api/receipt_routes.py

from fastapi import APIRouter, File, UploadFile, HTTPException, Request, Depends
from receipt_processor.google_ocr import GoogleVisionOCR
from receipt_processor.parsers.parser_selector import ParserSelector
from db.receipt_repository import ReceiptRepository
from utils.rate_limiter import RateLimiter
import re

# Initialize OCR and parser selector
rate_limiter = RateLimiter(max_requests=200)
receipt_repository = ReceiptRepository()
ocr_processor = GoogleVisionOCR()
parser_selector = ParserSelector()

router = APIRouter(prefix="/receipts", tags=["Receipt Processing"])

@router.post("/scan")
async def process_receipt(
    request: Request,
    file: UploadFile = File(...),
    client_ip: str = Depends(rate_limiter.check_rate_limit)
):
    try:
        print(f"Received file: {file.filename} ({file.content_type})")
        
        # Read the file content
        content = await file.read()
        print(f"File size: {len(content)} bytes")
        
        # Extract text using Google OCR
        extracted_text = ocr_processor.extract_text(content)
        print(f"Extracted text length: {len(extracted_text)} chars")
        
        if not extracted_text:
            raise HTTPException(400, "No text extracted from image")

        # Select and use appropriate parser
        parser = parser_selector.get_store_parser(extracted_text)
        print(f"Using parser: {parser.__class__.__name__}")
        
        parsed_receipt = parser.parse(extracted_text)
        print("Parsing completed successfully")
        
        # Try to extract date from the receipt
        receipt_date = None
        date_match = re.search(r'(\d{2}[./]\d{2}[./]\d{2,4})', extracted_text)
        if date_match:
            receipt_date = date_match.group(1)
        
        # Store the receipt in Supabase
        receipt_repository.create_receipt_request(
            receipt_image=content,
            parsed_data=parsed_receipt,
            receipt_date=receipt_date,
            submission_ip=client_ip
        )
        
        return {
            "status": "success",
            "message": "Receipt submitted successfully and pending review."
        }

    except HTTPException as e:
        raise e
    except Exception as e:
        print(f"ERROR: {str(e)}")
        raise HTTPException(500, f"Receipt processing error: {str(e)}")