools.product tra ve [[1,3],[1,4],[2,3],[2,4]]
for x in itertools.product(*temp):
temp2 = []
for i in x:
temp2 = temp2 + i
#Xet xem bo x co phu hop voi temp1 hay khong?
if check(temp2, temp1):
data = []
temp3 = self.right.getDatas()
for j in temp3:
index = temp1.index(j)
data.append(temp2[index])
temp4 = Relation("Relation(Data)")
temp4.setDatas(data)
temp4.setName(self.right.getName())
if temp4 in KB:
continue
return temp4
return False
def tostring(self):
string_left = []
for i in self.left:
string_left.append(i.toString())
string_left = ' ^ '.join(string_left)
return string_left + ' -> ' + self.right.toString()
def from_file(cls, filepath):
relations = []
lines = open(filepath).readlines()
for line in lines:
j = json.loads(line)
r = Relation(**j)
r.entity1 = Entity(**r.entity1)
r.entity2 = Entity(**r.entity2)
relations.append(r)
return RelationCollection(relations)
def readLibrary(fl):
err_str = ''
war_str = ''
list_obj = []
list_rel = []
ext_obj = []
ext_rel = []
dict_ft = defaultdict(list)
try:
with open(Object.folder + fl) as f:
#with object_pairs_hook method, we can check "Error 07" at the same level
target = json.loads(f.read(), object_pairs_hook=Object.dupe_checking_hook)
reldict = defaultdict(list)
err_str, war_str = Object.val_lib(fl, target, target, True, list_obj,
list_rel, reldict, ext_obj, ext_rel, dict_ft)
#pprint.pprint(reldict)
#print('\n'*2)
err_str += Object.cycle_test(reldict, [], '', True)
err_str += Object.notdefined(list_obj, list_rel, ext_obj, ext_rel, dict_ft)
if not err_str:
for key in target.keys():
if key == "objects":
for objectName in target[key].keys():
#We add all the objects contained in "objectName" into global dictionary "Object.
#objects" and their name into local "objects" list to give a tree structure
obj = Object(objectName, target[key][objectName])
#add object into Object.objects
Object.addObject(objectName, obj)
#add object into Object.objectsLib
if not Object.comp:
Object.objectsLib[objectName] = obj
else:
Object.objectsLib2[objectName] = obj
elif key == "relations":
for relationName in target[key].keys():
relation = Relation(relationName, target[key][relationName])
Relation.addRelation(relationName, relation)
Object.addRelation(relationName, relation)
if not Object.comp:
Object.relationsLib[relationName] = relation
else:
Object.relationsLib2[relationName] = relation
except IOError:
err_str += 'Error 02: library file is not found!\n'
list_obj = ['Error 02']
return [err_str, war_str, list_obj, list_rel]
def start(self):
## summarize
self.sum = Summarize(self.sentences, self.num_cluster)
self.sum.save(self.collection, self.oid)
## time
self.time = Time(self.sentences, self.sum.clusters)
self.time.save(self.collection, self.oid)
## keyword
self.keyword = Tf_Idsf(self.sentences)
self.keyword.save(self.collection, self.oid)
## relation
self.relation = Relation(self.sentences, self.mid)
self.relation.save(self.collection, self.oid)
def load_relations(self):
# load relation from file
relation_data = list(open("../data/RE/relation2id.txt").readlines())
relation_data = [s.split() for s in relation_data]
for relation in relation_data:
r = Relation(relation[0], int(relation[1]))
self.relations[relation[0]] = r
for r in self.relations:
self.relations[r].generate_vector(len(self.relations))
print("RelationTotal: " + str(len(self.relations)))
class Analyzer():
def __init__(self, mid):
conn = pymongo.MongoClient('13.209.73.233', 27017)
db = conn.get_database('test')
self.collection = db.get_collection('meets')
self.mid = mid
self.oid = ObjectId(self.mid)
## get talk
result = self.collection.find({"_id": self.oid}, {
"_id": False,
"talk": True,
"num_cluster": True
})
if result[0]:
talk = sorted(result[0]['talk'], key=lambda x: x[1])
self.sentences = [Sentence(i, t) for i, t in enumerate(talk)]
self.num_cluster = result[0]['num_cluster']
print(self.num_cluster)
self.error = False
else:
self.error = True
def start(self):
## summarize
self.sum = Summarize(self.sentences, self.num_cluster)
self.sum.save(self.collection, self.oid)
## time
self.time = Time(self.sentences, self.sum.clusters)
self.time.save(self.collection, self.oid)
## keyword
self.keyword = Tf_Idsf(self.sentences)
self.keyword.save(self.collection, self.oid)
## relation
self.relation = Relation(self.sentences, self.mid)
self.relation.save(self.collection, self.oid)
def generate(self, KB):
#Chua cac gia tri của cac bien ben trai co the thay the vao trong infer
temp = []
#Chua ten cac bien ben trai
temp1 = []
for i in self.left:
temp1 += i.getDatas()
#Luu nhung quan he tuong tu voi i
temp2 = []
for j in KB:
if i.isCompatible(j):
temp2.append(j.getDatas())
temp.append(temp2)
#Intertools.product(temp) tao mang ket hop cac mang trong temp lai voi nhau
#VD: temp = [[1,2],[3,4]]
#Intertools.product tra ve [[1,3],[1,4],[2,3],[2,4]]
for x in itertools.product(*temp):
temp2 = []
for i in x:
temp2 = temp2 + i
#Xet xem bo x co phu hop voi temp1 hay khong?
if check(temp2, temp1):
data = []
temp3 = self.right.getDatas()
for j in temp3:
index = temp1.index(j)
data.append(temp2[index])
temp4 = Relation("Relation(Data)")
temp4.setDatas(data)
temp4.setName(self.right.getName())
if temp4 in KB:
continue
return temp4
return False
def dbProcessSchemes(self):
relationList = []
for table in self.tableNames:
relObj = Relation()
relObj.Name = table
for columns in self.insp.get_columns(table):
attribute_object = Attribute(columns['name'])
attribute_object.Type = columns['type']
attribute_object.Autoincrement = columns['autoincrement']
attribute_object.Nullable = columns['nullable']
relObj.Attributes.append(attribute_object)
#attribute_object = Attribute()
#attribute_object.Name = columns['name']
key = Key()
for att in self.insp.get_primary_keys(table):
attribute_object = Attribute(att)
key.Attributes.append(attribute_object)
relObj.Keys.append(key)
#relObj.FDs = self.getFunctionalDependencies(relation)
relationList.append(relObj)
return relationList
def writeAnswer(self,questions):
# &: and
# |: or
# ~: not
fo=input('Nhập tên file output: ')
f=open(fo,'w')
ordinal_number = 0
for i in questions:
ordinal_number+=1
temp=i.split('?-')
f.write(str(ordinal_number))
f.write('.')
# f.write(temp[0])
f.write('\n')
temp1 = temp[1].split('?')
q = temp1[-1]
f.write('?-')
f.write(''.join(temp[1]))
f.write('\n')
if len(temp1)>1:
var = temp1[0:-1]
temp2=temp1[-1].split('&')
#Luu ket qua cua tung vi tu
res_temp=[]
#Luu cac bien
temp4 = []
#Giai ket qua cua tung vi tu
for k in temp2:
check_result1 = True
temp3=Relation(k)
for x in temp3.getDatas():
temp4.append(x)
#Luu cac ket qua cua 1 vi tu
res_temp1=[]
for j in self.KB:
if temp3.getName() == j.getName():
index =0
n = len(temp3.getDatas())
while index < n:
if temp3.getData(index) in var:
index+=1
continue
elif temp3.getData(index) == j.getData(index):
index+=1
else:
break
if index == n:
res_temp1.append(j.getDatas())
if len(res_temp1) ==0:
check_result1 = False
break
else:
res_temp.append(res_temp1)
if check_result1 == False:
f.write("No result!!!")
f.write('\n')
else:
if len(res_temp) >1:
check_result2 = False
#Ket hop nhung ket qua lai voi nhau tao thanh 1 bo
for x in itertools.product(*res_temp):
temp5= []
for y in x:
temp5.extend(y)
#Bo nao thoa dieu kien thi ghi ket qua bo do
if check(temp5,temp4):
check_result2 = True
for res_var in var:
index = temp4.index(res_var)
f.write(res_var +' = ' + temp5[index])
f.write('\t')
f.write('\n')
if check_result2 == False:
f.write("No result!!!")
f.write('\n')
else:
for x in res_temp[0]:
if check(x,temp4):
for res_var in var:
index = temp4.index(res_var)
f.write(res_var +' = ' + x[index])
f.write('\t')
f.write('\n')
else:
s = split_string(q)
for i in range(len(s)):
if s[i] != '&' and s[i] != '|' and s[i] != '[' and s[i] != ']':
temp1 = Relation(s[i])
res_temp = False
for j in self.KB:
if temp1 ==j:
res_temp=True
if res_temp == True:
s[i]='1'
else:
s[i]='0'
res = eval(''.join(s))
res_write = ''
if res == 1:
res_write = 'True'
else:
res_write = 'False'
f.write(res_write)
f.write('\n')
print("Xu ly query xong!!!")
f.close()
from Program import Program
from Relation import Relation
from Infer import Infer
#Doc file input
fp = input('Nhập tên file input: ')
f = open(fp, 'r')
f1 = f.read().splitlines()
KB = []
I = []
questions = []
flag = 0
for x in f1:
if x == '*':
flag += 1
continue
if flag == 0:
temp = Relation(x)
KB.append(temp)
elif flag == 1:
temp = Infer(x)
I.append(temp)
else:
questions.append(x)
f.close()
main = Program(KB, I)
main.run()
#main.printKB()
main.writeAnswer(questions)
Cntrs = Countries()
Cntrs.add(cntr1)
Cntrs.add(cntr2)
Cntrs.add(cntr3)
Cts = Cities()
Cts.add(ct1)
Cts.add(ct2)
Cts.add(ct3)
Cts.add(ct4)
Cts.add(ct5)
Cts.add(ct6)
ER = Relation()
ER.add(0, 0, Cntrs, Cts)
ER.add(0, 1, Cntrs, Cts)
ER.add(0, 2, Cntrs, Cts)
ER.add(1, 3, Cntrs, Cts)
ER.add(2, 4, Cntrs, Cts)
ER.add(2, 5, Cntrs, Cts)
with open('Relation.pickle', 'wb') as f:
pickle.dump(ER, f)
with open('Cities.pickle', 'wb') as f:
pickle.dump(Countries, f)
with open('Countries.pickle', 'wb') as f:
exit(0)
try:
test = int(sys.argv[1])
except ValueError:
print "Usage : python test.py TEST_NUMBER"
exit(0)
# create an SQLite test database
DB_NAME = 'test_db'
create_database(DB_NAME)
# get attributes from database
emp = SQLiteRelation(DB_NAME, 'emp')
dept = SQLiteRelation(DB_NAME, 'dept')
# manually create attributes, SQL can't be generated but errors will be checked
man_rel = Relation([Attribute('name', 'TEXT'), Attribute('age', 'INTEGER')])
if test == 0:
print str_attributes(man_rel.getAttributes())
# test SQLiteRelation
if test == 1:
rel = emp
if test == 2:
rel = dept
# Test SelectAttribute
if test == 10:
# error : attribute does not exists
rel = SelectAttribute('ename', 'nonexistant', emp)
if test == 11:
# error : not comparable attributes
import copy
from prettytable import PrettyTable
from GoodShit import GoodShit
from Relation import Relation
# Relacje - przyklad dla zadania ze slajdow
R1 = Relation(
'R1', 1000, {
'rows_identifiers': ['SIZE', 'VAL'],
'columns_identifiers': ['Property', 'A', 'B']
})
R1.populate_table([[4, 4], [200, 50]])
R2 = Relation(
'R2', 1000, {
'rows_identifiers': ['SIZE', 'VAL'],
'columns_identifiers': ['Property', 'A', 'B']
})
R2.populate_table([[4, 4], [40, 100]])
R3 = Relation(
'R3', 2000, {
'rows_identifiers': ['SIZE', 'VAL'],
'columns_identifiers': ['Property', 'B', 'D']
})
R3.populate_table([[4, 4], [400, 100]])
R4 = Relation(
'R4', 1000, {
'rows_identifiers': ['SIZE', 'VAL'],
def loadRelationFile(hin, relationFileName, startEntityType, endEntityType):
# Judge whether the entity type exists in the HIN
if startEntityType not in hin['EntityTypes']:
hin['EntityTypes'][startEntityType] = dict()
if endEntityType not in hin['EntityTypes']:
hin['EntityTypes'][endEntityType] = dict()
# Initialize relations list
startTypeToEndType = startEntityType + '-' + endEntityType
endTypeToStartType = endEntityType + '-' + startEntityType
hin['RelationTypes'][startTypeToEndType] = []
hin['RelationTypes'][endTypeToStartType] = []
hinStartToEndList = hin['RelationTypes'][startTypeToEndType]
hinEndToStartList = hin['RelationTypes'][endTypeToStartType]
hinRelationList = hin['Relations']
# Start reading data file
with open(DATASET_DIR + relationFileName, 'r') as relationFile:
while True:
relationLine = relationFile.readline()
if not relationLine:
break
relationLine = relationLine.strip()
relation = relationLine.split(' ')
# Data sensitive -> type ...
# Load the start entity and the end entity
startEntityId = int(relation[0])
endEntityId = relation[1]
if endEntityId.isdigit():
endEntityId = int(endEntityId)
weight = None
# The third column is regarded as the weight
if len(relation) > 2:
weightStr = relation[2]
# Judge whether the type of the weight should be float or int
if '.' in weightStr:
weight = float(weightStr)
else:
weight = int(weightStr)
# Entity location
tmpStartEntityInfo, tmpStartEntityIndex = locateEntity(
hin, startEntityId, startEntityType)
tmpEndEntityInfo, tmpEndEntityIndex = locateEntity(
hin, endEntityId, endEntityType)
tmpStartEntity = tmpStartEntityInfo.entity
tmpEndEntity = tmpEndEntityInfo.entity
# Construct a relation and its reverse one
tmpRelation = Relation(tmpStartEntity, tmpEndEntity, weight)
hinRelationList.append(tmpRelation)
tmpRelationIndex = len(hinRelationList) - 1
tmpReverseRelation = Relation(tmpEndEntity, tmpStartEntity)
hinRelationList.append(tmpReverseRelation)
tmpReverseRelationIndex = len(hinRelationList) - 1
# Add the relations in entity and HIN
tmpStartEntityInfo.addRelation(tmpRelation, tmpRelationIndex,
False)
tmpEndEntityInfo.addRelation(tmpRelation, tmpRelationIndex, True)
hinStartToEndList.append(tmpRelationIndex)
tmpStartEntityInfo.addRelation(tmpReverseRelation,
tmpReverseRelationIndex, True)
tmpEndEntityInfo.addRelation(tmpReverseRelation,
tmpReverseRelationIndex, False)
hinEndToStartList.append(tmpReverseRelationIndex)
relationFile.close()