将元数据的来源网站记入日志以便进行评估

ADC_function.py

@@ -1,6 +1,6 @@
 from os import replace
 import requests
-import hashlib
+#import hashlib
 from pathlib import Path
 import secrets
 import os.path
@@ -20,7 +20,7 @@ def getXpathSingle(htmlcode, xpath):
     return result1
 
 
-G_USER_AGENT = r'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/94.0.4606.71 Safari/537.36'
+G_USER_AGENT = r'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/94.0.4606.81 Safari/537.36'
 
 # 网页请求核心
 def get_html(url, cookies: dict = None, ua: str = None, return_type: str = None):

WebCrawler/__init__.py

@@ -115,6 +115,7 @@ def get_data_from_json(file_number):  # 从JSON返回元数据
             json_data = json.loads(pool.apply_async(func_mapping[source], (file_number,)).get())
             # if any service return a valid return, break
             if get_data_state(json_data):
+                print(f"[+]Find movie [{file_number}] metadata on website '{source}'")
                 break
         pool.close()
         pool.terminate()
@@ -126,6 +127,7 @@ def get_data_from_json(file_number):  # 从JSON返回元数据
                 json_data = json.loads(func_mapping[source](file_number))
                 # if any service return a valid return, break
                 if get_data_state(json_data):
+                    print(f"[+]Find movie [{file_number}] metadata on website '{source}'")
                     break
             except:
                 break

WebCrawler/storyline.py

@@ -34,7 +34,7 @@ def getStoryline(number, title):
     for i in range(cnt):
         sl = len(result[i])if isinstance(result[i], str) else 0
         if sl and first:
-            s += f'，[选中结果{apply_sites[i]}字数:{sl}]'
+            s += f'，[选中{apply_sites[i]}字数:{sl}]'
             first = False
             sel = result[i]
         elif sl:

config.py

@@ -394,10 +394,10 @@ if __name__ == "__main__":
         code = compile(evstr, "<string>", "eval")
         print('{}: "{}"'.format(evstr, eval(code)))
     config = Config()
-    mfilter = ('conf', 'proxy', '_exit', '_default_config', 'getboolean_override', 'getint_override', 'get_override', 'ini_path')
+    mfilter = {'conf', 'proxy', '_exit', '_default_config', 'getboolean_override', 'getint_override', 'get_override', 'ini_path'}
     for _m in [m for m in dir(config) if not m.startswith('__') and m not in mfilter]:
         evprint(f'config.{_m}()')
-    pfilter = ('proxies', 'SUPPORT_PROXY_TYPE')
+    pfilter = {'proxies', 'SUPPORT_PROXY_TYPE'}
     # test getInstance()
     assert(getInstance() == config)
     for _p in [p for p in dir(getInstance().proxy()) if not p.startswith('__') and p not in pfilter]: