Merge pull request #2065 from redapple/octet-stream-no-decompress

[MRG+2] [HttpCompressionMW] Do not decompress gzip binary/octet-stream responses

scrapy/utils/gz.py

@@ -50,9 +50,12 @@ def gunzip(data):
                 raise
     return output
 
-_is_gzipped_re = re.compile(br'^application/(x-)?gzip\b', re.I)
+_is_gzipped = re.compile(br'^application/(x-)?gzip\b', re.I).search
+_is_octetstream = re.compile(br'^(application|binary)/octet-stream\b', re.I).search
 
 def is_gzipped(response):
     """Return True if the response is gzipped, or False otherwise"""
     ctype = response.headers.get('Content-Type', b'')
-    return _is_gzipped_re.search(ctype) is not None
+    cenc = response.headers.get('Content-Encoding', b'').lower()
+    return (_is_gzipped(ctype) or
+            (_is_octetstream(ctype) and cenc in (b'gzip', b'x-gzip')))

tests/test_downloadermiddleware_httpcompression.py

@@ -145,6 +145,26 @@ class HttpCompressionTest(TestCase):
         self.assertEqual(response.headers['Content-Encoding'], b'gzip')
         self.assertEqual(response.headers['Content-Type'], b'application/gzip')
 
+    def test_process_response_gzip_app_octetstream_contenttype(self):
+        response = self._getresponse('gzip')
+        response.headers['Content-Type'] = 'application/octet-stream'
+        request = response.request
+
+        newresponse = self.mw.process_response(request, response, self.spider)
+        self.assertIs(newresponse, response)
+        self.assertEqual(response.headers['Content-Encoding'], b'gzip')
+        self.assertEqual(response.headers['Content-Type'], b'application/octet-stream')
+
+    def test_process_response_gzip_binary_octetstream_contenttype(self):
+        response = self._getresponse('x-gzip')
+        response.headers['Content-Type'] = 'binary/octet-stream'
+        request = response.request
+
+        newresponse = self.mw.process_response(request, response, self.spider)
+        self.assertIs(newresponse, response)
+        self.assertEqual(response.headers['Content-Encoding'], b'gzip')
+        self.assertEqual(response.headers['Content-Type'], b'binary/octet-stream')
+
     def test_process_response_head_request_no_decode_required(self):
         response = self._getresponse('gzip')
         response.headers['Content-Type'] = 'application/gzip'