The API for the gzip module is convenient for reading gzipped file but is rather awkward for reading bytes downloaded from a socket:

  >>> import gzip, io
  >>> from urllib.request import urlopen, Request
  >>> u = urlopen(Request('http://www.nytimes.com', headers={'Accept-Encoding': 'gzip'}))
  >>> content = gzip.GzipFile(fileobj=io.BytesIO(u.read()), mode='rb').read()

It would be better if the last line could be written:

  >>> content = gzip.decompress_file(u)