class CheckQualificationByRankList():
def __init__(self):
self.get_raw_data = GetRawData()
self.redis_client = RedisClient()
self.live_user_list = []
self.room_id_list = []
self.stupid_key_words = STUPID_KEY_WORDS
def get_users(self):
users = self.redis_client.get_users() #每次获取分数最低的10000个
return users
def is_live_user(self, sec_user_id):
return self.redis_client.is_live_user(sec_user_id)
def save_rooms(self):
for each in self.room_id_list:
self.redis_client.add_rooms(each)
def add_to_live_users(self):
for each in self.live_user_list:
self.redis_client.add_live_users(each, 1)
def increase_user_score(self, sec_user_id):
self.redis_client.increase_user_score(sec_user_id)
def is_qualified_user(self, user):
nickname = user.get('nickname')
for word in self.stupid_key_words:
if word in nickname:
return False
return True
def get_rank_list(self, sec_user_id):
if not self.is_live_user(sec_user_id):
try:
raw_data = self.get_raw_data.get_rank_list(sec_user_id)
except Exception as e:
logger.error('get_rank_list出错-' + e.args[0] + '-sec_user_id-' + sec_user_id)
return None
try:
user = raw_data.get('data').get('anchor_info').get('user')
except Exception as e:
logger.error('parse_current_room出错' + e.args[0] + '-sec_user_id-' + sec_user_id)
return None
if self.is_qualified_user(user):
own_room = user.get('own_room')
if own_room: #如果有这个，说明直播开始了
room_id = own_room.get('room_ids_str')[0]
self.live_user_list.append(sec_user_id)
self.room_id_list.append(room_id)
self.increase_user_score(sec_user_id)
else:
self.redis_client.delete_users(sec_user_id)
logger.info('删除user-sec_user_id-' + sec_user_id)
else:
self.increase_user_score(sec_user_id)
"""
try:
raw_data = self.get_raw_data.get_rank_list(sec_user_id)
except Exception as e:
logger.error('get_rank_list出错-' + e.args[0] + '-sec_user_id-' + sec_user_id)
return None
try:
nickname = raw_data.get('data').get('anchor_info').get('user').get('nickname')
print(nickname)
except Exception as e:
logger.error('parse_current_room出错' + e.args[0] + '-sec_user_id-' + sec_user_id)
return None
"""
def run(self):
users = self.get_users()
batch_size = 50 #这个接口80个一批可行的（总共近4000个），更多的我不敢再试了
for batch_limit in range(0, len(users), batch_size):
start = batch_limit
stop = min(batch_limit+batch_size, len(users))
logger.info('当前获取用户序号-' + str(start+1) + '-' + str(stop))
tasks = [gevent.spawn(self.get_rank_list, sec_user_id) for sec_user_id in users[start:stop]]
gevent.joinall(tasks)
self.save_rooms()
self.add_to_live_users()
logger.info('新增room_id-' + str(len(self.room_id_list)))
self.room_id_list.clear()
self.live_user_list.clear()
class SaveLiveUsers():
def __init__(self):
self.get_raw_data = GetRawData()
#self.db = pymysql.connect(host='47.114.166.130', port=13306, user='******', password='******', db='bxdb', charset='utf8mb4')
#self.cursor = self.db.cursor()
self.db = pymysql.connect(host='localhost', port=3306, user='******', password='******', db='bxmind', charset='utf8mb4')
self.cursor = self.db.cursor()
self.mysql_client = MysqlClient()
self.redis_client = RedisClient()
self.stupid_key_words = STUPID_KEY_WORDS
self.a_list = []
self.b_list = []
def into_mysql(self, data, table):
keys = ','.join(data.keys())
values = ','.join(['%s'] * len(data))
sql = 'insert into %s (%s) values (%s)' %(table, keys, values)
try:
self.cursor.execute(sql, tuple(data.values()))
self.db.commit()
except Exception as e:
print(e.args)
def is_qualified(self, nickname):
for word in self.stupid_key_words:
if word in nickname:
return False
return True
def run_a(self):
table = 'dy_live_lives'
with open('lives_20200614.txt', 'r', encoding='utf-8') as f:
lines = f.readlines()
for line in lines:
#print(line)
data = json.loads(line)
nickname = data.get('nickname')
sec_user_id = data.get('sec_user_id')
if data.get('status') == 4:
if self.is_qualified(nickname):
self.into_mysql(data, table)
else:
self.redis_client.delete_users(sec_user_id)
self.redis_client.delete_live_users(sec_user_id)
def replicate_table(self):
sql = 'CREATE TABLE dy_sample LIKE dy_live_lives'
self.cursor.execute(sql)
def select_users(self):
sql = 'SELECT room_id, sec_user_id, nickname, short_id, total_viewer, like_count, follower_count, signature, city FROM dy_live_lives WHERE total_viewer > 50000 AND follower_count > 500000'
self.cursor.execute(sql)
row = self.cursor.fetchone()
while row:
data = {}
data['room_id'] = row[0]
data['sec_user_id'] = row[1]
data['nickname'] = row[2]
data['short_id'] = row[3]
data['total_viewer'] = row[4]
data['like_count'] = row[5]
data['follower_count'] = row[6]
data['signature'] = row[7]
data['city'] = row[8]
self.a_list.append(data)
row = self.cursor.fetchone()
def get_cates(self, data):
sec_user_id = data['sec_user_id']
try:
cates_raw_data = self.get_raw_data.get_cates(sec_user_id)
except Exception as e:
logger.error('get_cates出错-' + e.args[0] + '-sec_user_id-' + sec_user_id)
return None
cate_list = cates_raw_data.get('user_shop_categories')
for each in cate_list:
cate = each['name']
number = each['count']
if cate in ['零食', '食品', '花茶', '果茶'] and number >= 3:
self.b_list.append(data)
break
def run_b(self):
self.select_users()
logger.info('a_list共有数据-' + str(len(self.a_list)))
batch_size = 100
for batch_limit in range(0, len(self.a_list), batch_size):
start = batch_limit
stop = min(batch_limit+batch_size, len(self.a_list))
logger.info('当前爬取用户序号-' + str(start+1) + '-' + str(stop))
tasks = [gevent.spawn(self.get_cates, data) for data in self.a_list[start:stop]]
gevent.joinall(tasks)
logger.info('b_list共有数据-' + str(len(self.b_list)))
for data in self.b_list:
self.into_mysql(data, 'dy_sample')
def run_c(self):
self.select_users()
for data in self.a_list:
self.into_mysql(data, 'dy_sample')
def select_rooms(self):
room_list = []
sql = 'SELECT room_id FROM dy_sample'
self.cursor.execute(sql)
row = self.cursor.fetchone()
while row:
room_list.append(row[0])
row = self.cursor.fetchone()
return room_list
def get_txt(self):
with open('lives_20200605.txt', 'r', encoding='utf-8') as f:
lines = f.readlines()
for line in lines:
print(line)
break
def write_to_file(self, item_list):
today = time.strftime('%Y-%m-%d', time.localtime())
today = today.replace('-', '')
with open (FILE_DIRECTORY + '/' + 'item_lists_sample'+ '_' + today + '.txt', 'a', encoding='utf-8') as file:
file.write(item_list + '\n')
def run_d(self):
url = 'https://detailskip.taobao.com/service/getData/1/p1/item/detail/sib.htm?itemId=606547898363&sellerId=2206709156233&modules=dynStock,qrcode,viewer,price,duty,xmpPromotion,delivery,activity,fqg,zjys,couponActivity,soldQuantity,page,originalPrice,tradeContract&callback=onSibRequestSuccess'
headers = {
'Referer': 'https://item.taobao.com/item.htm?id=606547898363',
'Sec-Fetch-Mode': 'no-cors',
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/76.0.3809.87 Safari/537.36'
}
response = requests.get(url, headers=headers, allow_redirects=False)
print(response.text)
def run_e(self):
file = 'lives_20200609.txt'
table = 'dy_live_lives'
data_batch = []
batch_size = 200
loop = asyncio.get_event_loop()
task = loop.create_task(self.mysql_client.connect_mysql(loop))
loop.run_until_complete(task)
with open(file, 'r', encoding='utf-8') as f:
lines = f.readlines()
for line in lines:
line = json.loads(line)
if self.is_qualified(line.get('nickname')):
if line.get('status') == 4:
data_batch.append(line) #最初数据没有入库然后程序很快就结束，是因为，4写成了’4‘。然后我以为其他地方有问题，就去打印，然后由于颜文字，打印出错（颜文字不会给入库造成错误），就在颜文字上卡了很久#最初数据没有入库然后程序很快就结束，是因为，4写成了’4‘。然后我以为其他地方有问题，就去打印，然后由于颜文字，打印出错（颜文字不会给入库造成错误），就在颜文字上卡了很久
line.pop('mobile')
if len(data_batch) >= batch_size:
tasks = [self.mysql_client.into_mysql(loop, i, table) for i in data_batch]
loop.run_until_complete(asyncio.wait(tasks))
#tasks = [gevent.spawn(self.into_mysql, line) for line in data_batch]
#gevent.joinall(tasks)
data_batch.clear()
else:
self.redis_client.delete_users(line.get('sec_user_id'))
self.redis_client.delete_live_users(line.get('sec_user_id'))
print('删除user', line.get('nickname'))
def run_f(self):
file_a = 'lives_20200623.txt'
file_b = '第四批_抖音主播_去重前.csv'
with open(file_a, 'r', encoding='utf-8') as f:
with open(file_b, 'a', encoding='utf-8-sig', newline='') as g:
lines = f.readlines()
first_line = json.loads(lines[0])
first_line.pop('mobile')
keys = list(first_line.keys())
writer = csv.DictWriter(g, fieldnames = keys)
for line in lines:
line = json.loads(line)
if self.is_qualified(line.get('nickname')):
if line.get('status') == 4:
line.pop('mobile')
writer.writerow(line)
def run_g(self):
read_workbook = xlrd.open_workbook('C:/Users/百芯科技/scraping/douyin7/第四批_抖音主播_去重后.xlsx')
write_workbook = copy(read_workbook)
read_sheet = read_workbook.sheet_by_name('Sheet1')
write_sheet = write_workbook.get_sheet(0)
nrows = read_sheet.nrows
ncolumns = read_sheet.ncols
for row in range(1, nrows):
text = read_sheet.row(row)[15].value
if text:
data = re.match('.*(1\d{10}).*', str(text), re.S)
if data:
mobile = data.group(1)
write_sheet.write(row, ncolumns, mobile)
write_workbook.save('C:/Users/百芯科技/scraping/douyin7/第四批_抖音主播_去重后_电话.xlsx')
class GetPromotions():
def __init__(self):
self.get_raw_data = GetRawData()
self.redis_client = RedisClient()
self.aweme_id_list = []
def get_users(self):
users = self.redis_client.get_users()
return users
def get_promotions(self, sec_user_id):
try:
raw_data = self.get_raw_data.get_promotions(sec_user_id)
except Exception as e:
logger.error('get_promotions出错-' + e.args[0] + '-sec_user_id-' +
sec_user_id)
return None
try:
promotions = self.parse_promotions(raw_data, sec_user_id)
except Exception as e:
logger.error('parse_promotions错误-' + e.args + '-sec_user_id-' +
sec_user_id)
return None
self.write_to_file(json.dumps(promotions, ensure_ascii=False))
def parse_promotions(self, raw_data, sec_user_id):
promotions = []
data = raw_data.get('promotions')
if data == []:
logger.info('该用户不带货，将删除，sec_user_id-' + sec_user_id)
self.redis_client.delete_users(sec_user_id)
return None
else:
for each in data:
promotion = {}
#promotion['user_id'] = user_id
promotion['sec_user_id'] = sec_user_id
promotion['price'] = each.get('price') / 100
promotion['cover_url'] = each.get('images')[0].get(
'url_list')[0]
promotion['title'] = each.get('title')
promotion['product_id'] = each.get('product_id')
promotion['product_url'] = each.get('detail_url')
promotion['min_price'] = str(int(each.get('min_price')) / 100)
promotion['douyin_sales'] = each.get('sales')
promotion['product_source'] = each.get('goods_source')
promotion['create_time'] = str(int(time.time()))
if each.get('market_price'):
promotion['market_price'] = each.get('market_price') / 100
if each.get('last_aweme_id'):
promotion['promotion_type'] = 'video'
promotion['aweme_id'] = int(each.get('last_aweme_id'))
self.aweme_id_list.append(promotion['aweme_id'])
else:
promotion['promotion_type'] = 'picture'
if each.get('taobao'):
taobao = each.get('taobao')
if taobao.get('coupon'):
promotion['coupon_amount'] = taobao.get('coupon').get(
'coupon_amount')
promotion['price_after_coupon'] = promotion[
'price'] - float(promotion['coupon_amount'])
promotion['coupon_url'] = taobao.get('coupon').get(
'coupon_web_url')
promotions.append(promotion)
return promotions
def write_to_file(self, promotions):
today = time.strftime('%Y-%m-%d', time.localtime())
today = today.replace('-', '')
harry_potter = str(random.choice(range(100)))
with open(FILE_DIRECTORY + '/' + 'promotions' + '_' + harry_potter +
'.txt',
'a',
encoding='utf-8') as file:
file.write(promotions + '\n')
def save_awemes(self):
for each in self.aweme_id_list:
self.redis_client.add_awemes(each)
def run(self):
users = self.get_users()
logger.info('共有用户数量：' + str(len(users)))
batch_size = 50 #尽管异步，还是很慢，200个就很慢很慢了，慢到跟同步一样，这可能是抖音某个神奇的特点吧
for batch_limit in range(0, len(users), batch_size):
start = batch_limit
stop = min(batch_limit + batch_size, len(users))
logger.info('get_promotions爬取当前用户序号-' + str(start + 1) + '-' +
str(stop))
tasks = [
gevent.spawn(self.get_promotions, sec_user_id)
for sec_user_id in users[start:stop]
]
gevent.joinall(tasks)
self.save_awemes()
self.aweme_id_list.clear()
class GetUserProfile():
def __init__(self):
self.get_raw_data = GetRawData()
self.redis_client = RedisClient()
self.stupid_key_words = STUPID_KEY_WORDS
def get_users(self):
users = self.redis_client.test_b()
return users
def get_user_profile(self, sec_user_id):
try:
raw_data = self.get_raw_data.get_user_profile(sec_user_id)
except Exception as e:
logger.error('get_user_profile出错-' + e.args[0] + '-sec_user_id-' +
sec_user_id)
return None
try:
user_profile = self.parse_user_profile(raw_data)
except Exception as e:
logger.error('parse_user_profile出错-' + e.args[0] +
'-sec_user_id-' + sec_user_id)
return None
if user_profile:
self.write_to_file(json.dumps(user_profile, ensure_ascii=False))
self.redis_client.test_c(sec_user_id)
def is_qualified_user(self, user):
nickname = user.get('nickname')
for word in self.stupid_key_words:
if word in nickname:
return False
if user.get('is_gov_media_vip'):
return False
if user.get('enterprise_verify_reason') != '':
return False
if user.get('custom_verify') != '':
return False
if not user.get('with_fusion_shop_entry'):
return False
if not user.get('live_commerce'):
return False
if not user.get('with_commerce_entry'):
return False
return True
def parse_user_profile(self, raw_data):
data = raw_data.get('user')
user_profile = {}
user_profile['sec_uid'] = data.get('sec_uid')
if self.is_qualified_user(data):
user_profile['follower_count'] = data.get('follower_count')
user_profile['nickname'] = data.get('nickname')
user_profile['gender'] = data.get('gender')
user_profile['location'] = data.get('location')
user_profile['birthday'] = data.get('birthday')
user_profile['avatar_url'] = data.get('avatar_larger').get(
'url_list')[0]
user_profile['school_name'] = data.get('school_name')
user_profile['signature'] = data.get('signature')
user_profile['uid'] = data.get('uid')
user_profile['short_id'] = data.get('short_id')
user_profile['unique_id'] = data.get('unique_id')
user_profile['star_atlas'] = data.get('commerce_user_info').get(
'star_atlas')
user_profile['aweme_count'] = data.get('aweme_count')
user_profile['dongtai_count'] = data.get('dongtai_count')
user_profile['following_count'] = data.get('following_count')
user_profile['favoriting_count'] = data.get('favoriting_count')
user_profile['total_favorited'] = data.get('total_favorited')
user_profile['live_commerce'] = data.get('live_commerce')
user_profile['create_time'] = str(int(time.time()))
return user_profile
else:
self.redis_client.delete_users(user_profile['sec_uid'])
#print(json.dumps(raw_data, ensure_ascii=False))
logger.info('删除user-sec_user_id-' + user_profile['sec_uid'])
return None
def write_to_file(self, user_profile):
today = time.strftime('%Y-%m-%d', time.localtime())
today = today.replace('-', '')
with open(FILE_DIRECTORY + '/' + 'user_profiles' + '_' + today +
'.txt',
'a',
encoding='utf-8') as file:
file.write(user_profile + '\n')
def run(self):
users = self.get_users()
logger.info('共有users-' + str(len(users)))
batch_size = 1 #50个会获取不到数据
for batch_limit in range(0, len(users), batch_size):
start = batch_limit
stop = min(batch_limit + batch_size, len(users))
logger.info('当前爬取用户序号-' + str(start + 1) + '-' + str(stop))
tasks = [
gevent.spawn(self.get_user_profile, sec_user_id)
for sec_user_id in users[start:stop]
]
gevent.joinall(tasks)
class CheckRooms():
def __init__(self):
self.get_raw_data = GetRawData()
self.redis_client = RedisClient()
self.lives_on_list = []
def get_rooms(self):
return self.redis_client.get_rooms(0, 0)
def change_room_status(self):
for each in self.lives_on_list:
self.redis_client.add_rooms(each, 1)
def check_room(self, room_id):
try:
room_raw_data = self.get_raw_data.get_live(room_id)
except Exception as e:
logger.error('get_live出错-' + e.args[0] + '-room_id-' + room_id)
return None
try:
owner = room_raw_data.get('data').get('owner')
follower_count = owner.get('follow_info').get('follower_count')
sec_user_id = owner.get('sec_uid')
except Exception as e:
logger.error('解析room_raw_data出错-' + e.args[0] + '-room_id-' +
room_id)
return None
if follower_count < 10000:
self.redis_client.delete_users(sec_user_id)
self.redis_client.delete_rooms(room_id)
else:
status = room_raw_data.get('data').get('status')
if status == 2:
try: #判断该场直播是否带货
item_list_raw_data = self.get_raw_data.get_item_list(
sec_user_id, room_id)
except Exception as e:
logger.error('get_item_list出错' + e.args[0] +
'-sec_user_id和room_id-' + sec_user_id + '-' +
room_id)
return None
if len(item_list_raw_data.get('promotions')) != 0:
self.lives_on_list.append(room_id)
def run(self):
all_room_ids = self.get_rooms()
logger.info('此前未在直播的直播间数量：' + str(len(all_room_ids)))
batch_size = 200
for batch_limit in range(0, len(all_room_ids), batch_size):
start = batch_limit
stop = min(batch_limit + batch_size, len(all_room_ids))
logger.info('待查看的此前未在直播的直播间-' + str(start + 1) + '-' + str(stop))
tasks = [
gevent.spawn(self.check_room, room_id)
for room_id in all_room_ids[start:stop]
]
gevent.joinall(tasks)
logger.info('新发现开始的直播数量-' + str(len(self.lives_on_list)))
self.change_room_status()
self.lives_on_list.clear()
