Flattentool

Added

- Add support for `"format": "date"` in a JSON schema

Changed

- All code has had black and isort applied. These have been added to Travis.

Fixed

- Remove extra lines in CSVs under Windows https://github.com/OpenDataServices/flatten-tool/pull/350

Added

- Add .ods (OpenDocument Spreadsheet) file format support https://github.com/OpenDataServices/flatten-tool/pull/326

Changed

- Drop Python 2 support https://github.com/OpenDataServices/flatten-tool/issues/299
- Documentation on Change Logs, Versioning and PyPi updated to match current practice.

Fixed

- Add BadXLSXZipFile exception type for BadZipFile exception when handling broken XLSX files related to cove issue https://github.com/OpenDataServices/cove/issues/514

Added

- Add an XML declaration to outputted XML files https://github.com/OpenDataServices/cove/issues/1206

Added

- --preserve-fields option to the cli so the user can specify which fields to keep when flattening, instead of defaulting to all of them

Fixed

- Correct cell source map when hashcomments is used.

Changed

- Using --rollup with `flatten` accepts fields to roll up as input directly and via a file, as well as via a schema.
- Fixed pytest version at <5 because of breaking changes to the tests.

Fixed

- Ignore NoneType in json instead of giving a TypeError.