Superdesk-core

Added

- Add `id` to authors in NINJS output.
- Add high priority publish celery queue.
- Default spellchecker implementation using new api.
- AP Media API ingest.
- Implement `Unpublish` action.

Fixed

- Fix custom renditions generated during ingest not cropping.
- Do not expire items from production if the reference an assignment.
- Fix saved search reports showing more items than visible in UI.
- Edit image action done from article should not modify the original image.

Changed

- Update celery version to 4.3.

Added

- Add generic spellcheckers management
- Add Grammalecte French spellchecker support
- Add Leuven University Dutch spellchecker support
- Add basic generic spellchecker

Fixed

- Make media description field multi-line.
- Fix subject required validation failing in some cases.

Added

- Ingest embargoed info in NewsML G2 and publish it in NINJS format.
- Add data required for dateline to geonames search.
- Add collation support for case insensitive mongo indexes.
- Add support for localization.
- Add source and dest language info to translate macro.
- Add desk routing support.
- Backend implementation for knowledge base.
- Create custom decorator for test scenarios for executing `app:initialize_data`.
- Allow item attachments in content editor.
- Implement `on_replace` and `on_replaced` hooks for PUT requests.
- Log execution time during FTP ingest.
- Add docs for `manage.py` commands.
- Add section attribute to content profile fields.
- Allow to restrict a feeding service to no parser.
- Add default author/editor role config.
- Add widgets config to content profiles.
- Add new `HTTPFeedinfServiceBase` helper class.
- Add FTP reading limit for files.
- Pass `req.arg.params` to search proxy.

Changed

- Publish queue should allow `item_id` for any resource type.
- Update PyYAML version.

Fixed

- Fix subject name parsing in STT NewsML parser.
- Enabled related items widget to work with media from external source.
- Publishing fails for missing embedded item.
- Auto-published updates appear as brand new stories in Newsroom.
- Handle `ingest_provider` in saved search.
- Send ingested status to feeding service.
- Keep removed associations as `null`.
- Add missing ingest error logs to ingest dashboard.
- Fix `original_source` parsing in NINJS parser.
- Fix unknown error is thrown on publishing a story that contains URL.
- Ingest subject scheme in NINJS parser.
- Ignore etag during update when `IF_MATCH` is switched off.
- Display latest properties for related items.
- Remove `httmock` from setup dependencies.
- Add missing cvs for media contacts.
- Fix creating a vocabulary with same ID like previously deleted one.
- Fix backend-meta return value for unknow superdesk revision.

Added

- Add `manage.py` command for generating vocabularies from text files.
- Make geonames feature class configurable for place autocomplete.
- Add api for custom item schema field registration.
- Add `embargoed` field to content api item schema.
- Attach featured image to email output.
- Add `manage.py` commands to docs.
- Add guid to associations in WXR parser.

Changed

- Return user `_etag` on password change.
- Improve component version info for about screen.

Fixed

- Fix failing update ingest tests.
- Preserve legal and sms flags on associate as update action.
- Do not overwrite existing item associations during archive rewrite.
- Fix `CRLF` handling in WXR parser.
- Handle ingest expiry set to very big number.
- Updated item preserves parent's featured image even if it was changed.
- Fix unlocking with invalid renditions data.
- Fix unlocking of item with expired embargo.
- Stop using self closing tags for non-void elements in WXR parser.
- When translated item is corrected, send notification to translators.
- Fix mandatory subject in schema when not present in the editor.

Fixed

- Set default log level to `INFO` for `content_api` module.
- Decode filter in response when new saved search is created.
- Fix readthedocs build.
- Handle timeouts from geonames api.
- Fix issue with related items validation.
- Fix `NewsML G2` parser to handle STT content.
- Various fixes in Wordpress import.
- Fix `on_item_locked` signal not passing updated item.
- Improvements for DPA feed parser.

Added

- Add `app:flush_elastic_index` manage command.
- Add `local_domains` cv for detecting local links.
- Add links to items in saved search report.
- Add ability to send email attachments.
- Add sync mode to `ingest:update` command.
- Add support for related items in `NINJS` output format.
- Add description and tags to vocabularies in metadata settings.

Changed

- Include lock/unlock info in archive history.
- Desk members from `/desks` are sorted by name.
- Use aggregations from query param sor class instance variable in search.
- Filter out `do not show` vocabularies from content profile editor.
- Replace `single_value` with `selection_type` in vocabularies.
- Don't store publish formatter instance in registry.
- Don't store feeding service instance in ingest.

TBD