Workday Event Collector

import json
from datetime import datetime, timedelta

import pytest
from freezegun import freeze_time
from WorkdayEventCollector import DATE_FORMAT, DEFAULT_MAX_FETCH, MAX_PAGE_SIZE, Client


def util_load_json(path):
    with open(path, encoding="utf-8") as f:
        return json.loads(f.read())


class TestFetchActivity:
    @pytest.fixture(autouse=True)
    def create_client(self, mocker):
        self.base_url = "https://test.com"
        self.tenant_name = "test"
        self.token_url = f"{self.base_url}/ccx/oauth2/{self.tenant_name}/token"
        mocker.patch.object(Client, "get_access_token", return_value="1234")
        self.client = Client(
            base_url=self.base_url,
            refresh_token="refresh_token",
            client_id="test12",
            client_secret="test_sec1234",
            token_url=self.token_url,
            verify=False,
            proxy=False,
            headers={"Accept": "application/json", "Content-Type": "application/json"},
            max_fetch=25000,
        )

    @staticmethod
    def create_response_by_limit(from_date, to_date, offset, user_activity_entry_count=False, limit=1):
        single_response = util_load_json("test_data/single_loggings_response.json")
        return [single_response.copy() for i in range(limit)]

    @staticmethod
    def create_response_with_duplicates(request_time, limit, number_of_different_time, id_to_start_from):
        """
        Creates response with different requestTime and ids.
        Args:
            request_time: request time to start from.
            limit: number of responses
            number_of_different_time: number of responses with different requestTime
            id_to_start_from: id to start from

        """
        single_response = util_load_json("test_data/single_loggings_response.json")
        request_time_date_time = datetime.strptime(request_time, DATE_FORMAT)
        output = []

        def create_single(single_response, time, id, output):
            single_response["requestTime"] = time
            single_response["taskId"] = str(id)
            output.append(single_response)
            id += 1
            return output, id

        for _i in range(limit - number_of_different_time):
            output, id_to_start_from = create_single(single_response.copy(), request_time, id_to_start_from, output)
        for _i in range(limit - number_of_different_time, limit):
            new_time = datetime.strftime(request_time_date_time + timedelta(seconds=10), DATE_FORMAT)
            output, id_to_start_from = create_single(single_response.copy(), new_time, id_to_start_from, output)
        return output

    @pytest.mark.parametrize("loggings_to_fetch", [1, 4, 6], ids=["Single", "Part", "All"])
    def test_get_max_fetch_activity_logging(self, loggings_to_fetch, requests_mock, mocker):
        """
        Given: number of logging to fetch.
        When: running get activity logging command or fetch.
        Then: return the correct number of loggings.

        """
        from WorkdayEventCollector import get_max_fetch_activity_logging

        mocker.patch.object(Client, "get_activity_logging_request", side_effect=self.create_response_by_limit)
        res = get_max_fetch_activity_logging(
            client=self.client,
            logging_to_fetch=loggings_to_fetch,
            from_date="2023-04-15T07:00:00.000Z",
            to_date="2023-04-16T07:00:00.000Z",
        )
        assert len(res) == loggings_to_fetch

    def test_get_max_fetch_activity_logging_paginates_above_page_size(self, mocker):
        """
        Given: a max_fetch (logging_to_fetch) larger than the per-call page cap (3 * MAX_PAGE_SIZE).
        When: running get_max_fetch_activity_logging.
        Then: the client is called multiple times (pagination), each call is capped at MAX_PAGE_SIZE,
              the offset advances by the number of events returned, and all events are collected.
        """
        from WorkdayEventCollector import get_max_fetch_activity_logging

        logging_to_fetch = 3 * MAX_PAGE_SIZE

        def response_by_requested_limit(from_date, to_date, offset, limit):
            # Workday returns up to `limit` events (page cap enforced by caller at MAX_PAGE_SIZE).
            single_response = util_load_json("test_data/single_loggings_response.json")
            return [single_response.copy() for _ in range(limit)]

        request_mock = mocker.patch.object(Client, "get_activity_logging_request", side_effect=response_by_requested_limit)

        res = get_max_fetch_activity_logging(
            client=self.client,
            logging_to_fetch=logging_to_fetch,
            from_date="2023-04-15T07:00:00.000Z",
            to_date="2023-04-16T07:00:00.000Z",
        )

        # logging_to_fetch / MAX_PAGE_SIZE => exactly 3 paginated API calls.
        assert request_mock.call_count == 3
        # Every call must be capped at the per-call page limit.
        assert [call.kwargs["limit"] for call in request_mock.call_args_list] == [MAX_PAGE_SIZE, MAX_PAGE_SIZE, MAX_PAGE_SIZE]
        # Offset must advance by the number of events collected so far.
        assert [call.kwargs["offset"] for call in request_mock.call_args_list] == [0, MAX_PAGE_SIZE, 2 * MAX_PAGE_SIZE]
        # All events are accumulated across the pages.
        assert len(res) == logging_to_fetch

    def test_get_max_fetch_activity_logging_stops_on_empty_page(self, mocker):
        """
        Given: a large max_fetch but the source runs out of events mid-pagination.
        When: running get_max_fetch_activity_logging.
        Then: pagination stops as soon as an empty page is returned (no infinite loop),
              and only the available events are collected.
        """
        from WorkdayEventCollector import get_max_fetch_activity_logging

        single_response = util_load_json("test_data/single_loggings_response.json")
        first_page = [single_response.copy() for _ in range(MAX_PAGE_SIZE)]
        second_page = [single_response.copy() for _ in range(MAX_PAGE_SIZE // 2)]
        request_mock = mocker.patch.object(
            Client,
            "get_activity_logging_request",
            side_effect=[first_page, second_page, []],
        )

        res = get_max_fetch_activity_logging(
            client=self.client,
            logging_to_fetch=3 * MAX_PAGE_SIZE,
            from_date="2023-04-15T07:00:00.000Z",
            to_date="2023-04-16T07:00:00.000Z",
        )

        available_events = MAX_PAGE_SIZE + MAX_PAGE_SIZE // 2
        # Third call returns empty -> loop breaks. Only the available events are collected.
        assert request_mock.call_count == 3
        assert [call.kwargs["offset"] for call in request_mock.call_args_list] == [0, MAX_PAGE_SIZE, available_events]
        assert len(res) == available_events

    def test_resolve_max_fetch_uses_provided_value(self):
        """
        Given: a max_fetch value is provided in the params.
        When: running resolve_max_fetch.
        Then: the provided value is returned (as an int).
        """
        from WorkdayEventCollector import resolve_max_fetch

        assert resolve_max_fetch({"max_fetch": "500"}) == 500

    @pytest.mark.parametrize("params", [{}, {"max_fetch": None}, {"max_fetch": ""}], ids=["missing", "none", "empty"])
    def test_resolve_max_fetch_falls_back_to_default(self, params):
        """
        Given: max_fetch is missing, None, or empty in the params.
        When: running resolve_max_fetch.
        Then: DEFAULT_MAX_FETCH is returned.
        """
        from WorkdayEventCollector import resolve_max_fetch

        assert resolve_max_fetch(params) == DEFAULT_MAX_FETCH

    DUPLICATED_ACTIVITY_LOGGINGS = [
        (("2023-04-15T07:00:00Z", 5, 2, 0), 5, {}),
        (("2023-04-15T07:00:00Z", 5, 1, 0), 4, {"last_log": 0}),
    ]

    @pytest.mark.parametrize("args, len_of_activity_loggings, last_run", DUPLICATED_ACTIVITY_LOGGINGS)
    def test_remove_duplicated_activity_logging(self, args, len_of_activity_loggings, last_run):
        """
        Given: responses with potential duplications from last fetch.
        When: running fetch command.
        Then: return last responses with the latest requestTime to check if there are duplications.

        """
        from WorkdayEventCollector import remove_duplications

        loggings = self.create_response_with_duplicates(*args)
        if "last_log" in last_run:
            last_run["last_log"] = loggings[last_run.get("last_log")]
        activity_loggings = remove_duplications(loggings, last_run)
        assert len(activity_loggings) == len_of_activity_loggings

    def test_remove_milliseconds_from_time_of_logging(self):
        """
        Given: loggings with time string with milliseconds
        When: Fetching loggings from Workday
        Then: Remove the milliseconds.

        """
        from WorkdayEventCollector import remove_milliseconds_from_time_of_logging

        activity_logging: dict = util_load_json("test_data/single_loggings_response.json")
        requests_time = "2023-04-24T07:00:00.123Z"
        final_time = "2023-04-24T07:00:00Z"
        activity_logging["requestTime"] = requests_time

        assert remove_milliseconds_from_time_of_logging(activity_logging) == final_time

    def test_get_activity_logging_command(self, mocker):
        """
        Given: params to run get_activity_logging_command
        When: running the command
        Then: Accurate response and readable output is returned.
        """
        from WorkdayEventCollector import get_activity_logging_command

        mocker.patch.object(Client, "get_activity_logging_request", side_effect=self.create_response_by_limit)
        activity_loggings, res = get_activity_logging_command(
            client=self.client, from_date="2023-04-15T07:00:00Z", to_date="2023-04-16T07:00:00Z", limit=4, offset=0
        )
        assert len(activity_loggings) == 4
        assert "Activity Logging List" in res.readable_output

    @freeze_time("2023-04-15 08:00:00")
    def test_fetch_activity_logging(self, mocker):
        """
        Tests the fetch_events function

        Given:
            - first_fetch_time
        When:
            - Running the 'fetch_activity_logging' function.
        Then:
            - Validates that the function generates the correct API requests with the expected parameters.
            - Validates that the function returns the expected events and next_run timestamps.
        """
        from WorkdayEventCollector import fetch_activity_logging

        first_fetch_time = datetime.strptime("2023-04-12T07:00:00Z", DATE_FORMAT)
        fetched_events = util_load_json("test_data/fetch_activity_loggings.json")
        http_responses = mocker.patch.object(
            Client,
            "get_activity_logging_request",
            side_effect=[
                fetched_events.get("fetch_loggings_before"),
                fetched_events.get("fetch_loggings"),
            ],
        )

        activity_loggings, new_last_run = fetch_activity_logging(
            self.client, last_run={}, first_fetch=first_fetch_time, max_fetch=3
        )

        assert http_responses.call_args_list[0][1] == {
            "limit": 3,
            "offset": 0,
            "from_date": "2023-04-12T07:00:00Z",
            "to_date": "2023-04-15T08:00:00Z",
        }
        assert http_responses.call_args_list[1][1] == {
            "limit": 2,
            "offset": 1,
            "from_date": "2023-04-12T07:00:00Z",
            "to_date": "2023-04-15T08:00:00Z",
        }

        assert activity_loggings == fetched_events.get("fetched_events")
        assert new_last_run.get("last_fetch_time") == "2023-04-15T07:00:00Z"
        assert new_last_run.get("last_log").get("taskId") == "3"

        # assert no new results when given the last_run:
        fetched_events = util_load_json("test_data/fetch_activity_loggings.json")
        http_responses = mocker.patch.object(
            Client, "get_activity_logging_request", side_effect=[fetched_events.get("fetch_loggings"), []]
        )

        activity_loggings, new_last_run = fetch_activity_logging(
            self.client, last_run=new_last_run, first_fetch=first_fetch_time, max_fetch=3
        )
        assert http_responses.call_args_list[0][1] == {
            "limit": 3,
            "offset": 0,
            "from_date": "2023-04-15T07:00:00Z",
            "to_date": "2023-04-15T08:00:00Z",
        }
        assert activity_loggings == []
        assert new_last_run.get("last_fetch_time") == "2023-04-15T07:00:00Z"
        assert new_last_run.get("last_log").get("taskId") == "3"

    @pytest.mark.parametrize("max_fetch, instance_returned", [(6000, 1), (15000, 2), (60000, 6)])
    def test_instance_returned_request(self, mocker, max_fetch, instance_returned):
        self.client.max_fetch = max_fetch
        http_request = mocker.patch.object(Client, "http_request")
        self.client.get_activity_logging_request(from_date="2023-04-15T07:00:00Z", to_date="2023-04-15T08:00:00Z")
        params_sent = http_request.call_args_list[0][1].get("params", {})
        assert params_sent.get("instancesReturned") == instance_returned