LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows
Python ai-agentsdata-engineeringdocument-aigenerative-aiidpjson-extractionllmmcp-serverocrpdf-extractionprompt-engineeringstructured-output
from typing import Any
from requests import Response
class X2TextUtil:
@staticmethod
def get_value_for_key(key: str, form_data: dict[str, Any]) -> str:
value: str = form_data.pop(key, None)
return value
@staticmethod
def get_text_content(json_response: dict[str, Any]) -> str:
combined_text: str = "\n".join(
item["text"]
for item in json_response # type:ignore
)
return combined_text
@staticmethod
def read_response(response: Response) -> dict[str, Any]:
if response.headers.get("Content-Type") == "application/json":
return response.json() # type: ignore
else:
return {"message": response.text}