feat: added response conversion node

[tanweersalah]

Description

Changes proposed in this pull request:

• Added response conversion functionality to finalizer

Related issue(s)

#276

[github-actions]

Note(s) for PR Auther:

• The integration test will be skipped for the PR. You can trigger it manually after adding the label: run-integration-test.
• The evaluation test will be skipped for the PR. You can trigger it manually after adding the label: evaluation requested.
• If any changes are made to the evaluation tests data, make sure that the integration tests are working as expected.
• If any changes are made to how to run the unit tests, make sure to update the steps for unit-tests in the create-release.yml workflow as well.

Note(s) for PR Reviewer(s):

• Make sure that the integration and evaluation tests are working as expected.

[muralov]

let's use TypedDict or Pydantic to parse:

class YAMLMetadata(TypedDict):
    namespace: str
    name: str

class YAMLConfig(TypedDict):
    metadata: YAMLMetadata
    kind: str

def _parse_yamls(self, yaml_config: str) -> Optional[YAMLConfig]:
    """
    Parse YAML string into Python object with error handling.
    Returns None if parsing fails.

    Args:
        yaml_config: YAML configuration string

    Returns:
        Parsed YAML object or None if parsing fails
    
    Example:
        >>> result = converter._parse_yamls("```yaml\\nkind: Deployment\\n```")
        >>> if result:
        >>>     print(f"Parsed YAML: {result}")
        >>> else:
        >>>     print("Failed to parse YAML")
    """
    try:
        if yaml_config[:7] == "```yaml":
            parsed_yaml = yaml.safe_load(yaml_config[8:-4])
        else:
            parsed_yaml = yaml.safe_load(yaml_config)
        
        return parsed_yaml

    except Exception as e:
        logger.error(
            f"Error while parsing the yaml : {yaml_config} , Exception : {e}"
        )
        return None

[muralov]

As I see if a yaml is generated but doesn't fit the schema, it returns None.
Have you observed the cases where partial yaml resources are generated? We can discuss this.

[tanweersalah]

As agreed will skip this

[muralov]

are both _generate_final_response and _convert_final_response methods are shown in Langfuse?

[tanweersalah]

i dont see either of them ,

https://langfuse.mur-nats.kymatunas.shoot.canary.k8s-hana.ondemand.com/project/cm4ty39mm000bifl0he3e61er/traces/1b0fd170-809b-47aa-82b7-b233a06db29b?timestamp=2025-01-23T15%3A34%3A09.422Z&observation=ca1b5345-a54d-4296-a05b-429fee764a86

[muralov]

Please rebase on main, there is conflict.

I'd add such instructions for prompts so that LLM outputs fit your logic. We can discuss this.

  1. For new resource deployments, wrap the YAML in <YAML-NEW> </YAML-NEW> tags
  2. For modifications to existing resources, wrap the YAML in <YAML-UPDATE> </YAML-UPDATE> tags
  3. Each YAML configuration must include:
     - apiVersion
     - kind
     - metadata (with at least name and namespace)
  4. Multiple YAML configurations should be separated into individual blocks with appropriate tags
  
    Example:
  <YAML-NEW>
  apiVersion: apps/v1
  kind: Deployment
  metadata:
    name: my-new-app
    namespace: default
  spec:
    ...
  </YAML-NEW>

  <YAML-UPDATE>
  apiVersion: v1
  kind: Service
  metadata:
    name: existing-service
    namespace: production
  spec:
    ...
  </YAML-UPDATE>