Update app.py

app.py

@@ -1,6 +1,5 @@
 from seleniumwire import webdriver
 from selenium.webdriver.chrome.options import Options
-from selenium.common.exceptions import WebDriverException
 from fastapi import FastAPI, Request
 import uvicorn
 import time
@@ -25,29 +24,14 @@ def get_root_domain(url):
         return '.'.join(parts[-2:])
     else:
         return domain
-
-# 网络抓包内容过滤
-def filter_type(_type: str):
-    types = [
-        'application/javascript', 'application/x-javascript', 'text/css', 'webp', 'image/png', 'image/gif',
-        'image/jpeg', 'image/x-icon', 'application/octet-stream'
-    ]
-    if _type not in types:
-        return True
-    return False
-    
+        
 @app.get("/")
 def main():
     return {"code": 200,"msg":"Success"}
     
 @app.get("/chrome")
 def chrome(url:str=None,wait:int=5,header:str=None,cookie:str=None):
-    # 设置日志性能监听参数
-    caps = {
-        "browserName": "chrome",
-        'goog:loggingPrefs': {'performance': 'ALL'}
-    }
-    
+
     # 必须有目标url
     if type(url) == str:
         target_url = unquote(url)
@@ -80,20 +64,17 @@ def chrome(url:str=None,wait:int=5,header:str=None,cookie:str=None):
     # 设置为无头模式
     options.add_argument('--headless')
 
-    # 开启日志性能监听
-    for key, value in caps.items():
-        options.set_capability(key, value)
-
     # 实例化
     driver = webdriver.Chrome(options=options)
 
     # 需要打开网址页面，才能用 driver.add_cookie 进行cookie追加
     driver.get(target_url)
 
-    # 清除本次打开网址页面，可能存储在本地的cookie、sessionStorage、localStorage
+    # 清除本次打开网址页面，可能存储在本地的cookie、sessionStorage、localStorage，并删除因此次访问所产生的 har
     driver.delete_all_cookies()
     driver.execute_script("window.sessionStorage.clear();")
     driver.execute_script("window.localStorage.clear();")
+    del driver.requests
 
     # 对浏览器追加我们传递进来的cookie
     if 'cookie' in header_array:
@@ -126,34 +107,19 @@ def chrome(url:str=None,wait:int=5,header:str=None,cookie:str=None):
 
     # 完全加载完成时，页面是否有发生过 301 302 跳转过
     is_jump = (target_url != current_url)
-    
-    network = []
-    performance_log = driver.get_log('performance')  # 获取名称为 performance 的日志
-    for packet in performance_log:
-        message = json.loads(packet.get('message')).get('message')  # 获取message的数据
-        if message.get('method') != 'Network.responseReceived':  # 如果method 不是 responseReceived 类型就不往下执行
-            continue
-        packet_type = message.get('params').get('response').get('mimeType')  # 获取该请求返回的type
-        if not filter_type(_type=packet_type):  # 过滤type
-            continue
-        requestId = message.get('params').get('requestId')  # 唯一的请求标识符。相当于该请求的身份证
-        url = message.get('params').get('response').get('url')  # 获取 该请求  url
-        headers = message.get('params').get('response').get('headers')  # 获取 该请求  url
-        status = message.get('params').get('response').get('status')  # 获取 该请求  url
-        try:
-            network.append({"status":status ,"url":url, "type":packet_type, "headers":headers})
-            # resp = driver.execute_cdp_cmd('Network.getResponseBody', {'requestId': requestId})
-            # print(f'response: {resp}')
-        except WebDriverException:  # 忽略异常
-            pass
-    print(driver.requests)
 
+    network = []
+    # 遍历输出过程中的har
+    for request in driver.requests:
+        if request.response:
+            network.append({"method":request.response.method, "status":request.response.status_code ,"url":request.url, "headers":request.response.headers})
+        
     data = {
         "url": current_url,
         "page_source": page_source,
         "cookies": cookies,
         "is_jump": is_jump,
-        "network": network,
+        "network": network
     }
     
     driver.quit()