Introduce query engine and schema generation #55
Conversation
The SpiceDB schema for permissions-api was previously hardcoded in text. This commit adds a template for the SpiceDB schema as well as data types for defining resources and actions. An instance of the Schema struct is currently hardcoded in the "schema" command to match existing output. Signed-off-by: John Schaeffer <[email protected]>
Signed-off-by: John Schaeffer <[email protected]>
Signed-off-by: John Schaeffer <[email protected]>
jnschaeffer
force-pushed
the
schema-generation
branch
from
3c286bd
to
6c61524
Compare
| } | ||
| {{end}} | ||
| `)) | ||
| ) |
There was a problem hiding this comment.
Should we already refer to the prefix as a namespace? This way it's easier to read and interpret that it's the intention.
| Optional: true, | ||
| Name: "tenant", | ||
| Type: "tenant", | ||
| Optional: true, |
There was a problem hiding this comment.
Given that the subject resource is being deleted from here, do you have a notion of how we'll register subjects? Would we auto-create them as role assignments happen?
There was a problem hiding this comment.
This got me thinking - RESTful management of resources may not necessarily be what we want here. Resources in SpiceDB are defined only in terms of their relationships, so it's not possible to create a resource that doesn't have relationships.
It may be better to rework this so that the resource API doesn't manage resources, but rather manages roles and relationships.
Signed-off-by: John Schaeffer <[email protected]>
|
|
||
| func init() { | ||
| rootCmd.AddCommand(schemaCmd) | ||
|
|
||
| schemaCmd.Flags().BoolVar(&dryRun, "dry-run", false, "dry run: print the schema instead of applying it") |
| schemaCmd = &cobra.Command{ | ||
| Use: "schema", | ||
| Short: "write the schema into SpiceDB", | ||
| Run: func(cmd *cobra.Command, args []string) { |
There was a problem hiding this comment.
total nit but I prefer RunE over a bunch of logger.Fatal - not sure how you feel about that (yes, i know you didn't really change this 😆 )
There was a problem hiding this comment.
I think it's a good suggestion for a subsequent PR. Can you open a GH issue? Would be a nice low hanging fruit.
| } | ||
| } | ||
|
|
||
| func (e *Engine) SubjectHasPermission(ctx context.Context, subject types.Resource, action string, resource types.Resource, queryToken string) error { |
There was a problem hiding this comment.
this signature be nicer as this IMO:
func (e *Engine) SubjectHasPermission(ctx context.Context, subject, resource types.Resource, action string, queryToken string) error
Co-authored-by: E Camden Fisher <[email protected]> Signed-off-by: John Schaeffer <[email protected]>
One of the challenges of permissions management in permissions-api is that we will (likely) need to persist data in two places: SpiceDB for permissions management, and CRDB for other things like update times, role names, and so on. To make this easier in the long run, this PR introduces a few things:
query.Engine: An engine for making authorization queries. The package name might change in the future, but since this is where the authorization logic lives today I figured I would keep itspicedbx.GenerateSchema: A function for generating a SpiceDB schema based on resource type definitionsAPI handlers have also been updated to explicitly create and assign roles rather than using the
querypackage to do so. This logic needs some refining, however, which will happen as this PR matures.This PR does not currently have the following:
query.Enginetestsspicedbx.GenerateSchematests