refactor(jans-cedarling): enhance schema parser and entity creation implementation

[rmarinn]

Prepare

• Read PR guidelines
• Read license information

---

Description

This PR refactors the JSON schema struct and the Cedar entity-building logic to improve code readability and maintainability.

Target issue

target issue: #10513

closes #10513

Implementation Details

1. Introduction of EntityBuilder Struct

   • A new EntityBuilder struct encapsulates configuration and logic for creating entities.
   • This improves modularity, making the entity-building process more self-contained and easier to use across the codebase.
   • Also fixes fix(jans-cedarling): error during creating entity of type record #10545

2. Refactoring CedarSchemaJson

   • The CedarSchemaJson struct has been redesigned for better usability.
   • Deserialization and JSON schema parsing logic is now self-contained, reducing complexity for other callers.

3. Unit Tests

   • Added unit tests for the refactored components.
   • Future test cases will be easier to implement due to reduced coupling and improved struct/function design.

4. Improved Automatically adding entities to the context

   • We don't need to do the "lower_snake_case convention" for entity names anymore to automatically add them to the context.

Entity Builder

The EntityBuilder struct centralizes entity creation logic and holds configuration on initialization.

Usage:

pub struct EntityBuilder {
    schema: CedarSchemaJson,
    entity_names: EntityNames,
    build_workload: bool,
    build_user: bool,
}

Function implementations on this struct will have the entity creation code. Other modules will only have to use these functions.

impl EntityBuilder {
    pub fn new(
        schema: CedarSchemaJson,
        entity_names: EntityNames,
        build_workload: bool,
        build_user: bool,
    ) -> Self {
        // ...
    }

    pub fn build_entities(
        &self,
        tokens: &DecodedTokens,
        resource: &ResourceData,
    ) -> Result<AuthorizeEntitiesData, BuildCedarlingEntityError> {
        // ...
    }
}

---

Test and Document the changes

• Static code analysis has been run locally and issues have been fixed
• Relevant unit and integration tests have been added/updated
• Relevant documentation has been updated if any (i.e. user guides, installation and configuration guides, technical design docs etc)

Please check the below before submitting your PR. The PR will not be merged if there are no commits that start with docs: to indicate documentation changes or if the below checklist is not selected.

• I confirm that there is no impact on the docs due to the code changes in this PR.

[djellemah]

Here's a suggestion to clean this up

18326b7

[rmarinn]

looks good! i didn't realize we can use #[serde(untagged)] like that. i merged your suggestion in 98f99b6. thanks

[abaghinyan]

"desrializing" should be corrected to "deserializing"

[rmarinn]

fixed in 821958b

[abaghinyan]

as your function explicitly depend on serde_json::Value, maybe you can simplify it directly to

pub fn deserialize_to_string(value: Value) -> Result<String, String> {
    serde_json::from_value::<String>(value).map_err(|e| format!("Error: {e}"))
}

[rmarinn]

i tried this and this requires me to map the error into a serde::de::Error on all the places it's called.

        let kind = deserialize_to_string(kind).map_err(|msg| de::Error::custom(msg))?;

[djellemah]

I think quite a nice way is in Attribute::deserialize and other call-sites of deserialize_to_string to have

let kind = String::deserialize(&kind).map_err(de::Error::custom)?;

then deserialize_to_string is no longer necessary, and its value parameter does not have to be passed by copy.

[djellemah]

The ordering of entity here causes an intermittent test failure.
That's solvable using
entity.sort_by(|a,b| a.uid().cmp(&b.uid()) );
and obviously changing to let mut entitity = ...

I suspect that's a result of my suggested UnifyClaims change from yesterday.