backend/tests/test_api_contract.py

"""End-to-end API contract tests.

Spec section 6.1: happy path, postcode regex (422), description length (422).
"""

from __future__ import annotations

from pathlib import Path

import pytest
from fastapi.testclient import TestClient

from app.config import PDF_DIR
from app.ingestion import SOURCE_REGISTRY
from app.main import app
from app.vector_store import get_vector_store


@pytest.fixture(scope="module", autouse=True)
def _bootstrap_corpus(tmp_path_factory) -> None:
    """Copy the on-disk PDFs into a temp dir and point the vector store there."""

    base = tmp_path_factory.mktemp("belfastbuild-api")
    target = base / "pdfs"
    target.mkdir(parents=True, exist_ok=True)
    for entry in SOURCE_REGISTRY:
        src = PDF_DIR / entry["filename"]
        if src.exists():
            (target / entry["filename"]).write_bytes(src.read_bytes())

    import shutil

    chroma_dir = base / "chroma"
    chroma_dir.mkdir(parents=True, exist_ok=True)

    from app.config import CHROMA_DIR
    from app.ingestion import PDF_DIR as INGEST_PDF_DIR
    from app import vector_store

    # Override module-level paths so the live vector store points at temp dirs.
    vector_store.PDF_DIR = INGEST_PDF_DIR
    vector_store.CHROMA_DIR = chroma_dir  # type: ignore[misc]

    # Reset the cached store and trigger a real re-index.
    vector_store.get_vector_store.cache_clear()
    store = vector_store.get_vector_store()
    store.index_corpus(force_reindex=True)


client = TestClient(app)


def test_health_endpoint() -> None:
    response = client.get("/api/health")
    assert response.status_code == 200
    body = response.json()
    assert body["status"] == "ok"
    assert body["indexed"] is True
    assert body["ready_for_screening"] is True
    assert "spatial_lookup_enabled" in body
    assert "corpus_fingerprint" in body


def test_happy_path_returns_200_with_integer_score() -> None:
    payload = {
        "postcode": "BT9 7AG",
        "description": "Single-storey rear extension to a semi-detached dwelling with off-street parking.",
    }
    response = client.post("/api/screen", json=payload)
    assert response.status_code == 200, response.text
    body = response.json()
    assert isinstance(body["compliance_percentage"], int)
    assert 0 <= body["compliance_percentage"] <= 100
    assert isinstance(body["policy_conflicts"], list)
    assert isinstance(body["source_citations"], list)
    assert "remediation_markdown" in body and body["remediation_markdown"]


def test_non_ni_postcode_rejected_with_422() -> None:
    response = client.post(
        "/api/screen",
        json={"postcode": "SW1A 1AA", "description": "Single-storey rear extension with off-street parking."},
    )
    assert response.status_code == 422


def test_short_description_rejected_with_422() -> None:
    response = client.post(
        "/api/screen",
        json={"postcode": "BT9 7AG", "description": "tiny"},
    )
    assert response.status_code == 422


def test_malformed_postcode_rejected_with_422() -> None:
    response = client.post(
        "/api/screen",
        json={"postcode": "BLAH", "description": "Single-storey rear extension to a residential property."},
    )
    assert response.status_code == 422