class DoubleBloom:
def __init__(self, values, probability):
self.zero_bloom = BloomFilter(values, probability, False)
self.one_bloom = BloomFilter(values, probability, False)
self.next_level = None
def insert(self, key, value):
if value == '0':
self.zero_bloom.insert(key)
else:
self.one_bloom.insert(key)
def get_value(self, key):
if self.zero_bloom.contains(key):
if self.one_bloom.contains(key):
if self.next_level is not None:
return self.next_level.get_value(key)
return 'Both'
return '0'
elif self.one_bloom.contains(key):
return '1'
return None
def add_level(self, values, probability):
self.next_level = DoubleBloom(values, probability)
def test_bloom_filter():
with pytest.assertRaises(ValueError):
blf = BloomFilter(n=-1, p=0.1)
with pytest.assertRaises(ValueError):
blf = BloomFilter(n=1, p=-0.1)
with pytest.assertRaises(ValueError):
blf = BloomFilter(n=1.0, p=0.1)
# aiming to cause collision
blf = BloomFilter(n=3, p=0.1)
blf.insert("Gondor")
assert blf.is_present("Gondor 1") == True
assert blf.is_present("Isenguard") == False
class Node:
def __init__(self, k, expected_num, fp_prob):
"""
Represents a single node of Bloom Tree
"""
self.children: List[Node] = []
self.parent: Optional[Node] = None
self.filter = BloomFilter(expected_num, fp_prob)
self.dataset_id: Optional[str] = None
self.k = k
def populate_dataset_info(self, dataset: List[Read]) -> None:
self.dataset_id = dataset[0].filename
self.insert_kmers_from_dataset(dataset)
def insert_kmers_from_dataset(self, dataset: List[Read]) -> None:
for read in dataset:
for kmer in read.kmers(self.k):
self.filter.insert(kmer)
def add_node_kmers(self, other: 'Node') -> None:
self.filter.filter |= other.filter.filter
def num_children(self) -> int:
return len(self.children)
def score(self, other: 'Node') -> int:
"""
"Hamming distance" score where lower is better
:param other: The node to compare against
"""
return count_xor(self.filter.filter, other.filter.filter)
def get_size(self):
"""
Returns the total number of bytes occupied by the filter object
"""
return (sys.getsizeof(self.children) + sys.getsizeof(self.parent) +
sys.getsizeof(self.dataset_id) + sys.getsizeof(self.k) +
self.filter.get_size())
def test_can_insert(self):
bloom = BloomFilter(2000, 4)
bloom.insert(5)
def test_inserted_is_probably_contained(self):
bloom = BloomFilter(2000, 4)
bloom.insert(42)
self.assertTrue(42 in bloom)
class TwitterApiSpider(object):
# 初始化，连接数据库，建表
def __init__(self, seed):
self.db = pymysql.connect(host=host,
port=3306,
user='******',
passwd='fit4-305',
db=db,
charset='utf8mb4')
self.cursor = self.db.cursor()
self.seed = seed
self.api_number = 0 #记录API顺序
self.location = 0 #记录checkpoints位置
self.api_error_times = 0 #记录API连续error次数
self.bloom_filter = BloomFilter(name=name,
length=length,
number=number,
save_frequency=save_frequency)
# checkpoints_id表
sql = 'create table if not exists checkpoints_id(' \
'checkpoints bigint not null)' \
'ENGINE=InnoDB DEFAULT CHARSET=utf8mb4'
self.cursor.execute(sql)
print('建立checkpoints_id表')
# checkpoints_info表
sql = 'create table if not exists checkpoints_info(' \
'checkpoints bigint not null)' \
'ENGINE=InnoDB DEFAULT CHARSET=utf8mb4'
self.cursor.execute(sql)
print('建立checkpoints_info表')
# user_id表
sql = 'create table if not exists user_id(' \
'id bigint not null auto_increment,' \
'userid bigint not null,' \
'primary key(id))' \
'ENGINE=InnoDB DEFAULT CHARSET=utf8mb4'
self.cursor.execute(sql)
print('建立user_id表')
sql = 'create table if not exists user_info(' \
'id bigint not null auto_increment,' \
'userid bigint not null,' \
'username varchar(64) not null,' \
'userscreenname varchar(64) not null,' \
'description mediumtext,' \
'createat timestamp not null default "1970-01-02 00:00:00",' \
'url mediumtext,' \
'profileimageurl mediumtext,' \
'profilebackgroundimageurl mediumtext,' \
'location mediumtext,' \
'timezone mediumtext,' \
'accesslevel bigint not null,' \
'statuscount bigint not null,' \
'followerscount bigint not null,' \
'friendscount bigint not null,' \
'favouritescount bigint not null,' \
'listedcount bigint not null,' \
'isprotected tinyint not null,' \
'isgeoenabled tinyint not null,' \
'isshowallinlinemedia tinyint not null,' \
'iscontributorsenable tinyint not null,' \
'isfollowrequestsent tinyint not null,' \
'isprofilebackgroundtiled tinyint not null,' \
'isprofileusebackgroundtiled tinyint not null,' \
'istranslator tinyint not null,' \
'isverified tinyint not null,' \
'vtcoffset bigint,' \
'lang varchar(64) default "en",' \
'biggerprofileimageurl mediumtext,' \
'biggerprofileimageurlhttps mediumtext,' \
'miniprofileimageurl mediumtext,' \
'miniprofileimageurlhttps mediumtext,' \
'originalprofileimageurl mediumtext,' \
'originalprofileimageurlhttps mediumtext,' \
'profilebackgroundimageurlhttps mediumtext,' \
'profilebanneripadurl mediumtext,' \
'profilebanneripadretinaurl mediumtext,' \
'profilebannermobileurl mediumtext,' \
'profilebannermobileretinaurl mediumtext,' \
'profilebannerretinaurl mediumtext,' \
'profilebannerurl mediumtext,' \
'profileimageurlhttps mediumtext,' \
'updatetime timestamp not null default now(),' \
'sensitivity float not null,' \
'sensitivity2 float not null,' \
'primary key(id))ENGINE=InnoDB default CHARSET=utf8mb4'
self.cursor.execute(sql)
print('建立user_info表')
# 将seed加入到user_id中,如果user_id中没有内容
sql = 'select userid from user_id where id = 1'
self.check_status()
self.cursor.execute(sql)
if self.cursor.fetchall():
pass
else:
sql = 'insert into user_id (userid) value (%s)' % self.seed
self.check_status()
self.cursor.execute(sql)
self.db.commit()
# 初始化checkpoints_id/info
sql = 'select * from checkpoints_id'
self.check_status()
self.cursor.execute(sql)
if self.cursor.fetchall():
pass
else:
sql = 'insert into checkpoints_id (checkpoints) value (0)'
self.check_status()
self.cursor.execute(sql)
self.db.commit()
sql = 'select * from checkpoints_info'
self.check_status()
self.cursor.execute(sql)
if self.cursor.fetchall():
pass
else:
sql = 'insert into checkpoints_info (checkpoints) value (0)'
self.check_status()
self.cursor.execute(sql)
self.db.commit()
# 随机返回api
def get_random_api(self):
# 从tokens.txt中读取需要用到的密钥
file = open('tokens.txt', 'r')
tokens = []
for line in file:
tokens.append(line.split(','))
random_int = self.api_number % len(tokens)
print('使用app_key序号：' + str(random_int))
self.api_number = self.api_number + 1
consumer_key = tokens[random_int][0]
consumer_secret = tokens[random_int][1]
access_token = tokens[random_int][2]
access_token_secret = tokens[random_int][3][0:-1]
auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
auth.set_access_token(access_token, access_token_secret)
return tweepy.API(auth,
proxy=proxy,
wait_on_rate_limit=True,
wait_on_rate_limit_notify=True)
# 根据id，获取其朋友和粉丝，加入到user_id中，此函数用户中文用户id爬取
def search_id_chinese(self, id):
# 获取粉丝id
cursor = -1
while cursor != 0:
api = self.get_random_api()
try:
result = api.followers_ids(user_id=id,
cursor=cursor,
count=5000)
self.api_error_times = 0 #成功使用API，将错误次数计数清零
except:
self.api_error_times = self.api_error_times + 1
if self.api_error_times % 10 == 0: #连续10次出错，则休息300s，并打印提示信息
self.api_error_times = 0
print(
'something goes wrong with followers_ids api, we will retry it in 300s'
)
time.sleep(300)
else:
followers_ids = result[0]
for follower_id in followers_ids:
if self.bloom_filter.is_contain(
str(follower_id)): #判断是否在BF中，是则跳过，不是则进一步判断是不是中国人
pass
else:
self.bloom_filter.insert(
str(follower_id)) #不论是不是中国人都要将其插入BF，这样可以减少后面重复判断
if self.is_chinese(follower_id):
sql = 'insert into user_id (userid) value (%s)' % follower_id
self.check_status()
self.cursor.execute(sql)
else:
pass
cursor = result[1][1] #获得新的cursor
self.db.commit()
# 处理朋友
cursor = -1
while cursor != 0:
api = self.get_random_api()
try:
result = api.friends_ids(user_id=id, cursor=cursor, count=5000)
self.api_error_times = 0
except:
self.api_error_times = self.api_error_times + 1
if self.api_error_times % 10 == 0:
self.api_error_times = 0
print(
'something goes wrong with followers_ids api, we will retry it in 300s'
)
time.sleep(300)
else:
friends_ids = result[0]
for friend_id in friends_ids:
# sql = 'select * from searched_list where userid = %s limit 1' % friend_id
# self.cursor.execute(sql)
if self.bloom_filter.is_contain(str(friend_id)):
pass
else:
self.bloom_filter.insert(str(friend_id))
if self.is_chinese(friend_id):
sql = 'insert into user_id (userid) value (%s)' % friend_id
self.check_status()
self.cursor.execute(sql)
else:
pass
cursor = result[1][1]
self.db.commit()
# 此函数用户全网用户id爬取，不会判断是否是中国人
def search_id(self, id):
# 获取粉丝
cursor = -1
while cursor != 0:
api = self.get_random_api()
try:
result = api.followers_ids(user_id=id,
cursor=cursor,
count=5000)
self.api_error_times = 0
except:
self.api_error_times = self.api_error_times + 1
if self.api_error_times % 10 == 0:
self.api_error_times = 0
print(
'something goes wrong with followers_ids api, we will retry it in 300s'
)
time.sleep(300)
else:
followers_ids = result[0]
for follower_id in followers_ids:
# sql = 'select * from searched_list where userid = %s limit 1' % follower_id
# self.cursor.execute(sql)
if self.bloom_filter.is_contain(str(follower_id)):
pass
else:
sql = 'insert into user_id (userid) value (%s)' % follower_id
self.check_status()
self.cursor.execute(sql)
self.bloom_filter.insert(str(follower_id))
cursor = result[1][1]
self.db.commit()
# 获取朋友
cursor = -1
while cursor != 0:
api = self.get_random_api()
try:
result = api.friends_ids(user_id=id, cursor=cursor, count=5000)
self.api_error_times = 0
except:
self.api_error_times = self.api_error_times + 1
if self.api_error_times % 10 == 0:
self.api_error_times = 0
print(
'something goes wrong with followers_ids api, we will retry it in 300s'
)
time.sleep(300)
else:
friends_ids = result[0]
for friend_id in friends_ids:
# sql = 'select * from searched_list where userid = %s limit 1' % friend_id
# self.cursor.execute(sql)
if self.bloom_filter.is_contain(str(friend_id)):
pass
else:
sql = 'insert into user_id (userid) value (%s)' % friend_id
self.check_status()
self.cursor.execute(sql)
self.bloom_filter.insert(str(friend_id))
cursor = result[1][1]
self.db.commit()
# 根据user_id获取用户信息,
def get_user_info(self):
print('开始获取用户信息')
sql = 'select checkpoints from checkpoints_info'
self.check_status()
self.cursor.execute(sql)
result = self.cursor.fetchall()
if result:
if result == -1:
print('已经获取了全部user_info')
return
else:
self.location = result[0][0]
else:
self.location = self.location
flag = 1
while flag:
api = self.get_random_api()
sql = 'select userid from user_id limit %s, 100' % self.location # 每次处理100个用户
self.check_status()
self.cursor.execute(sql)
result = self.cursor.fetchall()
if result:
userids = self.tuple_to_list(result)
try:
users_info = api.lookup_users(user_ids=userids)
self.api_error_times = 0
except:
self.api_error_times = self.api_error_times + 1
if self.api_error_times % 10 == 0:
self.api_error_times = 0
print(
'something goes wrong with lookup_users api, we will retry it in 300s'
)
time.sleep(300)
else:
for user_info in users_info:
info = user_info._json
userid = info['id']
username = info['name']
userscreenname = info['screen_name']
description = info['description'].replace('\"', '')
createat = info['created_at']
createat_timestamp = datetime.datetime.strptime(
createat, '%a %b %d %H:%M:%S +0000 %Y')
createat = createat_timestamp.strftime(
'%Y-%m-%d %H:%M:%S')
url = info['url']
profileimageurl = info['profile_image_url']
profilebackgroundimageurl = info[
'profile_background_image_url']
location = info['location']
timezone = info['time_zone']
accesslevel = 0
statuscount = info['statuses_count']
followerscount = info['followers_count']
friendscount = info['friends_count']
favouritescount = info['favourites_count']
listedcount = info['listed_count']
isprotected = info['protected']
isgeoenabled = info['geo_enabled']
isshowallinlinemedia = False
iscontributorsenable = info['contributors_enabled']
isfollowrequestsent = info['follow_request_sent']
isprofilebackgroundtiled = info[
'profile_background_tile']
isprofileusebackgroundtiled = False
istranslator = info['is_translator']
isverified = info['verified']
vtcoffset = info['utc_offset']
lang = info['lang']
biggerprofileimageurl = info['profile_image_url']
biggerprofileimageurlhttps = 'NULL'
miniprofileimageurl = 'NULL'
miniprofileimageurlhttps = 'NULL'
originalprofileimageurl = 'NULL'
originalprofileimageurlhttps = 'NULL'
profilebackgroundimageurlhttps = info[
'profile_background_image_url_https']
profilebanneripadurl = 'NULL'
profilebanneripadretinaurl = 'NULL'
profilebannermobileurl = 'NULL'
profilebannermobileretinaurl = 'NULL'
profilebannerretinaurl = 'NULL'
profilebannerurl = 'NULL'
profileimageurlhttps = info['profile_image_url_https']
updatetime = datetime.date.today()
sensitivity = 0
sensitivity2 = 0
sql = 'insert into user_info (userid, username, userscreenname, description, createat, url, ' \
'profileimageurl, profilebackgroundimageurl, location, timezone, accesslevel, statuscount, ' \
'followerscount, friendscount, favouritescount, listedcount, isprotected, isgeoenabled, ' \
'isshowallinlinemedia, iscontributorsenable, isfollowrequestsent, isprofilebackgroundtiled, ' \
'isprofileusebackgroundtiled, ' \
'istranslator, isverified, vtcoffset, lang, biggerprofileimageurl, biggerprofileimageurlhttps, ' \
'miniprofileimageurl, miniprofileimageurlhttps, originalprofileimageurl, originalprofileimageurlhttps, ' \
'profilebackgroundimageurlhttps, profilebanneripadurl, profilebanneripadretinaurl, ' \
'profilebannermobileurl, profilebannermobileretinaurl, profilebannerretinaurl, profilebannerurl, ' \
'profileimageurlhttps, updatetime, sensitivity, sensitivity2) value (%s, "%s", "%s", "%s", "%s", "%s", "%s", ' \
'"%s", "%s", "%s", %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, "%s", "%s", %s, %s, %s, %s, ' \
'%s, "%s", %s, %s, %s, %s, %s, %s, "%s", "%s", %s, %s)' % (
userid, username, userscreenname, description, createat, url, profileimageurl,
profilebackgroundimageurl, location, timezone, accesslevel,
statuscount, followerscount, friendscount, favouritescount, listedcount, isprotected,
isgeoenabled, isshowallinlinemedia,
iscontributorsenable, isfollowrequestsent, isprofilebackgroundtiled,
isprofileusebackgroundtiled,
istranslator, isverified, vtcoffset,
lang, biggerprofileimageurl, biggerprofileimageurlhttps, miniprofileimageurl,
miniprofileimageurlhttps, originalprofileimageurl,
originalprofileimageurlhttps, profilebackgroundimageurlhttps, profilebanneripadurl,
profilebanneripadretinaurl, profilebannermobileurl,
profilebannermobileretinaurl, profilebannerretinaurl, profilebannerurl, profileimageurlhttps,
updatetime, sensitivity, sensitivity2)
sql = sql.replace('\"None\"',
'NULL').replace('None', 'NULL')
self.check_status()
# 这里用try和except是为了防止sql语句格式错误导致程序崩溃。毕竟我也很难预料到会出现什么奇怪的字符！
try:
self.cursor.execute(sql)
except:
print(sql)
self.db.commit()
self.location = self.location + 100
self.save_checkpoints_info(self.location)
else:
flag = 0
self.save_checkpoints_info(-1)
print('get all users info, all jobs done')
# 获取所有的user_id
def get_user_id(self):
# 从checkpoints中读取出数据
sql = 'select checkpoints from checkpoints_id'
self.check_status()
self.cursor.execute(sql)
result = self.cursor.fetchall()
if result:
if result == -1:
print('已经获取了全部user_id')
return
else:
self.location = result[0][0]
else:
self.location = self.location
# 开始迭代遍历user_id中的用户
flag = 1
while flag:
sql = 'select userid from user_id limit %s, 1' % self.location
self.check_status()
self.cursor.execute(sql)
result = self.cursor.fetchall()
if result:
userid = result[0][0]
print('将要处理id:' + str(userid))
if functions == 'whole_net_chinese':
self.search_id_chinese(userid)
if functions == 'whole_net':
self.search_id(userid)
self.location = self.location + 1
self.save_checkpoints_id(self.location)
else:
flag = 0
self.save_checkpoints_id(-1)
print('已经获取了全部user_id')
# 判断是否是中国人
def is_chinese(self, id):
# 使用正则表达式判断文本内容中是否有中文字符，python3 的str本身就是unicode编码，所以不用decode()
# 判断是否是中文的原则就是，最近的10条推文中，出现过某条推文中包含中文但是不包含日文，因为日文中某些字与中文一致
pattern_zh = re.compile(u'[\u4e00-\u9fa5]+')
pattern_ja_ka = re.compile(u'[\u30a0-\u30ff]+')
pattern_ja_hi = re.compile(u'[\u3040-\u309f]+')
api = self.get_random_api()
try:
tweets = api.user_timeline(id=id, count=10)
self.api_error_times = 0
for tweet in tweets:
text = tweet.text
# print(tweet.text)
if pattern_zh.search(text) and not pattern_ja_hi.search(
text) and not pattern_ja_ka.search(text):
print(text)
print('包含中文，不包含日文')
return True
else:
pass
except:
self.api_error_times = self.api_error_times + 1
if self.api_error_times % 10 == 0:
self.api_error_times = 0
print(
'failed to get the tweets, so we take it as no chinese!, but we will take 300s for break.'
)
time.sleep(300)
return False
# 保存id的checkpoints
def save_checkpoints_id(self, num):
sql = 'update checkpoints_id set checkpoints = %s' % num
self.check_status()
self.cursor.execute(sql)
self.db.commit()
# 保存info的checkpoints
def save_checkpoints_info(self, num):
sql = 'update checkpoints_info set checkpoints = %s' % num
self.check_status()
self.cursor.execute(sql)
self.db.commit()
# tuple 转 list
def tuple_to_list(self, tuple):
list = []
for i in tuple:
list.append(i[0])
return list
# check status
def check_status(self):
try:
self.db.ping()
except:
print("lost connection to database server, we need to reconnect!")
self.db = pymysql.connect(host=host,
port=3306,
user='******',
passwd='fit4-305',
db=db,
charset='utf8')
self.cursor = self.db.cursor()
class WebParser(object):
def __init__(self, redis_key):
self.bloom_filter = BloomFilter(
redis.StrictRedis(host='localhost', port=6379), redis_key)
def list_zhilian(self, response):
urls = []
page = json.loads(response.text)
if not len(page['data']['results']):
return None
for info in page['data']['results']:
url = info.get('positionURL')
if not self.bloom_filter.exists(url):
self.bloom_filter.insert(url)
urls.append(url)
return urls
def list_qiancheng(self, response):
response.encoding = response.apparent_encoding
selector = Selector(response, type='html')
list = selector.xpath('//div[@class="el"]/p/span/a/@href').getall()
urls = self._filter_list(list)
return urls
def list_liepin(self, response):
selector = Selector(response, type='html')
list = selector.xpath('//span[@class="job-name"]/a/@href').getall()
urls = self._filter_list(list)
return urls
def list_boss(self):
pass
def list_shixi(self):
pass
def list_lagou(self):
pass
def content_zhilian(self, response, database, url):
response.encoding = response.apparent_encoding
selector = Selector(response)
title = selector.xpath(
'//*[@class="summary-plane__title"]/text()').get()
salary = selector.xpath(
'//*[@class="summary-plane__salary"]/text()').get()
city = selector.xpath(
'//*[@class="summary-plane__info"]/li/a/text()').get()
company = selector.xpath('//a[@class="company__title"]/text()').get()
company_url = selector.xpath(
'//a[@class="company__title"]/@href').get()
# description 结构上比较混乱,先爬取再说
description = selector.xpath(
'//*[@class="describtion__detail-content"]').getall()
summary_info = selector.xpath(
'//*[@class="summary-plane__info"]/li/text()').getall()
if len(summary_info) == 3:
experience = summary_info[0]
education = summary_info[1]
data = {
'url': url,
"title": title,
"salary": salary,
"city": city,
"company": company,
"company_url": company_url,
"experience": experience,
"education": education,
"description": description
}
else:
data = {
'url': url,
"title": title,
"salary": salary,
"city": city,
"company": company,
"company_url": company_url,
"summary_info": summary_info,
"description": description
}
database.insert(data)
def content_qiancheng(self, response, database, url):
response.encoding = response.apparent_encoding
selector = Selector(response)
title = selector.xpath('//div[@class="cn"]/h1/text()').get()
salary = selector.xpath('//div[@class="cn"]/strong/text()').get()
company = selector.xpath('//div[@class="com_msg"]/a/p/text()').get()
company_url = selector.xpath('//div[@class="com_msg"]/a/@href').get()
# description 结构上比较混乱,先爬取再说
description = selector.xpath(
'//div[@class="bmsg job_msg inbox"]').getall()
summary_info = city = selector.xpath(
'//p[@class="msg ltype"]/@title').get()
if summary_info:
summary_info = summary_info.split('\xa0\xa0|\xa0\xa0')
if len(summary_info) >= 3:
city = summary_info[0]
experience = summary_info[1]
education = summary_info[2]
data = {
'url': url,
"title": title,
"salary": salary,
"city": city,
"company": company,
"company_url": company_url,
"experience": experience,
"education": education,
"description": description
}
else:
data = {
'url': url,
"title": title,
"salary": salary,
"company": company,
"company_url": company_url,
"summary_info": summary_info,
"description": description
}
database.insert(data)
def content_liepin(self, response, database, url):
selector = Selector(response)
title = selector.xpath('//h1[@title]/text()').get()
salary = selector.xpath('//p[@class="job-item-title"]/text()').get()
if salary:
salary = salary.strip()
city = selector.xpath('//p[@class="basic-infor"]/span/a/text()').get()
if city:
city = city.split('-')[0]
company = selector.xpath(
'//div[@class="company-logo"]/p/a/text()').get()
company_url = selector.xpath(
'//div[@class="company-logo"]/p/a/@href').get()
# description 结构上比较混乱,先爬取再说
description = selector.xpath(
'//div[@class="job-item main-message job-description"]').getall()
summary_info = selector.xpath(
'//div[@class="job-qualifications"]/span/text()').getall()
if len(summary_info) >= 2:
education = summary_info[0]
experience = summary_info[1]
data = {
'url': url,
"title": title,
"salary": salary,
"city": city,
"company": company,
"company_url": company_url,
"experience": experience,
"education": education,
"description": description
}
else:
data = {
'url': url,
"title": title,
"salary": salary,
"city": city,
"company": company,
"company_url": company_url,
"summary_info": summary_info,
"description": description
}
database.insert(data)
def content_boss(self):
pass
def content_shixi(self):
pass
def content_lagou(self):
pass
def _filter_list(self, list):
urls = []
for url in list:
if not self.bloom_filter.exists(url):
self.bloom_filter.insert(url)
urls.append(url)
print('done one page')
return urls
