backend/endpoints/audit.py

import io
import logging
from typing import Optional
from fastapi import APIRouter, Depends, HTTPException, UploadFile, File, Form
from core.subscription.middleware import verify_token
from agents.auditor import audit_final_document, GlobalAuditOutput
from pypdf import PdfReader
import docx

logger = logging.getLogger(__name__)

router = APIRouter(prefix="/api/audit", tags=["audit"])

def extract_text_from_file(filename: str, content: bytes) -> str:
    ext = filename.split(".")[-1].lower() if "." in filename else ""
    if ext == "pdf":
        try:
            reader = PdfReader(io.BytesIO(content))
            text = []
            for page in reader.pages:
                text.append(page.extract_text() or "")
            return "\n".join(text)
        except Exception as e:
            logger.error(f"Error parsing PDF: {e}")
            raise HTTPException(status_code=400, detail="Nie można odczytać pliku PDF.")
    elif ext == "docx":
        try:
            doc = docx.Document(io.BytesIO(content))
            return "\n".join([para.text for para in doc.paragraphs])
        except Exception as e:
            logger.error(f"Error parsing DOCX: {e}")
            raise HTTPException(status_code=400, detail="Nie można odczytać pliku DOCX.")
    else:
        try:
            return content.decode("utf-8")
        except:
            raise HTTPException(status_code=400, detail="Nieobsługiwany format pliku. Obsługiwane są PDF, DOCX oraz TXT.")


@router.post("/reverse-audit", response_model=GlobalAuditOutput)
async def reverse_audit(
    file: UploadFile = File(...),
    program_name: Optional[str] = Form("Nieznany program"),
    token_data: dict = Depends(verify_token)
):
    """
    Endpoint obsługujący funkcję Reverse Audit.
    Przyjmuje plik PDF/DOCX wniosku i uruchamia audytor GraphRAG.
    """
    if not file:
        raise HTTPException(status_code=400, detail="Brak pliku.")

    file_content = await file.read()
    extracted_text = extract_text_from_file(file.filename, file_content)

    if not extracted_text or len(extracted_text.strip()) < 50:
        raise HTTPException(status_code=400, detail="Plik jest pusty lub nie udało się odczytać wystarczającej ilości tekstu.")

    try:
        audit_result = audit_final_document(
            project_id="reverse_audit",
            program_name=program_name,
            content=extracted_text,
            enable_multi_perspective=True,
            is_external_audit=True
        )
        return audit_result
    except Exception as e:
        logger.error(f"Error in reverse_audit: {e}")
        raise HTTPException(status_code=500, detail=f"Błąd podczas przeprowadzania audytu: {str(e)}")