Looking towards future development, it would be far quicker and easier to develop new tools if the data structure was represented using an ORM with proper versioning and automatic upgrades. I've designed a far more flexible general content representation, which would also necessitate a restructure of how processors work (for a future task).
In addition to this general layout, specific processors should manage their own tables/databases, specific to their use case. Currently, this is only needed for the PhashProcessor, for which a schema has already been designed.
Looking towards future development, it would be far quicker and easier to develop new tools if the data structure was represented using an ORM with proper versioning and automatic upgrades. I've designed a far more flexible general content representation, which would also necessitate a restructure of how processors work (for a future task).
In addition to this general layout, specific processors should manage their own tables/databases, specific to their use case. Currently, this is only needed for the PhashProcessor, for which a schema has already been designed.