🐛🔥 bugfix & remove useless codes

kaixinol · Nov 20, 2023 · 15b68e2 · 15b68e2
1 parent 8ed55d4
commit 15b68e2
Show file tree

Hide file tree

Showing 4 changed files with 4 additions and 11 deletions.
diff --git a/config.yaml b/config.yaml
@@ -1,7 +1,6 @@
 { "proxy":
     {"http": "socks5://127.0.0.1:7890", "https": "socks5://127.0.0.1:7890" }, "max_threads": 1,
-  "header": {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/118.0.0.0 Safari/537.36"}
-,"user_data_dir": "/media/Data/Project/twitter_user_tweet_crawler/twitter_user_tweet_crawler/userdata",
+  "header": {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/118.0.0.0 Safari/537.36"},
   "inject_js": "/media/Data/Project/twitter_user_tweet_crawler/script.js",
   "save": "/media/Data/Project/twitter_user_tweet_crawler/output",
   "user": "s_nample"

diff --git a/test.py b/test.py
@@ -10,13 +10,12 @@ def get_tweet():
     set_work_directory(Path(__file__).absolute().parent)
     config.load({"proxy": {"http": "socks5://127.0.0.1:7890", "https": "socks5://127.0.0.1:7890"}, "max_threads": 2,
                  "header": {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, "
-                                          "like Gecko) "},
-                 "user_data_dir": "/media/Data/Project/twitter_user_tweet_crawler/twitter_user_tweet_crawler/userdata"
+                                          "like Gecko) "}
                     , "inject_js": "/media/Data/Project/twitter_user_tweet_crawler/script.js",
                  "save": "/media/Data/Project/twitter_user_tweet_crawler/output/", }
                 )
     from twitter_user_tweet_crawler.tweet import Tweet
-    Path(config.save / 'res').mkdir(parents=True, exist_ok=True)
+    (Path(config.save) / 'res').mkdir(parents=True, exist_ok=True)
     browser = get_browser()
     browser.get('https://twitter.com/404')
     cookie: list[dict]

diff --git a/twitter_user_tweet_crawler/__main__.py b/twitter_user_tweet_crawler/__main__.py
@@ -39,7 +39,7 @@ def get_items_need_handle():
         return driver.find_elements(*selector)
 
     selector = (By.XPATH, '//*/div[2]/div/div[3]/a[@role="link"]')
-    Path(config.save / 'res').mkdir(exist_ok=True, parents=True)
+    (Path(config.save) / 'res').mkdir(exist_ok=True, parents=True)
 
     driver = get_browser()
 

diff --git a/twitter_user_tweet_crawler/browser.py b/twitter_user_tweet_crawler/browser.py
@@ -14,11 +14,6 @@ def get_browser(headless: bool = False, id=None) -> WebDriver:
     chrome_options.add_argument('--disable-gpu')
     chrome_options.add_argument('--no-sandbox')
     chrome_options.add_argument('--window-size=1200x600"')
-
-    if not id:
-        chrome_options.add_argument(f'user-data-dir={config["user_data_dir"] + "/" + str(browsers := browsers + 1)}')
-    else:
-        chrome_options.add_argument(f'user-data-dir={config["user_data_dir"] + "/" + str(id)}')
     if headless:
         chrome_options.add_argument('--headless')
     driver = webdriver.Chrome(options=chrome_options)