"""메타데이터 추출 F1 평가 단위테스트 (성능지표 No.3)."""
from __future__ import annotations

import pytest

from app.engine.metadata_eval import set_prf, evaluate_extraction


def test_set_prf_perfect():
    r = set_prf(["홍길동", "활빈당"], ["홍길동", "활빈당"])
    assert r.f1 == 1.0 and r.fp == 0 and r.fn == 0


def test_set_prf_partial():
    r = set_prf(["홍길동", "임꺽정"], ["홍길동", "활빈당"])
    assert r.tp == 1 and r.fp == 1 and r.fn == 1
    assert 0.0 < r.f1 < 1.0


def test_set_prf_case_insensitive():
    r = set_prf(["Hong"], ["hong"])
    assert r.f1 == 1.0


def test_evaluate_extraction_micro():
    preds = [{"characters": ["홍길동"], "motifs": ["복수"], "keywords": ["활빈당"], "genre": "역사"}]
    golds = [{"characters": ["홍길동"], "motifs": ["복수"], "keywords": ["활빈당"], "genre": "역사"}]
    out = evaluate_extraction(preds, golds)
    assert out["micro_avg"]["f1"] == 1.0
    assert out["genre_accuracy"]["accuracy"] == 1.0


def test_evaluate_extraction_genre_mismatch():
    preds = [{"characters": [], "motifs": [], "keywords": [], "genre": "SF"}]
    golds = [{"characters": [], "motifs": [], "keywords": [], "genre": "역사"}]
    out = evaluate_extraction(preds, golds)
    assert out["genre_accuracy"]["accuracy"] == 0.0


def test_length_mismatch_raises():
    with pytest.raises(ValueError):
        evaluate_extraction([{}], [{}, {}])