mcsimps2 / redpanda Goto Github PK

You can currently set it as <doc>.id = <ObjectID> and then do <doc>.save(). You may also be able to set in in the constructor. This behavior should be documented.

This will make the codebase more lightweight. Lodash is only used a couple times and is not necessary.

Instead of “await obj.foreignRef”, instead have “obj.foreignRef.fetch()” return the promise, and load this class with just the id so you can do “obj.foreignRef.id” without having to do a fetch.

Right now, only collections are supported.

This code doesn't look like it'll allow null values

if (schema[property.toString()].describe()['type'] === 'dbref') {
                        if (typeof value === 'string') {
                            
                            return Reflect.set(target, '__id__' + property.toString(), value)
                                    && Reflect.set(target, '__obj__' + property.toString(), null);
                        }
                        else if (value instanceof Document) {
                            return Reflect.set(target, '__id__' + property.toString(), value.id)
                                    && Reflect.set(target, '__obj__' + property.toString(), value);
                        }
                        else {
                            throw new Error("Can only set a foreign reference with an id or Document subclass instance")
                        }
                    }

For example, a dictionary field that maps to IDs. Right now, support is only for individual fields and arrays of foreign references.

Load should have the option of maxDepth for recursive loading.

Instead of making the collection using .toLowerCase(), the collection should be separated by underscores.

Need to find a way to support the Firestore function .runTransaction() in an ODM like manner

Add support to recursively save/update/reload documents. (And document ability to reload).

Probably want to do it as so when specifying a schema

{
   first_name: RedPanda.types.string().required(),
   street_address: {
         default: "",
         type: RedPanda.types.string().required()
   }
}

There is support for non-RedPanda types.

For example, when specifying a schema, it is possible to do
{
business: RedPanda.types.dbref().collection('business')
}

instead of {
business: RedPanda.types.dbref().collection(BusinessDocument)
}
because we may not have created a BusinessDocument type (perhaps we are migrated to RedPanda ODM). In this case, a snapshot is returned instead of a BusinessDocument. Add documentation about this.

Also add documentation about the ability to access the underlying doc reference, snapshot, collection, etc...