test_poll_adapters.py

"""ATS adapters: Greenhouse and Lever (roadmap M4, ADR-0025).

Fixtures (`tests/fixtures/{greenhouse,lever}_sample.json`) are real recorded responses from
each platform's own public demo/reference board — GitLab's Greenhouse board and Lever's own
`leverdemo` board — truncated for size, not fabricated. Only `poll.http.get_json` is mocked;
everything downstream of it (parsing, the adapter's own logic) runs for real against that data.
"""
from __future__ import annotations

import json
from pathlib import Path
from unittest.mock import patch

import pytest

from openjobradar.poll import Posting, UnknownAtsTypeError, adapter_for
from openjobradar.poll.adapters import greenhouse, lever
from openjobradar.poll.adapters.base import InvalidBoardIdError, validate_board_id
from openjobradar.poll.http import UnsupportedUrlError, get_json, strip_html

FIXTURES = Path(__file__).parent / "fixtures"


def _load(name: str):
    return json.loads((FIXTURES / name).read_text(encoding="utf-8"))


# --- registry -----------------------------------------------------------------


def test_registry_resolves_both_adapters() -> None:
    assert adapter_for("greenhouse") is greenhouse
    assert adapter_for("lever") is lever


def test_registry_rejects_unknown_ats_type() -> None:
    with pytest.raises(UnknownAtsTypeError):
        adapter_for("workday")


# --- board id validation --------------------------------------------------------


@pytest.mark.parametrize("bad", ["", "has space", "has/slash", "a" * 101, None])
def test_invalid_board_ids_rejected(bad) -> None:
    with pytest.raises(InvalidBoardIdError):
        validate_board_id(bad)


def test_valid_board_id_passes_through() -> None:
    assert validate_board_id("gitlab") == "gitlab"


# --- greenhouse -----------------------------------------------------------------


def test_greenhouse_parses_real_fixture_into_postings() -> None:
    with patch("openjobradar.poll.adapters.greenhouse.get_json", return_value=_load("greenhouse_sample.json")):
        postings = greenhouse.list_postings("gitlab", "gitlab", "GitLab")

    assert len(postings) == 2
    first = postings[0]
    assert isinstance(first, Posting)
    assert first.org_slug == "gitlab"
    assert first.ats_type == "greenhouse"
    assert first.job_id
    assert first.title
    assert "Remote" in first.location_raw
    assert first.apply_url.startswith("https://")
    assert "<" not in first.jd_text  # HTML was stripped
    assert first.stable_id == f"greenhouse:gitlab:{first.job_id}"


def test_greenhouse_falls_back_to_offices_when_location_name_missing() -> None:
    data = {
        "jobs": [
            {
                "id": 1,
                "title": "Remote Engineer",
                "location": {"name": ""},
                "offices": [{"name": "Remote - US"}],
                "absolute_url": "https://boards.greenhouse.io/acme/jobs/1",
                "content": "<p>Build things.</p>",
                "departments": [{"name": "Engineering"}],
            }
        ]
    }
    with patch("openjobradar.poll.adapters.greenhouse.get_json", return_value=data):
        postings = greenhouse.list_postings("acme", "acme", "Acme")

    assert postings[0].location_raw == "Remote - US"
    assert postings[0].department == "Engineering"
    assert postings[0].jd_text == "Build things."


def test_greenhouse_handles_empty_board() -> None:
    with patch("openjobradar.poll.adapters.greenhouse.get_json", return_value={"jobs": []}):
        assert greenhouse.list_postings("acme", "acme", "Acme") == []


# --- lever ------------------------------------------------------------------------


def test_lever_parses_real_fixture_into_postings() -> None:
    with patch("openjobradar.poll.adapters.lever.get_json", return_value=_load("lever_sample.json")):
        postings = lever.list_postings("leverdemo", "leverdemo", "Lever")

    assert len(postings) == 2
    first = postings[0]
    assert first.org_slug == "leverdemo"
    assert first.ats_type == "lever"
    assert first.job_id
    assert first.title
    assert first.apply_url.startswith("https://")
    assert first.stable_id == f"lever:leverdemo:{first.job_id}"


def test_lever_location_prefers_all_locations_over_primary_only() -> None:
    data = [
        {
            "id": "1",
            "text": "Engineer",
            "categories": {"location": "San Francisco", "allLocations": ["San Francisco", "Remote"]},
            "hostedUrl": "https://jobs.lever.co/acme/1",
            "descriptionPlain": "Do the work.",
        }
    ]
    with patch("openjobradar.poll.adapters.lever.get_json", return_value=data):
        postings = lever.list_postings("acme", "acme", "Acme")
    assert postings[0].location_raw == "San Francisco / Remote"


def test_lever_falls_back_to_remote_when_no_locations_but_workplace_type_remote() -> None:
    data = [
        {
            "id": "1",
            "text": "Engineer",
            "categories": {},
            "workplaceType": "remote",
            "hostedUrl": "https://jobs.lever.co/acme/1",
        }
    ]
    with patch("openjobradar.poll.adapters.lever.get_json", return_value=data):
        postings = lever.list_postings("acme", "acme", "Acme")
    assert postings[0].location_raw == "Remote"


def test_lever_jd_text_assembles_lists_when_description_plain_absent() -> None:
    data = [
        {
            "id": "1",
            "text": "Engineer",
            "categories": {},
            "hostedUrl": "https://jobs.lever.co/acme/1",
            "description": "<p>Intro</p>",
            "lists": [{"text": "Requirements", "content": "<ul><li>Python</li></ul>"}],
            "additionalPlain": "Equal opportunity employer.",
        }
    ]
    with patch("openjobradar.poll.adapters.lever.get_json", return_value=data):
        postings = lever.list_postings("acme", "acme", "Acme")
    jd = postings[0].jd_text
    assert "Intro" in jd
    assert "Requirements" in jd and "Python" in jd
    assert "Equal opportunity employer." in jd


def test_lever_handles_empty_board() -> None:
    with patch("openjobradar.poll.adapters.lever.get_json", return_value=[]):
        assert lever.list_postings("acme", "acme", "Acme") == []


# --- http helpers ------------------------------------------------------------------


def test_strip_html_removes_tags_and_collapses_whitespace() -> None:
    assert strip_html("<p>Hello   <b>world</b></p>\n<p>!</p>") == "Hello world !"


def test_strip_html_of_empty_string_is_empty() -> None:
    assert strip_html("") == ""


def test_get_json_rejects_non_https_urls() -> None:
    with pytest.raises(UnsupportedUrlError):
        get_json("http://example.com/insecure")
    with pytest.raises(UnsupportedUrlError):
        get_json("file:///etc/passwd")