From 10a2c46e12e54e666de743b9ceca09c18370ccec Mon Sep 17 00:00:00 2001 From: Paul Tremberth Date: Mon, 20 Jun 2016 13:39:37 +0200 Subject: [PATCH] [HttpCompressionMiddleware] Do not decompress binary/octet-stream responses --- scrapy/utils/gz.py | 9 +++++++-- ...st_downloadermiddleware_httpcompression.py | 20 +++++++++++++++++++ 2 files changed, 27 insertions(+), 2 deletions(-) diff --git a/scrapy/utils/gz.py b/scrapy/utils/gz.py index d035f9fdf..b2f737193 100644 --- a/scrapy/utils/gz.py +++ b/scrapy/utils/gz.py @@ -53,5 +53,10 @@ def gunzip(data): def is_gzipped(response): """Return True if the response is gzipped, or False otherwise""" - ctype = response.headers.get('Content-Type', b'') - return ctype in (b'application/x-gzip', b'application/gzip') + ctype = response.headers.get('Content-Type', b'').lower() + cenc = response.headers.get('Content-Encoding', b'').lower() + return ( + ctype in (b'application/x-gzip', b'application/gzip') or + (ctype in (b'application/octet-stream', b'binary/octet-stream') and + cenc in (b'gzip', b'x-gzip')) + ) diff --git a/tests/test_downloadermiddleware_httpcompression.py b/tests/test_downloadermiddleware_httpcompression.py index 24955a515..b2426946d 100644 --- a/tests/test_downloadermiddleware_httpcompression.py +++ b/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'