Pyclowder

This version adds Clowder 2 support and removes the old method of extractor registration in favor of reliance on
heartbeats.

Added

- api/v1 and api/v2 code split for back compatibility as v2 is introduced.
- new simplified ClowderClient is used in the new split endpoints for future refactoring.

Removed

- remove RABBITMQ_EXCHANGE parameter and REGISTRATION_URL parameter.
- remove DatasetsAPI and object-oriented ClowderClient.

When extractors download a file from clowder it will ask clowder to not track that download.
This will result in only those donwloads to be counted by users, not extractors.

Changed

- Ask not to track a download from an extractor.

Unreleased

Added
- Add support for `EXTRACTOR_KEY` and `CLOWDER_EMAIL` environment variables to register
an extractor for just one user.

This will change how clowder sees the extractors. If you have an extractor, and you specify
the queue name (eiter as command line argument or environment variable) the name of the
extractor shown in clowder, will be the name of the queue.

Fixed

- both heartbeat and nax_retry need to be converted to in, not string

Changed

- when you set the RABBITMQ_QUEUE it will change the name of the extractor as well in the
extractor_info document. [47](https://github.com/clowder-framework/pyclowder/issues/47)
- environment variable CLOWDER_MAX_RETRY is now MAX_RETRY

Changed

- updated pypi documentation

Fixed

- extractor would fail on empty dataset download [36](https://github.com/clowder-framework/pyclowder/issues/36)

Added

- ability to set the heartbeat for an extractractor [42](https://github.com/clowder-framework/pyclowder/issues/42)

Changed

- update wordcount extractor to not use docker image
- using piptools for requirements

Added

- Add `--max_retry` CLI flag and `CLOWDER_MAX_RETRY` environment variable.

Changed

- updated all of the requirements to latest versions
- updated github actions to automatically create releases
- use thread.daemon = True to fix python 3.10