# retoor <retoor@molodetz.nl>

from datetime import datetime, timedelta, timezone

import pytest

from devplacepy.services.sync.normalize import (
    clamp_instant,
    clean_text,
    fingerprint,
    instant_iso,
    normalize_origin,
    normalize_origin_uid,
    normalize_username,
    parse_instant,
    username_candidates,
)
from devplacepy.services.sync.spec import LIMITS


@pytest.mark.parametrize(
    "raw,expected",
    [
        ("Other.Example.ORG", "other.example.org"),
        ("  a1  ", "a1"),
        ("devplace-net", "devplace-net"),
    ],
)
def test_origin_is_normalized(raw, expected):
    assert normalize_origin(raw) == expected


@pytest.mark.parametrize("raw", ["", "a", "-bad", "bad-", "a..b", "has space", "x" * 64, "ünï.org"])
def test_origin_is_rejected(raw):
    with pytest.raises(ValueError):
        normalize_origin(raw)


@pytest.mark.parametrize("raw", ["", "  ", "a b", "x" * 65, "a/b"])
def test_origin_uid_is_rejected(raw):
    with pytest.raises(ValueError):
        normalize_origin_uid(raw)


def test_origin_uid_accepts_integers_and_uuids():
    assert normalize_origin_uid(42) == "42"
    assert normalize_origin_uid("01a139e9-a2f0-7fa1-8635-1a98fbf95e3d")


def test_clean_text_strips_control_characters_and_newlines():
    assert clean_text("a\x00b\r\nc\rd\x07", multiline=True) == "ab\nc\nd"
    assert clean_text("a \n  b\t c", multiline=False) == "a b c"


def test_username_always_valid_across_inputs():
    samples = ["", "a", "ab", "Ünï cøde!", "x" * 80, "--__--", "john.doe@mail", "日本語", "ok_name-1"]
    for sample in samples:
        name = normalize_username(sample, "uid-1")
        assert LIMITS.min_username_length <= len(name) <= LIMITS.max_username_length
        assert all(ch.isalnum() or ch in "_-" for ch in name) and name.isascii()


def test_username_is_stable_for_same_input():
    assert normalize_username("a", "u1") == normalize_username("a", "u1")
    assert normalize_username("a", "u1") != normalize_username("a", "u2")


def test_username_candidates_are_distinct_and_bounded():
    seen = []
    for candidate in username_candidates("x" * 32, "other.example.org"):
        seen.append(candidate)
        assert len(candidate) <= LIMITS.max_username_length
        if len(seen) == 50:
            break
    assert len(set(seen)) == 50
    assert seen[1].endswith("-other")


def test_instants_are_utc_iso_and_clamped():
    assert instant_iso("2025-01-01T01:00:00+01:00") == "2025-01-01T00:00:00+00:00"
    assert instant_iso("2025-01-01").startswith("2025-01-01T00:00:00")
    assert parse_instant(1700000000).year == 2023
    now = datetime.now(timezone.utc)
    future = now + timedelta(days=30)
    assert clamp_instant(future, now) == now
    assert instant_iso("1970-01-01").startswith("2000-01-01")


@pytest.mark.parametrize("raw", ["", None, "nope", True, "2025-13-45"])
def test_instant_rejects_garbage(raw):
    with pytest.raises(ValueError):
        parse_instant(raw)


def test_fingerprint_ignores_key_order_and_detects_change():
    assert fingerprint({"a": 1, "b": 2}) == fingerprint({"b": 2, "a": 1})
    assert fingerprint({"a": 1}) != fingerprint({"a": 2})