天眼查核心人员更新

comData/Tyc/CorePerson.py

@@ -45,15 +45,18 @@ def get_html(tycid, s, headers):
         # div_part.find('div', class_='dimHeader_root__XTCLe')
     except:
         return -1
-    try:
-        tmp_field = div_part.find('div', class_='dim-tab-root').find('span').text
-        if '最新公示' in tmp_field:
-            total = div_part.find('div', class_='dim-tab-root').find('span').get_text().split('最新公示')[1].replace(' ', '')
-            return int(total)
-        else:
-            return -1
-    except:
-        return 0
+    if div_part is None:
+        return -2
+    else:
+        try:
+            tmp_field = div_part.find('div', class_='dim-tab-root').find('span').text
+            if '最新公示' in tmp_field:
+                total = div_part.find('div', class_='dim-tab-root').find('span').get_text().split('最新公示')[1].replace(' ', '')
+                return int(total)
+            else:
+                return -1
+        except:
+            return 0
 
 
 @retry(tries=3, delay=1)
@@ -64,7 +67,10 @@ def get_page(url, s, headers):
     if res.status_code != 200:
         raise
     data_page = res.json()
-    total_page_ = data_page['data']['total']
+    try:
+        total_page_ = data_page['data']['total']
+    except:
+        raise
     return total_page_
 
 
@@ -77,7 +83,7 @@ def doJob():
             'Accept-Encoding': 'gzip, deflate, br',
             'Accept-Language': 'zh-CN,zh;q=0.9',
             'Cache-Control': 'max-age=0',
-            'Connection': 'keep-alive',
+            # 'Connection': 'keep-alive',
             'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36',
             'version': 'TYC-Web'
         }
@@ -90,7 +96,7 @@ def doJob():
         # 根据从Redis中拿到的社会信用代码,在数据库中获取对应基本信息
         # social_code = baseCore.redicPullData('CorPersonEnterprise:gnqy_socialCode')
         # 判断 如果Redis中已经没有数据，则等待
-        social_code = '91440300MA5EU1QM0T'
+        social_code = '91110108780992804C'
         if social_code == None:
             time.sleep(20)
             continue
@@ -163,6 +169,11 @@ def doJob():
                 log.info(f"{id}---{xydm}----{tycid}----请求失败----重新放入redis")
                 time.sleep(2)
                 continue
+            elif charge == -2:
+                # 该企业没有人员信息
+                log.info(f"{id}---{xydm}----{tycid}----没有核心人员")
+                continue
+
             elif charge == 0:
                 log.info(f"{id}---{xydm}----{tycid}----没有最新公示")
                 url1 = f'https://capi.tianyancha.com/cloud-company-background/company/dim/staff?_={t}&gid={tycid}&pageSize=20&pageNum=1'
@@ -240,6 +251,8 @@ def doJob():
                     pass
                 else:
                     log.info(f'{id}---{xydm}----{tycid}----没有高管信息')
+                # todo: 关闭连接
+                res.close()
                 if flag == 1:
                     for one_info in list_all:
                         name = one_info['name']