cnki:headers中间件处理cookie字符串

science_article_cnki/science_article_cnki/middlewares.py

@@ -104,13 +104,25 @@ from scrapy.http.headers import Headers
 
 
 class CnkiSearchHeadersDownloaderMiddleware:
-    def __init__(self, custom_headers: dict):
+    def __init__(self, custom_headers: dict, cookies_str: str):
         self.custom_headers = custom_headers
+        self.custom_cookies = self._parse_cookies_str(cookies_str)
 
     @classmethod
     def from_crawler(cls, crawler):
-        return cls(custom_headers=crawler.settings['SEARCH_REQUEST_HEADERS'])
+        return cls(
+            custom_headers=crawler.settings['SEARCH_REQUEST_HEADERS'],
+            cookies_str=crawler.settings['SEARCH_REQUEST_COOKIES_STR']
+        )
+
+    def _parse_cookies_str(self, cookies_str):
+        cookies = {}
+        for cookie in cookies_str.split(';'):
+            key, value = cookie.split('=', 1)
+            cookies[key.strip()] = value.strip()
+        return cookies
 
     def process_request(self, request, spider):
         request.headers = Headers(self.custom_headers)
+        request.cookies = self.custom_cookies
         return None