We already have a pretty generic docker image that should handle most use-cases:
- allows mounting config file
- allows overriding configuration values with
G_env variables, see: https://guillotina.readthedocs.io/en/latest/developer/commands.html?highlight=override#overridding-configuration
The helm chart should have support for configuring
n number of databases and all the settings that guillotina supports. It would package these configuration settings into a
config.json(or yaml if you prefer) that is mounted and referenced in the command arguments.
Make sure to provide the ability to also customize the image because people using this will likely be customizing the docker image used to package other things with guillotina(like customize addons).
This helm chart would serve as a generic chart to be able to deploy guillotina with.
Requests coming to the API were not being validated, task was to validate all the request hitting the API. Jsonschema is already defined for most of the endpoints, only a validation wrapper was required to be added to all the endpoint(along with jsonschema for missing endpoints and testcases for validation). Now all the request hitting being validated against a well defined schema and through 412(Precondition failed) along with proper error in case of a bad request.
Added validation is a wrapper over all the endpoint causing too much redundancy in code base because its need to be added to all the endpoint. Next thing is to replace the validation with a decorator, which will validate the request automatically and will have very less code redundancy