Python自定义scrapy中间模块避免重复采集的方法

本文实例讲述了Python自定义scrapy中间模块避免重复采集的方法。分享给大家供大家参考。具体如下：

from scrapy import log
from scrapy.http import Request
from scrapy.item import BaseItem
from scrapy.utils.request import request_fingerprint
from myproject.items import MyItem
class IgnoreVisitedItems(object):
 """Middleware to ignore re-visiting item pages if they
 were already visited before.
 The requests to be filtered by have a meta['filter_visited']
 flag enabled and optionally define an id to use
 for identifying them, which defaults the request fingerprint,
 although you'd want to use the item id,
 if you already have it beforehand to make it more robust.
 """
 FILTER_VISITED = 'filter_visited'
 VISITED_ID = 'visited_id'
 CONTEXT_KEY = 'visited_ids'
 def process_spider_output(self, response, result, spider):
   context = getattr(spider, 'context', {})
   visited_ids = context.setdefault(self.CONTEXT_KEY, {})
   ret = []
   for x in result:
     visited = False
     if isinstance(x, Request):
       if self.FILTER_VISITED in x.meta:
         visit_id = self._visited_id(x)
         if visit_id in visited_ids:
           log.msg("Ignoring already visited: ％s" ％ x.url,
               level=log.INFO, spider=spider)
           visited = True
     elif isinstance(x, BaseItem):
       visit_id = self._visited_id(response.request)
       if visit_id:
         visited_ids[visit_id] = True
         x['visit_id'] = visit_id
         x['visit_status'] = 'new'
     if visited:
       ret.append(MyItem(visit_id=visit_id, visit_status='old'))
     else:
       ret.append(x)
   return ret
 def _visited_id(self, request):
   return request.meta.get(self.VISITED_ID) or request_fingerprint(request)

希望本文所述对大家的Python程序设计有所帮助。