财务 收据匹配ing 代理

import pytesseract, json, re, os

OCR_PATH = "/data/receipts/"
LEDGER_PATH = "/data/expenses.json"

def ocr_receipt(file):
    text = pytesseract.image_to_string(file)
    amount = re.search(r'RM\s*([\d,]+)', text)
    return {"amount": int(amount.group(1).replace(',', ''))} if amount else None

def match_receipts():
    matches = []
    for r in os.listdir(OCR_PATH):
        rcpt = ocr_receipt(OCR_PATH + r)
        entry = next((e for e in ledger if e["receipt_id"] == r), None)
        status = "Matched" if entry and entry["amount"] == rcpt["amount"] else "Discrepancy"
        matches.append({"receipt": r, "entry": entry, "status": status})
    return matches

matches = match_receipts()
print(json.dumps(matches, indent=2))