Udata-ckan

- Only store `url` field in `remote_url` extra if this is an URL otherwise store it in `ckan:source` [30](https://github.com/opendatateam/udata-ckan/pull/30)
- Properly handle geometry errors [31](https://github.com/opendatateam/udata-ckan/pull/31)
- Improve extras parsing [32](https://github.com/opendatateam/udata-ckan/pull/32):
- Skip empty extras
- Parse update frequencies as RDF URI or udata frequency identifier
- Parse `spatial-text` matching a known zone name or slug
- Store unknown `spatial-uri`, `spatial-text` and `frequency` as `ckan:spatial-uri`, `ckan:spatial-text` and `ckan:frequency`

- Test against a real CKAN instance [23](https://github.com/opendatateam/udata-ckan/pull/23)
- Allows to filter on Organizations and Tags [26](https://github.com/opendatateam/udata-ckan/pull/26)
- Register `ckan:` prefixed extras [28](https://github.com/opendatateam/udata-ckan/pull/28)

- Fix packaging [2](https://github.com/opendatateam/udata-ckan/pull/2)
- Make use of [udata pytest plugin](opendatateam/udata1400) [3](https://github.com/opendatateam/udata-ckan/pull/3)
- Enable the `ckan` plugin in test (plugin needs to be enabled to use the harvester) [8](https://github.com/opendatateam/udata-ckan/pull/8)

- Initial release