import csv
import re
import sys
from collections import Counter
from datetime import datetime
from decimal import Decimal
from pathlib import Path
# convert revolut account statement csv exports into the same shape as nlb.py / trade_republic.py
STATEMENTS = "account-statement*.csv"
IBAN = re.compile(r"\b[A-Z]{2}\d{2}[A-Z0-9]{11,30}\b")
FIELDS = ["date", "desc", "account", "amount", "balance", "currency", "type"]


def decimals(text):
    return len(text.partition(".")[2])


def scale(source):
    """Revolut writes JPY without decimals and EUR with two, so follow the file."""
    places = {}
    for r in source:
        seen = [decimals(r[c]) for c in ("Amount", "Fee", "Balance") if r[c]]
        places[r["Currency"]] = max(seen + [places.get(r["Currency"], 0)])
    return places


def extract_transactions(path):
    print(path)
    with open(path, newline="") as f:
        source = list(csv.DictReader(f))
    source.sort(key=lambda r: r["Completed Date"])
    places = scale(source)

    rows = []
    skipped = Counter()
    for r in source:
        if r["State"] != "COMPLETED" or not r["Balance"]:
            skipped[r["State"]] += 1
            continue
        digits = places[r["Currency"]]
        fee = Decimal(r["Fee"] or 0)
        # the fee is deducted alongside the amount rather than booked separately
        amount = Decimal(r["Amount"]) - fee
        desc = r["Description"]
        if fee:
            desc += f" (fee {fee:.{digits}f})"
        account = IBAN.search(desc)
        rows.append({
            "date": datetime.strptime(r["Completed Date"][:10], "%Y-%m-%d").strftime("%d.%m.%y"),
            "desc": desc,
            "account": account.group() if account else "",
            "amount": f"{amount:+.{digits}f}",
            "balance": f"{Decimal(r['Balance']):.{digits}f}",
            "currency": r["Currency"],
            "type": r["Type"],
        })
    for state, count in sorted(skipped.items()):
        print(f"  skipped {count} {state or 'unfinished'} transactions")
    return rows


def check_balances(rows, path):
    """Every balance must be the previous one plus the transaction amount."""
    for currency in dict.fromkeys(row["currency"] for row in rows):
        chain = [row for row in rows if row["currency"] == currency]
        for prev, row in zip(chain, chain[1:]):
            expected = Decimal(prev["balance"]) + Decimal(row["amount"])
            if expected != Decimal(row["balance"]):
                print(f"  warning: {path.name} {row['date']} {row['desc'][:40]!r} "
                      f"balance {row['balance']} != {expected}")


def main():
    target = Path(sys.argv[1])
    paths = sorted(target.glob(STATEMENTS)) if target.is_dir() else [target]
    out_path = target.with_suffix(".csv")
    if out_path in paths:
        sys.exit(f"{out_path} would overwrite the statement it reads; pass the directory instead")

    all_rows = []
    for path in paths:
        rows = extract_transactions(path)
        check_balances(rows, path)
        all_rows.extend(rows)

    with open(out_path, "w", newline="") as f:
        writer = csv.DictWriter(f, fieldnames=FIELDS)
        writer.writeheader()
        writer.writerows(all_rows)

    print(f"Wrote {len(all_rows)} transactions to {out_path}")


if __name__ == "__main__":
    main()