import csv
from collections import defaultdict
from datetime import date, datetime

from src.generate_data import generate_data


def test_record_count_and_header(tmp_path):
    path = tmp_path / "orders.csv"
    generate_data(path)
    with path.open(encoding="utf-8", newline="") as source:
        records = csv.reader(source)
        assert next(records) == ["order_id", "order_date", "category", "quantity", "unit_price"]
        assert sum(1 for _ in records) == 1500


def test_repeated_generation_is_byte_identical(tmp_path):
    path = tmp_path / "orders.csv"
    generate_data(path)
    first_run = path.read_bytes()
    generate_data(path)
    assert path.read_bytes() == first_run


def test_planned_examples_are_present_in_generated_csv(tmp_path):
    path = tmp_path / "orders.csv"
    generate_data(path)
    with path.open(encoding="utf-8", newline="") as source:
        rows = list(csv.DictReader(source))

    expected_values = {
        "order_id": {""},
        "order_date": {"", "2024-02-30", "03/15/2024"},
        "category": {"", "  тОВАРЫ   для  ДОМА  "},
        "quantity": {"", "1.5", "0", "-2", "три"},
        "unit_price": {"", "0", "-12.50", "бесплатно", "12.340", "12,50"},
    }
    for field, expected in expected_values.items():
        assert expected <= {row[field] for row in rows}, field

    by_id = defaultdict(list)
    for row in rows:
        by_id[row["order_id"]].append(row)
    first, repeat = by_id["1"]
    assert first == repeat

    iso, local = sorted(by_id["2"], key=lambda row: "." in row["order_date"])
    assert date.fromisoformat(iso["order_date"]) == datetime.strptime(
        local["order_date"], "%d.%m.%Y"
    ).date()
    assert iso == local | {"order_date": iso["order_date"]}

    first, conflict = by_id["3"]
    assert int(first["quantity"]) != int(conflict["quantity"])
    assert first == conflict | {"quantity": first["quantity"]}