test_api_endpoints.py

"""Tests for the FastAPI endpoints. Databricks is stubbed out with mock data."""
from __future__ import annotations

import sys
import types
from contextlib import asynccontextmanager
from datetime import datetime, timezone
from pathlib import Path
from unittest.mock import MagicMock

import pandas as pd
import pytest
from fastapi.testclient import TestClient

_fake_db = types.ModuleType("db")
_fake_db.get_spark = MagicMock()
sys.modules.setdefault("db", _fake_db)


MOCK_SCORES = pd.DataFrame(
    [
        {
            "package_name": "requests",
            "github_score": 85.0,
            "pypi_score": 90.0,
            "community_score": 80.0,
            "sentiment_score": 70.0,
            "overall_health_score": 82.25,
            "health_tier": "A",
            "scored_at": pd.Timestamp("2026-03-26 12:00:00"),
        },
        {
            "package_name": "flask",
            "github_score": 60.0,
            "pypi_score": 65.0,
            "community_score": 55.0,
            "sentiment_score": 50.0,
            "overall_health_score": 58.25,
            "health_tier": "B",
            "scored_at": pd.Timestamp("2026-03-26 12:00:00"),
        },
    ]
)

MOCK_SENTIMENT = pd.DataFrame(
    [
        {
            "package_name": "requests",
            "so_question_sentiment_avg": 0.4,
            "so_answer_sentiment_avg": 0.5,
            "readme_sentiment_compound": 0.3,
            "pypi_desc_sentiment_compound": 0.2,
            "overall_sentiment": 0.38,
        },
        {
            "package_name": "flask",
            "so_question_sentiment_avg": 0.2,
            "so_answer_sentiment_avg": 0.3,
            "readme_sentiment_compound": 0.1,
            "pypi_desc_sentiment_compound": 0.15,
            "overall_sentiment": 0.19,
        },
    ]
)


@pytest.fixture(scope="module")
def client():
    import main as m

    m._scores_df = MOCK_SCORES
    m._sentiment_df = MOCK_SENTIMENT

    @asynccontextmanager
    async def _noop_lifespan(app):
        yield

    m.app.router.lifespan_context = _noop_lifespan

    with TestClient(m.app) as c:
        yield c


class TestHealth:
    def test_returns_200(self, client):
        r = client.get("/health")
        assert r.status_code == 200

    def test_status_is_ok(self, client):
        r = client.get("/health")
        assert r.json()["status"] == "ok"

    def test_packages_cached_count(self, client):
        r = client.get("/health")
        assert r.json()["packages_cached"] == 2

    def test_timestamp_present(self, client):
        r = client.get("/health")
        assert "timestamp" in r.json()


class TestListPackages:
    def test_returns_all_packages(self, client):
        r = client.get("/packages")
        assert r.status_code == 200
        assert len(r.json()) == 2

    def test_sorted_descending_by_score(self, client):
        r = client.get("/packages")
        scores = [p["overall_health_score"] for p in r.json()]
        assert scores == sorted(scores, reverse=True)

    def test_each_item_has_required_fields(self, client):
        r = client.get("/packages")
        for item in r.json():
            assert "package_name" in item
            assert "overall_health_score" in item
            assert "health_tier" in item


class TestGetPackage:
    def test_known_package_returns_200(self, client):
        r = client.get("/packages/requests")
        assert r.status_code == 200

    def test_response_has_scores_and_sentiment(self, client):
        r = client.get("/packages/requests")
        body = r.json()
        assert "scores" in body
        assert "sentiment" in body

    def test_scores_sub_schema(self, client):
        r = client.get("/packages/requests")
        scores = r.json()["scores"]
        for field in ("github_score", "pypi_score", "community_score",
                      "sentiment_score", "overall_health_score", "health_tier"):
            assert field in scores

    def test_correct_tier_returned(self, client):
        r = client.get("/packages/requests")
        assert r.json()["scores"]["health_tier"] == "A"

    def test_unknown_package_returns_404(self, client):
        r = client.get("/packages/nonexistent_xyz_pkg")
        assert r.status_code == 404

    def test_404_detail_message(self, client):
        r = client.get("/packages/nonexistent_xyz_pkg")
        assert "not found" in r.json()["detail"].lower()


class TestGetScores:
    def test_returns_200(self, client):
        r = client.get("/packages/flask/scores")
        assert r.status_code == 200

    def test_all_score_fields_are_floats(self, client):
        r = client.get("/packages/flask/scores")
        body = r.json()
        for field in ("github_score", "pypi_score", "community_score",
                      "sentiment_score", "overall_health_score"):
            assert isinstance(body[field], float)

    def test_unknown_returns_404(self, client):
        r = client.get("/packages/unknown_abc/scores")
        assert r.status_code == 404


class TestGetSentiment:
    def test_returns_200(self, client):
        r = client.get("/packages/requests/sentiment")
        assert r.status_code == 200

    def test_overall_sentiment_present(self, client):
        r = client.get("/packages/requests/sentiment")
        assert "overall_sentiment" in r.json()

    def test_package_name_in_response(self, client):
        r = client.get("/packages/flask/sentiment")
        assert r.json()["package_name"] == "flask"


class TestLeaderboard:
    def test_returns_all_packages(self, client):
        r = client.get("/scores/leaderboard")
        assert r.status_code == 200
        assert len(r.json()) == 2

    def test_sorted_descending(self, client):
        r = client.get("/scores/leaderboard")
        scores = [p["overall_health_score"] for p in r.json()]
        assert scores == sorted(scores, reverse=True)

    def test_items_have_full_score_schema(self, client):
        r = client.get("/scores/leaderboard")
        for item in r.json():
            assert "health_tier" in item
            assert "scored_at" in item