Merge pull request #433 from aio-libs/fix-gzip

bring back original Accept-Encoding

aiobotocore/endpoint.py

@@ -224,6 +224,19 @@ def __init__(self, host,
 
     @asyncio.coroutine
     def _request(self, method, url, headers, data):
+        # Note: When using aiobotocore with dynamodb, requests fail on crc32
+        # checksum computation as soon as the response data reaches ~5KB.
+        # When AWS response is gzip compressed:
+        # 1. aiohttp is automatically decompressing the data
+        # (http://aiohttp.readthedocs.io/en/stable/client.html#binary-response-content)
+        # 2. botocore computes crc32 on the uncompressed data bytes and fails
+        # cause crc32 has been computed on the compressed data
+        # The following line forces aws not to use gzip compression,
+        # if there is a way to configure aiohttp not to perform decompression,
+        # we can remove the following line and take advantage of
+        # aws gzip compression.
+        # https://github.com/boto/botocore/issues/1255
+        headers['Accept-Encoding'] = 'identity'
         headers_ = MultiDict(
             (z[0], text_(z[1], encoding='utf-8')) for z in headers.items())