"""Tests for the scoreboard schema + CSV snapshot (no live Mongo needed)."""

import csv
from pathlib import Path

from harness import scoreboard
from harness.scoreboard import ScoreRow

REPO_ROOT = Path(__file__).resolve().parent.parent


def test_score_row_field_order():
    names = ScoreRow.field_names()
    assert names[:9] == [
        "video_id", "upload_date", "variant_id", "genome_hash", "parent_genome",
        "APV", "VSA", "fitness", "channel_status",
    ]
    # context fields exist and come after the objective
    for ctx in ["views", "likes", "comments", "shares", "avg_view_duration_sec", "subscribers_gained"]:
        assert ctx in names


def test_committed_metrics_csv_header_matches_schema():
    with (REPO_ROOT / "metrics.csv").open(newline="", encoding="utf-8") as fh:
        header = next(csv.reader(fh))
    assert header == ScoreRow.field_names()


def test_from_doc_ignores_unknown_keys():
    row = ScoreRow.from_doc({
        "video_id": "v", "upload_date": "2026-06-10", "variant_id": "A",
        "genome_hash": "h", "parent_genome": "seed", "APV": 50.0, "VSA": 0.5,
        "fitness": 5.0, "_id": "should-be-ignored", "junk": 123,
    })
    assert row.video_id == "v" and row.fitness == 5.0


def test_score_row_holds_no_secret_fields():
    # Defense-in-depth: the schema must not contain anything that could carry a credential.
    forbidden = {"token", "secret", "password", "api_key", "mongo_url", "credentials", "oauth"}
    for name in ScoreRow.field_names():
        assert not any(bad in name.lower() for bad in forbidden)


def test_snapshot_to_csv(monkeypatch, tmp_path):
    rows = [
        ScoreRow("b", "2026-06-11", "A", "h", "seed", 50, 0.5, 5.0),
        ScoreRow("a", "2026-06-10", "A", "h", "seed", 80, 0.5, 6.8, views=10, likes=2),
    ]
    monkeypatch.setattr(scoreboard, "read_all", lambda readonly=True: rows)
    out = tmp_path / "snap.csv"
    n = scoreboard.snapshot_to_csv(out, readonly=True)
    assert n == 2
    with out.open(newline="", encoding="utf-8") as fh:
        reader = list(csv.DictReader(fh))
    # sorted by (upload_date, variant_id) → 'a' (06-10) before 'b' (06-11)
    assert [r["video_id"] for r in reader] == ["a", "b"]
    assert reader[0]["likes"] == "2"