def rank(self):
dictionary = rw.readFile(self.dict_dir).split("\n")[0:-1]
dic = set()
sample = rw.readFile(self.res_dir).split("\n")[0:-1]
words = {}
high_freq = []
content = ""
for line in dictionary:
temp = line.split("\t")
dic.add(temp[0])
for line in sample:
temp = line.split(":")
if (temp[0] in dic):
high_freq.append(line)
limit = int(round(len(high_freq) * self.ratio))
for line in high_freq[:limit]:
#content += (line + "\n")
temp = line.split(':')
content += (temp[0] + '\n')
self.dic_word[temp[0]] = int(temp[1])
rw.writeFile(self.rank_dir,content)
def readDictionary(self):
files = os.listdir(self.dic_dir)
for filename in files:
if filename.endswith('.txt'):
key = filename[0:-4]
if (key not in self.dic):
path = self.dic_dir + '/' + filename
self.dic[key] = rw.readFile(path).split('\r\n')[0:-1]
def get_text(self):
files = os.listdir(self.src_dir)
content = ""
for filename in files:
if filename == 'total.txt':
path = self.src_dir + '/' + filename
txt = rw.readFile(path)
content += txt
return content
def addToDic(self, filename):
src_path = self.src_dir + '/' + filename
lines = rw.readFile(src_path).split('\n')[0:-1]
for line in lines:
temp = line.split('\t')
if temp[0] in self.query_dic:
self.query_dic[temp[0]] += int(temp[1])
else:
self.query_dic[temp[0]] = int(temp[1])
def get_text(self):
files = os.listdir(self.src_dir)
content = ""
for filename in files:
if filename == 'total.txt':
path = self.src_dir + '/' + filename
txt = rw.readFile(path)
content += txt
return content
def find_NE(self, filename):
src_path = self.src_dir + '/' + filename
res_path = self.res_dir + '/' + filename
txt = rw.readFile(src_path)
content = ""
for key in self.dic:
txt = self.find_key(key,txt)
p = re.compile('.*\[.+\].*')
lines = p.findall(txt)
for line in lines:
content += (line + '\n')
rw.writeFile(res_path,content)
return content
def analyze(self):
files = os.listdir(self.src_dir)
for filename in files:
path = self.src_dir + '/' + filename
if os.path.isdir(path) == False:
txt = rw.readFile(path)
segs = self.ch.findall(txt)
for seg in segs:
words = self.get_words(seg)
for word in words:
if (word in self.dictionary):
self.dictionary[word] += 1
else:
self.dictionary[word] = 1
return self.dictionary
def sub(self, filename):
print "********************"
print "Substituting File: %s" % filename
src_path = self.src_dir + '/' + filename
res_path = self.res_dir + '/' + filename
txt = pre.sort_txt(rw.readFile(src_path), [0, 1], 20)
query_log = txt.split('\n')[0:self.num]
content = ""
for line in query_log:
temp = line.split('\t')
substituted = self.run(temp[0])
if substituted != "":
print "Substituting: %s" % temp[0]
query = temp[0] + '\t' + substituted + '\t' + temp[1] + '\n'
content += query
rw.writeFile(res_path, content)
def sub(self,filename):
print "********************"
print "Substituting File: %s" % filename
src_path = self.src_dir + '/' + filename
res_path = self.res_dir + '/' + filename
txt = pre.sort_txt(rw.readFile(src_path),[0,1],20)
query_log = txt.split('\n')[0:self.num]
content = ""
for line in query_log:
temp = line.split('\t')
substituted = self.run(temp[0])
if substituted != "":
print "Substituting: %s" % temp[0]
query = temp[0] + '\t' + substituted + '\t' + temp[1] + '\n'
content += query
rw.writeFile(res_path,content)
def segment(self,filename):
print "********************"
print "Segmenting File: %s" % filename
src_path = self.src_dir + '/' + filename
res_path = self.res_dir + '/' + filename
query_log = rw.readFile(src_path).split('\n')[0:self.num]
content = ""
for line in query_log:
temp = line.split('\t')
#print "Segmenting: %s" % temp[0]
segmented = self.run(temp[0])
if segmented != "":
print "Segmenting: %s" % temp[0]
query = temp[0] + '\t' + segmented + '\t' + temp[1] + '\n'
content += query
rw.writeFile(res_path,content)
def sort_file(self,filename):
src_path = self.src_dir + "/" + filename
tar_path = self.tar_dir + "/" + filename
content = rw.readFile(src_path)
query_list = content.split("\n")
#constants
num = len(query_list)
num_sorted = 0
freq = 0
freq_sorted = 0
query_list_sorted = []
content_sorted = ""
# sort the queries over min_char
for query in query_list:
temp = query.split("\t")
if len(temp) > 1:
freq += int(temp[-1])
if len(temp[0]) >= self.min_char:
query_list_sorted.append(temp[0] + "\t" + temp[-1])
freq_sorted += int(temp[-1])
for target in self.targets:
index = self.targets.index(target)
count = (temp[0].count(target) > 0) * 1
self.target_num_l[index] += count#temp[0].count(target)
self.target_freq_l[index] += count*int(temp[-1])
num_sorted = len(query_list_sorted)
self.total_num += num
self.total_num_sorted += num_sorted
self.total_freq += freq
self.total_freq_sorted += freq_sorted
result = ("Query Log: %s\n" % filename) \
+ ("Number of queries: %d\n" % num) \
+ ("Queries over %d bytes: %d\n" % (self.min_char,num_sorted)) \
+ ("Queries frequency: %d\n" % freq) \
+ ("Long queries frequency: %d\n" % freq_sorted) \
+ (("Long query ratio: %0.3f\n") % (float(num_sorted)/num)) \
+ (("Long query frequency ratio: %0.3f\n") % (float(freq_sorted)/freq))
# recombine the sorted queries
for query in query_list_sorted:
content_sorted += (query + "\n")
rw.writeFile(tar_path,content_sorted)
return result
def sort_file(src_dir, res_dir, filename, cols, min_f):
print "Pre-processing file: %s" % filename
src_path = src_dir + '/' + filename
res_path = res_dir + '/' + filename
content = rw.readFile(src_path)
query_list = content.split('\n')[0:-1]
#constants
query_list_sorted = []
content_sorted = ""
for query in query_list:
temp = query.split("\t")
freq = int(temp[cols[-1]])
if freq >= min_f:
line = ""
for i in xrange(len(cols)-1):
line += (temp[cols[i]] + '\t')
line += temp[cols[-1]]
content_sorted += (line + '\n')
rw.writeFile(res_path,content_sorted)
return content_sorted
def run(self):
#define the server thread
self.s = socket.socket()
self.host = socket.gethostname()
self.port = 2222 + self.pid * 10
self.s.bind((self.host, self.port))
self.s.listen(5)
while True:
#listen for connection
c, addr = self.s.accept()
print '\nGot connection from ', addr
#initialize file index
original_index = self.fileindex('original')
download_index = self.fileindex('download')
index = original_index + download_index
#receive message
msg = c.recv(1024)
#parse & analyze message
mid = msg.split(',')[0].split(':')[1]
action = msg.split(',')[1].split(':')[0]
name = msg.split(',')[1].split(':')[1]
if action == 'search':
print 'search ' + name
ttl = msg.split(',')[2].split('=')[1]
#read mid_list from file 'msg_list'
mid_list = rw.readList('msg_list.txt')
#decide broadcast or not
if int(ttl) == 1 or mid in mid_list:
print 'file dont need to pass'
else:
#update mid_list in file 'msg_list'
mid_list.append(mid)
rw.write('msg_list.txt', mid_list)
#decrease the ttl value by 1 after broadcast for once
rmsg = msg.rsplit(ttl, 1)
msg = str(int(ttl)-1).join(rmsg)
#start autobroadcast thread (details in autobroadcast.py module)
broadcast = autobroadcast.Auto(msg, self.neighbor)
broadcast.start()
broadcast.join()
if name in index:
#check file's state
meta_dict = rw.readDict('metadata.txt')
if (name in original_index and (meta_dict['original'][name]['state'] == 'valid' or meta_dict['download'][name]['state'] == 'valid')):
#start hitresponse thread (details in hitresponse.py module)
msg = 'mid:' + str(self.pid) + '|' + str(self.port) + ',response:' + name
hit = hitresponse.Hit(msg, mid)
hit.start()
hit.join()
else:
'file state is not qualified'
else:
print 'no file match, pass to neighbors'
elif action == 'update':
print 'update ' + name
#read mid_list from file 'msg_list'
mid_list = rw.readList('msg_list.txt')
#decide broadcast or not
if mid in mid_list:
print 'update dont need to pass'
else:
#update mid_list in file 'msg_list'
mid_list.append(mid)
rw.write('msg_list.txt', mid_list)
#start autobroadcast thread (details in autobroadcast.py module)
broadcast = autobroadcast.Auto(msg, self.neighbor)
broadcast.start()
broadcast.join()
if name in download_index:
#set file state to invalid
meta_dict = rw.readDict('metadata.txt')
meta_dict['download'][name]['state'] = 'invalid'
rw.write('metadata.txt', meta_dict)
else:
print 'no file need to update, pass to neighbors'
elif action == 'check':
print 'check ' + name
#check the original file's version
version = int(msg.split(',')[2].split(':')[1])
TTR = int(msg.split(',')[3].split(':')[1])
meta_dict = rw.readDict('metadata.txt')
if meta_dict['original'][name]['version'] == version:
#if same version, send a new TTR.
newTTR = 2 * TTR
msg = 'mid:' + str(self.pid) + '|' + str(self.port) + ',checkresponse:' + name + ',state:valid,TTR:' + str(newTTR)
else:
#send Invalid & new version exist.
msg = 'mid:' + str(self.pid) + '|' + str(self.port) + ',checkresponse:' + name + ',state:invalid'
hit = hitresponse.Hit(msg, mid)
hit.start()
hit.join()
elif action == 'checkresponse':
print 'checkresponse ' + name
#update the metadata
state = str(msg.split(',')[2].split(':')[1])
if state == 'valid':
TTR = int(msg.split(',')[3].split(':')[1])
#change TTR to 2TTR. (details in pullrenew.py module)
pullrenew.renew('TTR', name, TTR)
#change state from 'TTR Expired' back to 'valid' (details in pullrenew.py module)
pullrenew.renew('state', name, state)
elif state == 'invalid':
#change state from 'TTR Expired' back to 'invalid' (details in pullrenew.py module)
pullrenew.renew('state', name, state)
elif action == 'response':
#start reconnect thread (details in reconnect.py module)
print 'response ' + name + ' from ' + mid
#read name_list from file 'req_list'
name_list = rw.readList('req_list.txt')
#decide reconnect or not
if name in name_list:
msg = 'mid:' + str(self.pid) + '|' + str(self.port) + ',obtain:' + name
connect = reconnect.Connect(msg, mid, name)
connect.start()
connect.join()
else:
print 'file has been obtained'
elif action == 'obtain':
print 'obtain ' + name
#start transfer directly to original peer
if name in original_index:
path = os.path.join(os.getcwd(), 'files', 'original', name)
metadata = rw.readDict('metadata.txt')['original'][name]
elif name in download_index:
path = os.path.join(os.getcwd(), 'files', 'download', name)
metadata = rw.readDict('metadata.txt')['download'][name]
#send metadata & file
c.sendall(json.dumps(metadata))
#slice file into chunks by buffer
content = rw.readFile(path)
i = 0
while i <= len(content):
chunk = buffer(content, i, 1024)
c.sendall(chunk)
i += 1024
print 'send file: ' + name
c.close()
def init_dict(self):
words = rw.readFile(self.dict_dir).split('\n')[0:-1]
for word in words:
self.dictionary.add(word)
def init_dic(self):
lines = rw.readFile(self.dic_dir).split('\n')[0:-1]
for line in lines:
temp = line.split('\t')
self.dictionary[temp[0]] = temp[1]
return self.dictionary
def init_dic(self):
lines = rw.readFile(self.dic_dir).split('\n')[0:-1]
for line in lines:
temp = line.split('\t')
self.dictionary[temp[0]] = temp[1]
return self.dictionary
def init_dict(self):
words = rw.readFile(self.dict_dir).split('\n')[0:-1]
for word in words:
self.dictionary.add(word)
def init_model(self):
lines = rw.readFile(self.src_dir).split('\n')[0:-1]
for line in lines:
temp = line.split('\t')
self.models.append(temp[0])
def init_model(self):
lines = rw.readFile(self.src_dir).split('\n')[0:-1]
for line in lines:
temp = line.split('\t')
self.models.append(temp[0])