LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows
Python ai-agentsdata-engineeringdocument-aigenerative-aiidpjson-extractionllmmcp-serverocrpdf-extractionprompt-engineeringstructured-output
import logging
from typing import Any
from plugins import get_plugin
from api_v2.postman_collection.dto import PostmanCollection
logger = logging.getLogger(__name__)
class ApiDeploymentDTORegistry:
_dto_class: Any | None = None # Store the selected DTO class (cached)
@classmethod
def load_dto(cls) -> Any | None:
"""Load DTO from plugin or return default PostmanCollection.
Checks if the api_dto plugin is available and gets the Postman DTO
via the service, otherwise returns the base PostmanCollection class.
"""
plugin = get_plugin("api_dto")
if plugin:
service = plugin["service_class"]()
return service.get_postman_dto()
return PostmanCollection
@classmethod
def get_dto(cls) -> type | None:
"""Returns the first available DTO class, or None if unavailable."""
return cls.load_dto()