Ppx

Added
- New `timeout` parameter for most functions and classes. This specifies the
maximum amount of time to wait for a response from the server.

Changed
- The backend for MassIVE now uses the GNPS API to list files and projects,
only falling back to scraping the FTP server on failure. This should make
it much faster. Thanks mwang87!
- Files and projects are now returned in sorted order.

Fixed
- Poor connections with PRIDE were leading to a number of occasional errors.
Multiple reconnect attempts are now tried for a wider variety of FTP
operations.

Fixed
- Downloading files is now more robust. ppx will now retry FTP connections up
to 10 times if the connection was dropped or refused.
- Partial downloads are now continued automatically by comparing the local
file size to the remote file size.

Fixed
- The PRIDE REST API was not yielding all of the available files on the their
FTP server for a project. We changed the backend use the FTP server directly
instead. **Note that this may change the number, identity, and order of the
file that were previously returned for PRIDE projects!**
- Small documentation updates.

Added
- Caching of the remote files and directories found for a project. If a
project's `fetch` attribute is `False`, then we'll rely on this cached
data, so long as it is available. Setting `fetch=True` will always refresh
the data from the repository.

Changed
- **We did a complete rework of the API!** This will break nearly all previous
code using ppx, but greatly improves its versatility.
See the [docs](https://ppx.readthedocs.io) for more details
- Updated the build to align with
[PEP517](https://www.python.org/dev/peps/pep-0517/)

Added
- A command line interface for downloading files from PRIDE and MassIVE
projects.
- Additional unit tests.
- A ppx logo
- This changelog.
- ppx is now available on bioconda!