CIAgent/apps/api/tests/unit/test_sources_api.py

"""Sources API: ownership isolation, user-creatable type restriction, and
the ad-hoc test action - via HTTP, with respx mocking the network call the
test action makes."""

from __future__ import annotations

import uuid
from unittest.mock import patch

import httpx
import respx


def _register_and_login(client) -> dict[str, str]:
    email = f"user-{uuid.uuid4().hex[:12]}@example.com"
    client.post(
        "/api/v1/auth/register",
        json={"email": email, "password": "correct-horse-1", "display_name": "Test User"},
    )
    tokens = client.post(
        "/api/v1/auth/login", json={"email": email, "password": "correct-horse-1"}
    ).json()
    return {"Authorization": f"Bearer {tokens['access_token']}"}


def _create_company(client, headers):
    return client.post(
        "/api/v1/companies",
        json={"name": f"Co {uuid.uuid4().hex[:6]}", "frequency_type": "weekly"},
        headers=headers,
    ).json()


def test_create_custom_url_source(client):
    headers = _register_and_login(client)
    company = _create_company(client, headers)

    resp = client.post(
        f"/api/v1/companies/{company['id']}/sources",
        json={"source_type": "custom_url", "name": "Pricing", "base_url": "example.com/pricing"},
        headers=headers,
    )
    assert resp.status_code == 201
    body = resp.json()
    assert body["base_url"] == "https://example.com/pricing"
    assert body["status"] == "active"


def test_create_source_rejects_non_user_creatable_type(client):
    headers = _register_and_login(client)
    company = _create_company(client, headers)

    resp = client.post(
        f"/api/v1/companies/{company['id']}/sources",
        json={"source_type": "github", "name": "GitHub", "base_url": "https://github.com/acme"},
        headers=headers,
    )
    assert resp.status_code == 422


def test_sources_scoped_to_owner(client):
    owner_headers = _register_and_login(client)
    other_headers = _register_and_login(client)
    company = _create_company(client, owner_headers)

    created = client.post(
        f"/api/v1/companies/{company['id']}/sources",
        json={
            "source_type": "custom_url",
            "name": "Pricing",
            "base_url": "https://example.com/pricing",
        },
        headers=owner_headers,
    ).json()

    # Another user can't list this company's sources...
    resp = client.get(f"/api/v1/companies/{company['id']}/sources", headers=other_headers)
    assert resp.status_code == 404

    # ...or update/delete the source directly.
    resp = client.patch(
        f"/api/v1/sources/{created['id']}", json={"active": False}, headers=other_headers
    )
    assert resp.status_code == 404


def test_update_and_delete_source(client):
    headers = _register_and_login(client)
    company = _create_company(client, headers)
    created = client.post(
        f"/api/v1/companies/{company['id']}/sources",
        json={
            "source_type": "custom_url",
            "name": "Pricing",
            "base_url": "https://example.com/pricing",
        },
        headers=headers,
    ).json()

    resp = client.patch(f"/api/v1/sources/{created['id']}", json={"active": False}, headers=headers)
    assert resp.status_code == 200
    assert resp.json()["active"] is False

    resp = client.delete(f"/api/v1/sources/{created['id']}", headers=headers)
    assert resp.status_code == 204

    resp = client.get(f"/api/v1/companies/{company['id']}/sources", headers=headers)
    assert resp.json() == []


def test_update_source_sets_a_frequency_override(client):
    headers = _register_and_login(client)
    company = _create_company(client, headers)
    created = client.post(
        f"/api/v1/companies/{company['id']}/sources",
        json={
            "source_type": "custom_url",
            "name": "Pricing",
            "base_url": "https://example.com/pricing",
        },
        headers=headers,
    ).json()

    resp = client.patch(
        f"/api/v1/sources/{created['id']}", json={"frequency_type": "daily"}, headers=headers
    )
    assert resp.status_code == 200
    body = resp.json()
    assert body["frequency_type"] == "daily"
    assert body["next_check"] is None  # takes effect on the next scheduler tick

    # Clearing the override back to "same as company" is an explicit null.
    resp = client.patch(
        f"/api/v1/sources/{created['id']}", json={"frequency_type": None}, headers=headers
    )
    assert resp.status_code == 200
    assert resp.json()["frequency_type"] is None


def test_update_source_rejects_a_custom_frequency_below_the_minimum_interval(client):
    headers = _register_and_login(client)
    company = _create_company(client, headers)
    created = client.post(
        f"/api/v1/companies/{company['id']}/sources",
        json={
            "source_type": "custom_url",
            "name": "Pricing",
            "base_url": "https://example.com/pricing",
        },
        headers=headers,
    ).json()

    resp = client.patch(
        f"/api/v1/sources/{created['id']}",
        json={"frequency_type": "custom", "interval_minutes": 1},
        headers=headers,
    )
    assert resp.status_code == 400


def test_source_test_action_runs_a_real_collection(client):
    headers = _register_and_login(client)
    company = _create_company(client, headers)
    created = client.post(
        f"/api/v1/companies/{company['id']}/sources",
        json={
            "source_type": "custom_url",
            "name": "Pricing",
            "base_url": "https://example.com/pricing",
        },
        headers=headers,
    ).json()

    with patch("socket.getaddrinfo", return_value=[(2, 1, 6, "", ("93.184.216.34", 0))]):
        with respx.mock:
            respx.get("https://example.com/robots.txt").mock(return_value=httpx.Response(404))
            respx.get("https://example.com/pricing").mock(
                return_value=httpx.Response(
                    200,
                    html="<html><head><title>Pricing</title></head><body>"
                    "<article><h1>Pricing</h1><p>Plans start at $10/month.</p></article>"
                    "</body></html>",
                )
            )
            resp = client.post(f"/api/v1/sources/{created['id']}/test", headers=headers)

    assert resp.status_code == 200
    body = resp.json()
    assert body["status"] == "active"
    assert body["documents_found"] == 1