checking for subclass of URLWarning instead of checking error message text when URL in allowed_domains

scrapy/spidermiddlewares/offsite.py

@@ -56,10 +56,15 @@ class OffsiteMiddleware(object):
         for domain in allowed_domains:
             url_pattern = re.compile("^https?://.*$")
             if url_pattern.match(domain):
-                warnings.warn("allowed_domains accepts only domains, not URLs. Ignoring URL entry %s in allowed_domains." % domain, Warning)
+                warnings.warn("allowed_domains accepts only domains, not URLs. Ignoring URL entry %s in allowed_domains." % domain, URLWarning)
+                
         regex = r'^(.*\.)?(%s)$' % '|'.join(re.escape(d) for d in allowed_domains if d is not None)
         return re.compile(regex)
 
     def spider_opened(self, spider):
         self.host_regex = self.get_host_regex(spider)
         self.domains_seen = set()
+
+
+class URLWarning(Warning):
+    pass

tests/test_spidermiddleware_offsite.py

@@ -5,6 +5,7 @@ from six.moves.urllib.parse import urlparse
 from scrapy.http import Response, Request
 from scrapy.spiders import Spider
 from scrapy.spidermiddlewares.offsite import OffsiteMiddleware
+from scrapy.spidermiddlewares.offsite import URLWarning
 from scrapy.utils.test import get_crawler
 import warnings
 
@@ -78,4 +79,4 @@ class TestOffsiteMiddleware5(TestOffsiteMiddleware4):
         with warnings.catch_warnings(record=True) as w:
             warnings.simplefilter("always")
             self.mw.get_host_regex(self.spider)
-            assert "allowed_domains accepts only domains, not URLs." in str(w[-1].message)
+            assert issubclass(w[-1].category, URLWarning)