Prevent automatic decompression of gzipped files, fix for #710

This is a fix for Issue #710. It derives from the filename if the target file is a gzip file (i.e. has a .gz suffix) and uses the raw file if that is the case.

I also added a test for it, but could not manage to integrate it into the tests.py script. It works when I call snakemake in the test folder and when I import tests.py and run it (from IPython).

For the test I reused the landsat dataset found in the test_remote_gs folder. Both files (plain text and gzipped version of the file) are downloaded from bitbucket and compared using diff. I hope this is solid for most use cases.

The paths still point to my fork and need to be updated.

Commits

• Added test for downloading zipped and unzipped files via http and compare them.

• Finished test that fails with the current setup.

• Consistent naming of files.

• Added fix for the HTTP remote provider automagically decompressing .gz files.

• Added comment about test setup