veritone-json-schemas
changelog
Version 0.1.0
- initial version published to NPM
Version 0.1.1
- export a map of the validation contracts to the validation functions
as VALIDATOR
- basic README.md
- cleanup for object validation contract
Version 0.1.2
Version 0.1.3
- revise TRANSCRIPT validator to make words optional
- add { processed: } to positive validations to show only the
processed fields
- dynamic test loader
- empty words are no longer valid
Version 0.1.4
- Fix { processed } for transcript.
- More tests, addd vtn-standard headers to transcript
Version 0.1.5
- Change GUIDS from number to string
- Some minor changes in descriptions
Version 0.1.6
- Changes to type enum in series / object results
- Support for object recognition/detection
- More examples for transcript
Version 0.1.7
- Make object required if a series array is present
- Add test cases to prevent empty objects in summaries
- Add test case to make sure that an empty object summary with no
recognition/detection is a valid result.
Version 1.0.0
- BREAKING CHANGE Adds new restrictions to existing
object and transcription validation contracts
that caused some previously valid examples to now be categorized as
invalid examples. All of these centered around disallowing the inclusion
of additional properties not specified by the schema.
- Add a number of new json-schema validators for engine categories
centered around text analytics and translation.
- See 160904a
for full details.
Version 1.0.1
- Fixes the schema IDs to reflect their proper public hosting
location.
- Depending on your use of the json-schemas, this could be considered
a breaking change if you previously referenced the schema IDs.
Version 1.0.2
- Changes to examples to integrate nicely with our docs site.
Version 1.1.0
- Adds support for Anomaly Detection Engines
Version 1.1.1