- Specify qcode type in vocabulary schema. - Fix newsmlg2/ninjs parsers to handle sample iptc data. - Fix error sending on validation error in metadata. - Add signature header to assets http push.
Added
- Add BBC LDRS service and ninjs parser. (by CaerphillyMediaLtd) - Add `places_autocomplete` resource for using geonames. - Add prepopulate data for planning e2e tests. - Add support for `planning_types` in `app:initialize_data` cmd. - It should be possible to restrict feed parsers for a service. - Add new media editor endpoint.
1.19
Fixed
- Fix priority and urgency qcodes schema. - Fix vocabularies not validating item if unique field is missing.
Added
- Add event/planning item types to content filters. - Add docs how to create an output formatter. - Add schema fields for preffered cv items. - Add dev server config for content api.
Changed
- Push notification when new vocabulary is created. - Feeding services labels and fields should be set server side, not client. - Change email address field not to be case sensitive. - Default `unique_field` to `qcode` when suitable in vocabulary. - Enable getting image/video item id form image/video url.
1.18
Fixed
- Modify json loads to handle arrow `ParserError`. - Fix opening published image takes too much time.
Added
- Add tests for custom media multi-items ninjs output.
Changed
- Set `qcode` as unique field for `genre`, `priority`, `replace_words` and `annotation_types` cvs.
1.17
Fixed
- Allow item to be archived if the associated item is not expired. - Provide `_type` when fetching single item from api. - Empty qcode was not checked for some vocabularies. - Parse abstract metadata in STT parser. - Use first 100 body characters as headline if no headline is present in Ritzau feed.
Added
- Add endpoint for auto suggestions for contact organisations names. - Add new user type `support`. - Add config to toggle off error notification emails globally.
Changed
- Change email password field not to be case sensitive. - Pass destination desk/stage to incoming stage macro. - Make user optional when marking for a desk.
1.16
Fixed
- Use search providers proxy for saved search validation. - Allow contentapi elastic index re-indexing using rebuild command. - Map caption to `description_text` in image iptc parser.
Added
- Add `advanced_search` field to search provider schema. - Add `monitoring_default_view` field to desk schema. - Add macro to set item in progress when ingesting file with correspoind `assign_id`.
Changed
- Raise `IngestFileError` on ingest error. - Replace `editor_state` by `fields_meta` in item schema. - Allow forcing update of specific vocabulary using init command.
1.15
Fixed
- Use file extension when getting binary from amazon media storage. - Parse authors when ingesting ninjs. - Fix place elastic mapping to be consistent across resources. - Fix ftp not re-ingesting same file. - Avoid ssl verification on sentry. - Remove place field from Ritzau parser. - Fix ingest item expiry handling. - Add language to elastic aggregations when `apps.languages` is enabled. - Fix exception catching in ingest when content expiry is 0. - Remove content expiry from ingest settings if its <0.
Added
- Add source and ednote with tweet url to twitter ingest. - Add image feed ingest. - Add AP ingest service. - Add `--dir` param to `app:prepopulate` command.
Changed
- Use content profile for validation on auto-publishing. - Remove annotations processing from ninjs formatter. - Publish media embedded in item on item publishing.