123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246 |
- from selenium import webdriver
- from selenium.webdriver.common.desired_capabilities import DesiredCapabilities
- import time
- import os
- import urllib.parse
- from selenium.webdriver.support.ui import WebDriverWait
- from selenium.webdriver.common.by import By
- from selenium.webdriver.support import expected_conditions as EC
- import codecs
- import random
- from bs4 import BeautifulSoup
- import requests
- import time
- import rpyc
- import sys
- import docker
- import dataset
- headers = {
- "Authorization": "Bearer " + "t35vhZtWNgvDNWHc3DJh0OKll3mcB9GvC8K2EAkBug2",
- "Content-Type": "application/x-www-form-urlencoded"
- }
- def send_msg(kw):
- params = {"message": "處理關鍵字: "+kw}
- r = requests.post("https://notify-api.line.me/api/notify",headers=headers, params=params)
- def empty_query(q):
- global driver
- googleurl='https://www.google.com/search?q='+urllib.parse.quote(q)
- driver.get(googleurl)
- time.sleep(3)
- def process_query(q):
- global driver
- googleurl='https://www.google.com/search?q='+urllib.parse.quote(q)
- driver.get(googleurl)
- time.sleep(3)
- elmts=driver.find_elements_by_xpath("//div[@class='g']//div[@class='yuRUbf']//a")
- idx=1
- ranking=-1
- for elmt in elmts:
- href=elmt.get_attribute('href')
- txt=elmt.text
- if len(txt)>10:
- if 'hhh.com.tw' in href:
- webdriver.ActionChains(driver).move_to_element(elmt).perform()
- webdriver.ActionChains(driver).move_to_element(elmt).click().perform()
- break
- client = docker.from_env()
- ls=client.containers.list()
- print(ls)
- ls[0].restart()
- time.sleep(10)
- options = webdriver.ChromeOptions()
- options.add_argument("--proxy-server=socks5://172.104.67.159:8180")
- driver = webdriver.Remote(
- command_executor='http://127.0.0.1:4444/wd/hub',
-
- desired_capabilities=options.to_capabilities())
- driver.set_window_size(1400,1000)
- qlist=[]
- def get_list_from_db2():
- global qlist
- db = dataset.connect('mysql://choozmo:pAssw0rd@db.ptt.cx:3306/hhh?charset=utf8mb4')
- query="""SELECT query as kw FROM hhh.gsc_designer where
- (`page` = 'https://hhh.com.tw/designers/cases/491/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/31/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/293/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/278/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/24/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/594/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/356/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/307/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/491/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/33/1-page/new-sort/') or
- (`page` = 'https://hhh.com.tw/designers/cases/385/1-page/new-sort/') and position <=10
- order by impressions/position desc;
- """
- cursor=db.query(query)
- for c in cursor:
- qlist.append(c['kw'])
- get_list_from_db2()
- while True:
- q=random.choice(qlist)
- prob=random.randint(0,4)
- if prob <=1:
-
- print(q)
- process_query(q)
- send_msg(q)
- else:
- empty_query(q)
-
- intsleep=random.randint(7,20)
- time.sleep(intsleep)
|