Initial support for proxies. Related to #111 #86

Author: dayures

SPARQLWrapper/Wrapper.py

@@ -290,6 +290,7 @@ def __init__(self, endpoint, updateEndpoint=None, returnFormat=XML, defaultGraph
         self._defaultGraph = defaultGraph
         self.onlyConneg = False # Only Content Negotiation
         self.customHttpHeaders = {}
+        self.proxies = {}
 
         if returnFormat in _allowedFormats:
             self._defaultReturnFormat = returnFormat
@@ -344,6 +345,15 @@ def setTimeout(self, timeout):
         """
         self.timeout = int(timeout)
 
+    def setProxies(self, proxies):
+        """Set the proxies to use for querying the endpoint.
+        @since: 1.8.3
+
+        @param proxies: Dictionary mapping protocol names to URLs of proxies.
+        @type proxies: dict
+        """
+        self.proxies = proxies
+
     def setOnlyConneg(self, onlyConneg):
         """Set this option for allowing (or not) only HTTP Content Negotiation (so dismiss the use of HTTP parameters).
         @since: 1.8.1
@@ -745,6 +755,11 @@ def _createRequest(self):
         for customHttpHeader in self.customHttpHeaders:
             request.add_header(customHttpHeader, self.customHttpHeaders[customHttpHeader])
 
+        if self.proxies:
+            proxy_support = urllib2.ProxyHandler(self.proxies)
+            opener = urllib2.build_opener(proxy_support)
+            urllib2.install_opener(opener)
+
         return request
 
     def _query(self):

test/proxy_test.py

@@ -0,0 +1,100 @@
+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+
+import inspect
+import os
+import sys
+import urllib
+
+# prefer local copy to the one which is installed
+# hack from http://stackoverflow.com/a/6098238/280539
+_top_level_path = os.path.realpath(os.path.abspath(os.path.join(
+    os.path.split(inspect.getfile(inspect.currentframe()))[0],
+    ".."
+)))
+if _top_level_path not in sys.path:
+    sys.path.insert(0, _top_level_path)
+# end of hack
+
+from SPARQLWrapper import SPARQLWrapper, XML, GET, SELECT
+import unittest
+from urllib2 import URLError
+import socket
+import errno  
+
+endpoint = "http://dbpedia.org/sparql"
+endpoint_ssl = "https://dbpedia.org/sparql"
+
+prefixes = """
+    PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+"""
+
+selectQuery = """
+    SELECT ?label
+    WHERE {
+    <http://dbpedia.org/resource/Asturias> rdfs:label ?label .
+    }
+"""
+
+proxiesWorking = {
+            "http":"177.22.107.236:8080", 
+            "https":"177.22.107.236:8080"} # Checked using http://www.proxy-checker.org/
+
+proxiesNotWorking = {"http":"127.0.0.1:80", 
+                     "https":"127.0.0.1:80"}
+
+class SPARQLWrapperProxyTests(unittest.TestCase):
+
+    def __generic(self, endpoint):
+        sparql = SPARQLWrapper(endpoint)
+        sparql.setQuery(prefixes + selectQuery)
+        sparql.setReturnFormat(XML)
+        sparql.setMethod(GET)
+        return sparql
+
+    def testProxyWorking(self):
+        sparql = self.__generic(endpoint)
+        sparql.setProxies(proxiesWorking)
+        self.assertTrue(sparql.proxies) # assert no empty
+
+        try:
+            result = sparql.query() 
+        except URLError as error:
+            print "The Proxy server is not responding"
+            print error # Because we are not sure that the proxy is working
+        else:
+            result.convert().toprettyxml()
+            self.assertTrue(True)
+
+    def testProxyWorkingSSL(self):
+        sparql = self.__generic(endpoint_ssl)
+        self.assertEqual(sparql.endpoint, endpoint_ssl) 
+        sparql.setProxies(proxiesWorking)
+        self.assertTrue(sparql.proxies) # assert no empty
+
+        try:
+            result = sparql.query() 
+        except URLError as error:
+            print "The Proxy server is not responding"
+            print error # Because we are not sure that the proxy is working
+        else:
+            result.convert().toprettyxml()
+            self.assertTrue(True)
+
+    def testProxyNotWorking(self):
+        sparql = self.__generic(endpoint)
+        sparql.setProxies(proxiesNotWorking)
+        self.assertTrue(sparql.proxies) # assert no empty
+
+        try:
+            result = sparql.query() 
+        except URLError as error:
+            self.assertTrue(True)
+        else:
+            self.assertTrue(False)
+
+if __name__ == "__main__":
+    unittest.main()
+    
+    

test/wrapper_test.py

@@ -312,6 +312,15 @@ def testSetHTTPAuth(self):
         self.wrapper.http_auth = "OAuth"
         self.assertRaises(NotImplementedError, self._get_request, self.wrapper)
 
+    def testSetProxies(self):
+        proxies = {"http": "127.0.0.1"}
+        self.assertFalse(self.wrapper.proxies)
+        self.wrapper.setProxies(proxies)
+        self.assertTrue(self.wrapper.proxies)
+        self.assertEqual(self.wrapper.proxies, proxies)
+        self.assertIsInstance(urllib2._opener, urllib2.OpenerDirector)
+
+
     def testSetQuery(self):
         self.wrapper.setQuery('PREFIX example: <http://example.org/INSERT/> SELECT * WHERE {?s ?p ?v}')
         self.assertEqual(SELECT, self.wrapper.queryType)