w

hhh/SEO/local_1777.py

@@ -19,35 +19,32 @@ import codecs
 import random
 import os
 import time
-
+from userAgentRandomizer import userAgents
 
 driver=None
 
-
 def re_get_webdriver():
     global driver
     result=[]
-#    client = docker.from_env()
-#    ls=client.containers.list()
-#    print(ls)
-#    for l in ls:
-#        if 'p17777' in l.name:
-#            ls[0].restart()
-
-#    time.sleep(4)
     if driver is not None:
         print('closing....')
         driver.quit()
-#    options = webdriver.EdgeOptions()
     try:
+        ua = userAgents()
+
+        user_agent = ua.random()        
 
         options = webdriver.ChromeOptions()
         options.add_argument("--no-sandbox")
         options.add_argument("--disable-dev-shm-usage")
         options.add_argument("--headless")
 
+        print(user_agent)
+        options.add_argument("--user-agent=" +user_agent)
+        options.add_argument("--incognito")
+
+
         driver = webdriver.Chrome(options=options)
-#        desired_capabilities=options.to_capabilities())
 
         driver.set_window_size(1400,1000)
         return
@@ -56,11 +53,10 @@ def re_get_webdriver():
         traceback.print_exc()
         driver=None
         return None
-    driver=None
 
 def run_once(url):
     global driver
-    i=random.randint(0,15)
+    i=random.randint(0,2)
     if i==0 or driver is None:
 #    if True:
         re_get_webdriver()
@@ -75,7 +71,6 @@ def run_once(url):
         print('exception')
 
 
-
 lst=[]
 
 
@@ -83,7 +78,7 @@ db = dataset.connect('mysql://choozmo:pAssw0rd@db.ptt.cx:3306/seo?charset=utf8mb
 
 cursor=db.query('SELECT * FROM columnids order by rand()')
 for c in cursor:
-    lst.append('https://www.hhh.com.tw/columns/detail/'+str(c['cid'])+'/index.php')
+    lst.append('https://www.hhh.com.tw/columns/detail/'+str(c['cid'])+'/')
 
 
 
@@ -93,5 +88,7 @@ while True:
     l=random.choice(lst)
     print(l)
     run_once(l)
+    time.sleep(5)
+    driver=None