OffsiteMiddleware: removed logging and simplified implementation

scrapy/contrib/spidermiddleware/offsite.py

@@ -3,16 +3,13 @@ OffsiteMiddleware: Filters out Requests for URLs outside the domains covered by
 the spider.
 """
 
-from scrapy import log
-from scrapy.http import Request
+from itertools import ifilter
+
+from scrapy.item import ScrapedItem
 from scrapy.utils.url import url_is_from_spider
 
 class OffsiteMiddleware(object):
-    def process_spider_output(self, response, result, spider):
-        def _filter(r):
-            if isinstance(r, Request) and not url_is_from_spider(r.url, spider):
-                log.msg("Ignoring link (offsite): %s " % r.url, level=log.DEBUG, domain=spider.domain_name)
-                return False
-            return True
-        return (r for r in result or () if _filter(r))
 
+    def process_spider_output(self, response, result, spider):
+        filter = lambda x: isinstance(x, ScrapedItem) or url_is_from_spider(x.url, spider)
+        return ifilter(filter, result or ())