SUMM.AI API

This package contains tests of the integreat_cms.summ_ai_api app

SUMM.AI Test

async tests.summ_ai_api.summ_ai_test.test_auto_translate_easy_german(login_role_user_async: tuple[AsyncClient, str], settings: SettingsWrapper, aiohttp_server: Callable, caplog: LogCaptureFixture) → None[source]

This test checks whether the SUMM.AI API client works as expected

Parameters:

  • login_role_user_async (tuple[AsyncClient, str]) – The fixture providing the http client and the current role (see login_role_user())

  • settings (SettingsWrapper) – The fixture providing the django settings

  • aiohttp_server (Callable) – The fixture providing the dummy aiohttp server used for faking the SUMM.AI API server

  • caplog (LogCaptureFixture) – The caplog fixture

Return type:

None

tests.summ_ai_api.summ_ai_test.test_check_rate_limit_exceeded(settings: SettingsWrapper, caplog: LogCaptureFixture) → None[source]

Test for check_rate_limit_exceeded method. Tests if return is True when response-status = 429 (too many requests) or 529 (site overloaded), otherwise False

Parameters:

  • settings (SettingsWrapper) – The Django settings

  • caplog (LogCaptureFixture) – Fixture for asserting log messages in tests (see caplog)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_missing_translation(settings: SettingsWrapper, aiohttp_raw_server: Callable) → None[source]

Test correct handling for a json response without the translated_text key by SUMM.AI. Tests if SummAiRuntimeError is raised when invalid JSON is sent as response.

Parameters:

  • settings (SettingsWrapper)

  • aiohttp_raw_server (Callable)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_patient_task_queue_hit_rate_limit() → None[source]

Test for PatientTaskQueue Class. It tests that when the maximum number of parallel requests to the summ_ai_api has been exceeded (rate limit exceeded), a defined time is waited until the next task object is returned via __anext__().

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_patient_task_queue_max_retries() → None[source]

Test for PatientTaskQueue Class. Tests that it stops early when requests are not successful after a cooldown, resulting in another cooldown repeatedly.

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_patient_task_queue_normal_deque() → None[source]

Test for PatientTaskQueue Class. Tests that a first-in/first out (fifo) deque-queue is built from a task-list Checks that __anext__() removes one element each from the queue

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_summ_ai_error_handling(login_role_user_async: tuple[AsyncClient, str], settings: SettingsWrapper, aiohttp_raw_server: Callable, caplog: LogCaptureFixture) → None[source]

This test checks whether the error handling of the SUMM.AI API client works as expected.

Parameters:

  • login_role_user_async (tuple[AsyncClient, str]) – The fixture providing the http client and the current role (see login_role_user())

  • settings (SettingsWrapper) – The fixture providing the django settings

  • aiohttp_raw_server (Callable) – The fixture providing the dummy aiohttp server used for faking the SUMM.AI API server

  • caplog (LogCaptureFixture) – The caplog fixture

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_translate_text_field_ddos_defense(settings: SettingsWrapper, aiohttp_raw_server: Callable, caplog: LogCaptureFixture) → None[source]

see test_translate_text_field_successful_translation()

Parameters:

  • settings (SettingsWrapper)

  • aiohttp_raw_server (Callable)

  • caplog (LogCaptureFixture)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_translate_text_field_hit_rate_limit(settings: SettingsWrapper, aiohttp_raw_server: Callable, caplog: LogCaptureFixture) → None[source]

see test_translate_text_field_successful_translation()

Parameters:

  • settings (SettingsWrapper)

  • aiohttp_raw_server (Callable)

  • caplog (LogCaptureFixture)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_translate_text_field_internal_server_error(settings: SettingsWrapper, aiohttp_raw_server: Callable, caplog: LogCaptureFixture) → None[source]

see test_translate_text_field_successful_translation()

Parameters:

  • settings (SettingsWrapper)

  • aiohttp_raw_server (Callable)

  • caplog (LogCaptureFixture)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_translate_text_field_successful_translation(settings: SettingsWrapper, aiohttp_raw_server: Callable, caplog: LogCaptureFixture) → None[source]
Parameters:

  • settings (SettingsWrapper) – The Django settings

  • aiohttp_raw_server (Callable) – The fixture providing the dummy aiohttp server used for faking the SUMM.AI API server

  • caplog (LogCaptureFixture) – Fixture for asserting log messages in tests (see caplog)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_translate_text_forbidden(settings: SettingsWrapper, aiohttp_raw_server: Callable, caplog: LogCaptureFixture) → None[source]

tests 403 response see test_translate_text_field_successful_translation()

Parameters:

  • settings (SettingsWrapper)

  • aiohttp_raw_server (Callable)

  • caplog (LogCaptureFixture)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_translate_text_with_empty_text_field(settings: SettingsWrapper, aiohttp_raw_server: Callable, caplog: LogCaptureFixture) → None[source]

see test_translate_text_field_successful_translation()

Parameters:

  • settings (SettingsWrapper)

  • aiohttp_raw_server (Callable)

  • caplog (LogCaptureFixture)

Return type:

None

async tests.summ_ai_api.summ_ai_test.test_unexpected_html(settings: SettingsWrapper, aiohttp_raw_server: Callable) → None[source]

Test correct handling for an unexpected HTML response by SUMM.AI. Tests if SummAiInvalidJSONError is raised when unexpected HTML is sent as response.

Parameters:

  • settings (SettingsWrapper)

  • aiohttp_raw_server (Callable)

Return type:

None

tests.summ_ai_api.summ_ai_test.test_validate_response_invalid() → None[source]

Test for validate_response method. Should throw exception when ‘translated_text’ is not found in request.

Return type:

None

tests.summ_ai_api.summ_ai_test.test_validate_response_valid() → None[source]

Test for validate_response method. Should return True when ‘translated_text’ is found in request.

Return type:

None

tests.summ_ai_api.summ_ai_test.test_worker() → None[source]

Test that the worker properly works through tasks from the PatientTaskQueue, returning the results.

Return type:

None

Utils

class tests.summ_ai_api.utils.MockedRequest[source]

Bases: object

Helper class mocking request, used for creating a SummAiApiclient instance. Region-property is needed therefore.

__init__() → None[source]
Return type:

None

tests.summ_ai_api.utils.enable_summ_api(region_slug: str) → None[source]

Enable SUMM.AI in the test region

Parameters:

region_slug (str) – The slug of the region in which we want to enable SUMM.AI

Return type:

None

tests.summ_ai_api.utils.get_changed_pages(settings: SettingsWrapper, ids: list[int]) → list[dict][source]

Load the changed pages with the specified ids from the database

Parameters:

  • settings (SettingsWrapper) – The fixture providing the django settings

  • ids (list[int]) – A list containing the requested pages including their translations in German and Easy German

Returns:

The changed pages

Return type:

list[dict]

tests.summ_ai_api.utils.make_mock_summ_ai_server(response_data: dict, status: int = 200) → Callable[source]

Creates a mock of SummAiServer-(Response)

Parameters:

  • response_data (dict) – data that mocked server returns

  • status (int) – http status code that mocked server returns (default = 200)

Returns:

Request handler for mocked SummAiServer

Return type:

Callable

tests.summ_ai_api.utils.make_mock_summ_ai_server_rate_limited() → Application[source]

Creates a mock of SUMM.AI server with rate limiting

Returns:

Mocked SUMM.AI server with rate limiting

Return type:

Application

tests.summ_ai_api.utils.make_rogue_summ_ai_server(response_text: str, status: int = 200) → Callable[source]

Creates a mock of SummAiServer-(Response) that can return any text as response (not just JSON)

Parameters:

  • response_text (str) – text that mocked server returns

  • status (int) – http status code that mocked server returns (default = 200)

Returns:

Request handler for mocked SummAiServer

Return type:

Callable