zulip/zerver/lib/test_classes.py

2431 lines
88 KiB
Python
Raw Normal View History

import base64
import os
import re
import shutil
import subprocess
import tempfile
from contextlib import contextmanager
from datetime import timedelta
from typing import (
TYPE_CHECKING,
Any,
Callable,
Collection,
Dict,
Iterable,
Iterator,
List,
Mapping,
Optional,
Sequence,
Set,
Tuple,
Union,
cast,
)
from unittest import TestResult, mock, skipUnless
from urllib.parse import parse_qs, quote, urlencode
import lxml.html
import orjson
import responses
from django.apps import apps
from django.conf import settings
from django.core.mail import EmailMessage
from django.core.signals import got_request_exception
from django.db import connection
from django.db.migrations.executor import MigrationExecutor
from django.db.migrations.state import StateApps
from django.db.utils import IntegrityError
from django.http import HttpRequest, HttpResponse, HttpResponseBase
from django.http.response import ResponseHeaders
from django.test import Client as TestClient
from django.test import SimpleTestCase, TestCase, TransactionTestCase
from django.test.client import BOUNDARY, MULTIPART_CONTENT, ClientHandler, encode_multipart
from django.test.testcases import SerializeMixin
from django.urls import resolve
from django.utils import translation
from django.utils.module_loading import import_string
from django.utils.timezone import now as timezone_now
from django_stubs_ext import ValuesQuerySet
from fakeldap import MockLDAP
from openapi_core.contrib.django import DjangoOpenAPIRequest, DjangoOpenAPIResponse
from requests import PreparedRequest
from two_factor.plugins.phonenumber.models import PhoneDevice
from typing_extensions import override
from corporate.models import Customer, CustomerPlan, LicenseLedger
from zerver.actions.message_send import check_send_message, check_send_stream_message
from zerver.actions.realm_settings import (
do_change_realm_permission_group_setting,
do_set_realm_property,
)
from zerver.actions.streams import bulk_add_subscriptions, bulk_remove_subscriptions
from zerver.decorator import do_two_factor_login
from zerver.lib.cache import bounce_key_prefix_for_testing
from zerver.lib.initial_password import initial_password
from zerver.lib.message import access_message
from zerver.lib.notification_data import UserMessageNotificationsData
per-request caches: Add per_request_cache library. We have historically cached two types of values on a per-request basis inside of memory: * linkifiers * display recipients Both of these caches were hand-written, and they both actually cache values that are also in memcached, so the per-request cache essentially only saves us from a few memcached hits. I think the linkifier per-request cache is a necessary evil. It's an important part of message rendering, and it's not super easy to structure the code to just get a single value up front and pass it down the stack. I'm not so sure we even need the display recipient per-request cache any more, as we are generally pretty smart now about hydrating recipient data in terms of how the code is organized. But I haven't done thorough research on that hypotheseis. Fortunately, it's not rocket science to just write a glorified memoize decorator and tie it into key places in the code: * middleware * tests (e.g. asserting db counts) * queue processors That's what I did in this commit. This commit definitely reduces the amount of code to maintain. I think it also gets us closer to possibly phasing out this whole technique, but that effort is beyond the scope of this PR. We could add some instrumentation to the decorator to see how often we get a non-trivial number of saved round trips to memcached. Note that when we flush linkifiers, we just use a big hammer and flush the entire per-request cache for linkifiers, since there is only ever one realm in the cache.
2023-07-14 19:46:50 +02:00
from zerver.lib.per_request_cache import flush_per_request_caches
from zerver.lib.redis_utils import bounce_redis_key_prefix_for_testing
from zerver.lib.sessions import get_session_dict_user
from zerver.lib.soft_deactivation import do_soft_deactivate_users
from zerver.lib.stream_subscription import get_subscribed_stream_ids_for_user
from zerver.lib.streams import (
create_stream_if_needed,
get_default_value_for_history_public_to_subscribers,
)
from zerver.lib.subscription_info import gather_subscriptions
from zerver.lib.test_console_output import (
ExtraConsoleOutputFinder,
ExtraConsoleOutputInTestError,
tee_stderr_and_find_extra_console_output,
tee_stdout_and_find_extra_console_output,
)
from zerver.lib.test_helpers import (
cache_tries_captured,
find_key_by_email,
instrument_url,
queries_captured,
)
from zerver.lib.topic import RESOLVED_TOPIC_PREFIX, filter_by_topic_name_via_message
from zerver.lib.user_groups import get_system_user_group_for_user
from zerver.lib.users import get_api_key
from zerver.lib.webhooks.common import (
check_send_webhook_message,
get_fixture_http_headers,
standardize_headers,
)
from zerver.models import (
Client,
Message,
NamedUserGroup,
PushDeviceToken,
2021-12-05 11:43:34 +01:00
Reaction,
Realm,
2021-12-05 11:43:34 +01:00
RealmEmoji,
Recipient,
Stream,
Subscription,
UserGroupMembership,
tests: Ensure stream senders get a UserMessage row. We now complain if a test author sends a stream message that does not result in the sender getting a UserMessage row for the message. This is basically 100% equivalent to complaining that the author failed to subscribe the sender to the stream as part of the test setup, as far as I can tell, so the AssertionError instructs the author to subscribe the sender to the stream. We exempt bots from this check, although it is plausible we should only exempt the system bots like the notification bot. I considered auto-subscribing the sender to the stream, but that can be a little more expensive than the current check, and we generally want test setup to be explicit. If there is some legitimate way than a subscribed human sender can't get a UserMessage, then we probably want an explicit test for that, or we may want to change the backend to just write a UserMessage row in that hypothetical situation. For most tests, including almost all the ones fixed here, the author just wants their test setup to realistically reflect normal operation, and often devs may not realize that Cordelia is not subscribed to Denmark or not realize that Hamlet is not subscribed to Scotland. Some of us don't remember our Shakespeare from high school, and our stream subscriptions don't even necessarily reflect which countries the Bard placed his characters in. There may also be some legitimate use case where an author wants to simulate sending a message to an unsubscribed stream, but for those edge cases, they can always set allow_unsubscribed_sender to True.
2021-12-10 13:55:48 +01:00
UserMessage,
UserProfile,
UserStatus,
)
from zerver.models.groups import SystemGroups
from zerver.models.realms import clear_supported_auth_backends_cache, get_realm
from zerver.models.streams import get_realm_stream, get_stream
from zerver.models.users import get_system_bot, get_user, get_user_by_delivery_email
from zerver.openapi.openapi import validate_test_request, validate_test_response
from zerver.tornado.event_queue import clear_client_event_queues_for_testing
if settings.ZILENCER_ENABLED:
from zilencer.models import RemoteZulipServer, get_remote_server_by_uuid
if TYPE_CHECKING:
from django.test.client import _MonkeyPatchedWSGIResponse as TestHttpResponse
class EmptyResponseError(Exception):
pass
class UploadSerializeMixin(SerializeMixin):
"""
We cannot use override_settings to change upload directory because
because settings.LOCAL_UPLOADS_DIR is used in URL pattern and URLs
are compiled only once. Otherwise using a different upload directory
for conflicting test cases would have provided better performance
while providing the required isolation.
"""
lockfile = "var/upload_lock"
@classmethod
@override
def setUpClass(cls: Any) -> None:
if not os.path.exists(cls.lockfile):
with open(cls.lockfile, "w"): # nocoverage - rare locking case
pass
super().setUpClass()
class ZulipClientHandler(ClientHandler):
@override
def get_response(self, request: HttpRequest) -> HttpResponseBase:
got_exception = False
def on_exception(**kwargs: object) -> None:
nonlocal got_exception
if kwargs["request"] is request:
got_exception = True
request.body # noqa: B018 # prevents RawPostDataException
got_request_exception.connect(on_exception)
try:
response = super().get_response(request)
finally:
got_request_exception.disconnect(on_exception)
if (
not got_exception # Django will reraise this exception
and request.method != "OPTIONS"
and isinstance(response, HttpResponse)
and not (
response.status_code == 302 and response.headers["Location"].startswith("/login/")
)
):
openapi_request = DjangoOpenAPIRequest(request)
openapi_response = DjangoOpenAPIResponse(response)
response_validated = validate_test_response(openapi_request, openapi_response)
if response_validated:
validate_test_request(
openapi_request,
str(response.status_code),
request.META.get("intentionally_undocumented", False),
)
return response
class ZulipTestClient(TestClient):
def __init__(self) -> None:
super().__init__()
self.handler = ZulipClientHandler(enforce_csrf_checks=False)
class ZulipTestCaseMixin(SimpleTestCase):
# Ensure that the test system just shows us diffs
maxDiff: Optional[int] = None
migrations: Backfill extra_data_json for audit log entries. This migration is reasonably complex because of various anomalies in existing data. Note that there are cases when extra_data does not contain data that is proper json with possibly single quotes. Thus we need to use "ast.literal_eval" to cover that. There is also a special case for "event_type == USER_FULL_NAME_CHANGED", where extra_data is a plain str. This event_type is only used for RealmAuditLog, so the zilencer migration script does not need to handle it. The migration does not handle "event_type == REALM_DISCOUNT_CHANGED" because ast.literal_eval only allow Python literals. We expect the admin to populate the jsonified extra_data for extra_data_json manually beforehand. This chunks the backfilling migration to reduce potential block time. The migration for zilencer is mostly similar to the one for zerver; except that the backfill helper is added in a wrapper and unrelated events are removed. **Logging and error recovery** We print out a warning when the extra_data_json field of an entry would have been overwritten by a value inconsistent with what we derived from extra_data. Usually this only happens when the extra_data was corrupted before this migration. This prevents data loss by backing up possibly corrupted data in extra_data_json with the keys "inconsistent_old_extra_data" and "inconsistent_old_extra_data_json". More roundtrips to the database are needed for inconsistent data, which are expected to be infrequent. This also outputs messages when there are audit log entries with decimals, indicating that such entries are not backfilled. Do note that audit log entries with decimals are not populated with "inconsistent_old_extra_data_*" in the JSONField, because they are not overwritten. For such audit log entries with "extra_data_json" marked as inconsistent, we skip them in the migration. Because when we have discovered anomalies in a previous run, there is no need to overwrite them again nesting the extra keys we added to it. **Testing** We create a migration test case utilizing the property of bulk_create that it doesn't call our modified save method. We extend ZulipTestCase to support verifying console output at the test case level. The implementation is crude but the use case should be rare enough that we don't need it to be too elaborate. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2023-07-15 18:43:23 +02:00
# This bypasses BAN_CONSOLE_OUTPUT for the test case when set.
# Override this to verify if the given extra console output matches the
# expectation.
expected_console_output: Optional[str] = None
client_class = ZulipTestClient
@override
def setUp(self) -> None:
super().setUp()
python: Convert assignment type annotations to Python 3.6 style. This commit was split by tabbott; this piece covers the vast majority of files in Zulip, but excludes scripts/, tools/, and puppet/ to help ensure we at least show the right error messages for Xenial systems. We can likely further refine the remaining pieces with some testing. Generated by com2ann, with whitespace fixes and various manual fixes for runtime issues: - invoiced_through: Optional[LicenseLedger] = models.ForeignKey( + invoiced_through: Optional["LicenseLedger"] = models.ForeignKey( -_apns_client: Optional[APNsClient] = None +_apns_client: Optional["APNsClient"] = None - notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - signup_notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + signup_notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - author: Optional[UserProfile] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) + author: Optional["UserProfile"] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) - bot_owner: Optional[UserProfile] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) + bot_owner: Optional["UserProfile"] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) - default_sending_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) - default_events_register_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_sending_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_events_register_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) -descriptors_by_handler_id: Dict[int, ClientDescriptor] = {} +descriptors_by_handler_id: Dict[int, "ClientDescriptor"] = {} -worker_classes: Dict[str, Type[QueueProcessingWorker]] = {} -queues: Dict[str, Dict[str, Type[QueueProcessingWorker]]] = {} +worker_classes: Dict[str, Type["QueueProcessingWorker"]] = {} +queues: Dict[str, Dict[str, Type["QueueProcessingWorker"]]] = {} -AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional[LDAPSearch] = None +AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional["LDAPSearch"] = None Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
2020-04-22 01:09:50 +02:00
self.API_KEYS: Dict[str, str] = {}
test_name = self.id()
bounce_key_prefix_for_testing(test_name)
bounce_redis_key_prefix_for_testing(test_name)
@override
def tearDown(self) -> None:
super().tearDown()
# Important: we need to clear event queues to avoid leaking data to future tests.
clear_client_event_queues_for_testing()
clear_supported_auth_backends_cache()
flush_per_request_caches()
translation.activate(settings.LANGUAGE_CODE)
# Clean up local uploads directory after tests:
assert settings.LOCAL_UPLOADS_DIR is not None
if os.path.exists(settings.LOCAL_UPLOADS_DIR):
shutil.rmtree(settings.LOCAL_UPLOADS_DIR)
# Clean up after using fakeldap in LDAP tests:
if hasattr(self, "mock_ldap") and hasattr(self, "mock_initialize"):
if self.mock_ldap is not None:
self.mock_ldap.reset()
self.mock_initialize.stop()
def get_user_from_email(self, email: str, realm: Realm) -> UserProfile:
return get_user(email, realm)
@override
def run(self, result: Optional[TestResult] = None) -> Optional[TestResult]: # nocoverage
migrations: Backfill extra_data_json for audit log entries. This migration is reasonably complex because of various anomalies in existing data. Note that there are cases when extra_data does not contain data that is proper json with possibly single quotes. Thus we need to use "ast.literal_eval" to cover that. There is also a special case for "event_type == USER_FULL_NAME_CHANGED", where extra_data is a plain str. This event_type is only used for RealmAuditLog, so the zilencer migration script does not need to handle it. The migration does not handle "event_type == REALM_DISCOUNT_CHANGED" because ast.literal_eval only allow Python literals. We expect the admin to populate the jsonified extra_data for extra_data_json manually beforehand. This chunks the backfilling migration to reduce potential block time. The migration for zilencer is mostly similar to the one for zerver; except that the backfill helper is added in a wrapper and unrelated events are removed. **Logging and error recovery** We print out a warning when the extra_data_json field of an entry would have been overwritten by a value inconsistent with what we derived from extra_data. Usually this only happens when the extra_data was corrupted before this migration. This prevents data loss by backing up possibly corrupted data in extra_data_json with the keys "inconsistent_old_extra_data" and "inconsistent_old_extra_data_json". More roundtrips to the database are needed for inconsistent data, which are expected to be infrequent. This also outputs messages when there are audit log entries with decimals, indicating that such entries are not backfilled. Do note that audit log entries with decimals are not populated with "inconsistent_old_extra_data_*" in the JSONField, because they are not overwritten. For such audit log entries with "extra_data_json" marked as inconsistent, we skip them in the migration. Because when we have discovered anomalies in a previous run, there is no need to overwrite them again nesting the extra keys we added to it. **Testing** We create a migration test case utilizing the property of bulk_create that it doesn't call our modified save method. We extend ZulipTestCase to support verifying console output at the test case level. The implementation is crude but the use case should be rare enough that we don't need it to be too elaborate. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2023-07-15 18:43:23 +02:00
if not settings.BAN_CONSOLE_OUTPUT and self.expected_console_output is None:
return super().run(result)
extra_output_finder = ExtraConsoleOutputFinder()
with tee_stderr_and_find_extra_console_output(
extra_output_finder
), tee_stdout_and_find_extra_console_output(extra_output_finder):
test_result = super().run(result)
if extra_output_finder.full_extra_output and (
test_result is None or test_result.wasSuccessful()
):
migrations: Backfill extra_data_json for audit log entries. This migration is reasonably complex because of various anomalies in existing data. Note that there are cases when extra_data does not contain data that is proper json with possibly single quotes. Thus we need to use "ast.literal_eval" to cover that. There is also a special case for "event_type == USER_FULL_NAME_CHANGED", where extra_data is a plain str. This event_type is only used for RealmAuditLog, so the zilencer migration script does not need to handle it. The migration does not handle "event_type == REALM_DISCOUNT_CHANGED" because ast.literal_eval only allow Python literals. We expect the admin to populate the jsonified extra_data for extra_data_json manually beforehand. This chunks the backfilling migration to reduce potential block time. The migration for zilencer is mostly similar to the one for zerver; except that the backfill helper is added in a wrapper and unrelated events are removed. **Logging and error recovery** We print out a warning when the extra_data_json field of an entry would have been overwritten by a value inconsistent with what we derived from extra_data. Usually this only happens when the extra_data was corrupted before this migration. This prevents data loss by backing up possibly corrupted data in extra_data_json with the keys "inconsistent_old_extra_data" and "inconsistent_old_extra_data_json". More roundtrips to the database are needed for inconsistent data, which are expected to be infrequent. This also outputs messages when there are audit log entries with decimals, indicating that such entries are not backfilled. Do note that audit log entries with decimals are not populated with "inconsistent_old_extra_data_*" in the JSONField, because they are not overwritten. For such audit log entries with "extra_data_json" marked as inconsistent, we skip them in the migration. Because when we have discovered anomalies in a previous run, there is no need to overwrite them again nesting the extra keys we added to it. **Testing** We create a migration test case utilizing the property of bulk_create that it doesn't call our modified save method. We extend ZulipTestCase to support verifying console output at the test case level. The implementation is crude but the use case should be rare enough that we don't need it to be too elaborate. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2023-07-15 18:43:23 +02:00
extra_output = extra_output_finder.full_extra_output.decode(errors="replace")
if self.expected_console_output is not None:
self.assertEqual(extra_output, self.expected_console_output)
return test_result
exception_message = f"""
---- UNEXPECTED CONSOLE OUTPUT DETECTED ----
To ensure that we never miss important error output/warnings,
we require test-backend to have clean console output.
This message usually is triggered by forgotten debugging print()
statements or new logging statements. For the latter, you can
use `with self.assertLogs()` to capture and verify the log output;
use `git grep assertLogs` to see dozens of correct examples.
You should be able to quickly reproduce this failure with:
./tools/test-backend --ban-console-output {self.id()}
Output:
migrations: Backfill extra_data_json for audit log entries. This migration is reasonably complex because of various anomalies in existing data. Note that there are cases when extra_data does not contain data that is proper json with possibly single quotes. Thus we need to use "ast.literal_eval" to cover that. There is also a special case for "event_type == USER_FULL_NAME_CHANGED", where extra_data is a plain str. This event_type is only used for RealmAuditLog, so the zilencer migration script does not need to handle it. The migration does not handle "event_type == REALM_DISCOUNT_CHANGED" because ast.literal_eval only allow Python literals. We expect the admin to populate the jsonified extra_data for extra_data_json manually beforehand. This chunks the backfilling migration to reduce potential block time. The migration for zilencer is mostly similar to the one for zerver; except that the backfill helper is added in a wrapper and unrelated events are removed. **Logging and error recovery** We print out a warning when the extra_data_json field of an entry would have been overwritten by a value inconsistent with what we derived from extra_data. Usually this only happens when the extra_data was corrupted before this migration. This prevents data loss by backing up possibly corrupted data in extra_data_json with the keys "inconsistent_old_extra_data" and "inconsistent_old_extra_data_json". More roundtrips to the database are needed for inconsistent data, which are expected to be infrequent. This also outputs messages when there are audit log entries with decimals, indicating that such entries are not backfilled. Do note that audit log entries with decimals are not populated with "inconsistent_old_extra_data_*" in the JSONField, because they are not overwritten. For such audit log entries with "extra_data_json" marked as inconsistent, we skip them in the migration. Because when we have discovered anomalies in a previous run, there is no need to overwrite them again nesting the extra keys we added to it. **Testing** We create a migration test case utilizing the property of bulk_create that it doesn't call our modified save method. We extend ZulipTestCase to support verifying console output at the test case level. The implementation is crude but the use case should be rare enough that we don't need it to be too elaborate. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2023-07-15 18:43:23 +02:00
{extra_output}
--------------------------------------------
"""
raise ExtraConsoleOutputInTestError(exception_message)
return test_result
"""
WRAPPER_COMMENT:
We wrap calls to self.client.{patch,put,get,post,delete} for various
reasons. Some of this has to do with fixing encodings before calling
into the Django code. Some of this has to do with providing a future
path for instrumentation. Some of it's just consistency.
The linter will prevent direct calls to self.client.foo, so the wrapper
functions have to fake out the linter by using a local variable called
django_client to fool the regex.
"""
DEFAULT_SUBDOMAIN = "zulip"
TOKENIZED_NOREPLY_REGEX = settings.TOKENIZED_NOREPLY_EMAIL_ADDRESS.format(
token=r"[a-z0-9_]{24}"
)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
def set_http_headers(self, extra: Dict[str, str], skip_user_agent: bool = False) -> None:
if "subdomain" in extra:
assert isinstance(extra["subdomain"], str)
extra["HTTP_HOST"] = Realm.host_for_subdomain(extra["subdomain"])
del extra["subdomain"]
elif "HTTP_HOST" not in extra:
extra["HTTP_HOST"] = Realm.host_for_subdomain(self.DEFAULT_SUBDOMAIN)
# set User-Agent
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
if "HTTP_AUTHORIZATION" in extra:
# An API request; use mobile as the default user agent
default_user_agent = "ZulipMobile/26.22.145 (iOS 10.3.1)"
else:
# A web app request; use a browser User-Agent string.
default_user_agent = (
"Mozilla/5.0 (Windows NT 10.0; Win64; x64)"
" AppleWebKit/537.36 (KHTML, like Gecko)"
" Chrome/79.0.3945.130 Safari/537.36"
)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
if skip_user_agent:
# Provide a way to disable setting User-Agent if desired.
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
assert "HTTP_USER_AGENT" not in extra
elif "HTTP_USER_AGENT" not in extra:
extra["HTTP_USER_AGENT"] = default_user_agent
@instrument_url
def client_patch(
self,
url: str,
info: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
intentionally_undocumented: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
"""
We need to urlencode, since Django's function won't do it for us.
"""
encoded = urlencode(info)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["content_type"] = "application/x-www-form-urlencoded"
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.patch(
url,
encoded,
follow=follow,
secure=secure,
headers=headers,
intentionally_undocumented=intentionally_undocumented,
**extra,
)
@instrument_url
def client_patch_multipart(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
info: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented: bool = False,
**extra: str,
) -> "TestHttpResponse":
"""
Use this for patch requests that have file uploads or
that need some sort of multi-part content. In the future
Django's test client may become a bit more flexible,
so we can hopefully eliminate this. (When you post
with the Django test client, it deals with MULTIPART_CONTENT
automatically, but not patch.)
"""
encoded = encode_multipart(BOUNDARY, dict(info))
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.patch(
url,
encoded,
content_type=MULTIPART_CONTENT,
follow=follow,
secure=secure,
headers=headers,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=intentionally_undocumented,
**extra,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
)
def json_patch(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
payload: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
**extra: str,
) -> "TestHttpResponse":
data = orjson.dumps(payload)
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.patch(
url,
data=data,
content_type="application/json",
follow=follow,
secure=secure,
headers=None,
**extra,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
)
@instrument_url
def client_put(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
info: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
encoded = urlencode(info)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["content_type"] = "application/x-www-form-urlencoded"
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.put(
url, encoded, follow=follow, secure=secure, headers=headers, **extra
)
def json_put(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
payload: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
data = orjson.dumps(payload)
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.put(
url,
data=data,
content_type="application/json",
follow=follow,
secure=secure,
headers=headers,
**extra,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
)
@instrument_url
def client_delete(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
info: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented: bool = False,
**extra: str,
) -> "TestHttpResponse":
encoded = urlencode(info)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["content_type"] = "application/x-www-form-urlencoded"
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.delete(
url,
encoded,
follow=follow,
secure=secure,
headers={
"Content-Type": "application/x-www-form-urlencoded", # https://code.djangoproject.com/ticket/33230
**(headers or {}),
},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=intentionally_undocumented,
**extra,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
)
@instrument_url
def client_options(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
info: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.options(
url, dict(info), follow=follow, secure=secure, headers=headers, **extra
)
@instrument_url
def client_head(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
info: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.head(url, info, follow=follow, secure=secure, headers=headers, **extra)
@instrument_url
def client_post(
self,
url: str,
info: Union[str, bytes, Mapping[str, Any]] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented: bool = False,
content_type: Optional[str] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
encoded = info
if content_type is None:
if isinstance(info, dict) and not any(
hasattr(value, "read") and callable(value.read) for value in info.values()
):
content_type = "application/x-www-form-urlencoded"
encoded = urlencode(info, doseq=True)
else:
content_type = MULTIPART_CONTENT
return django_client.post(
url,
encoded,
follow=follow,
secure=secure,
headers={
"Content-Type": content_type, # https://code.djangoproject.com/ticket/33230
**(headers or {}),
},
content_type=content_type,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=intentionally_undocumented,
**extra,
)
@instrument_url
def client_post_request(self, url: str, req: Any) -> "TestHttpResponse":
"""
We simulate hitting an endpoint here, although we
actually resolve the URL manually and hit the view
directly. We have this helper method to allow our
instrumentation to work for /notify_tornado and
future similar methods that require doing funny
things to a request object.
"""
match = resolve(url)
return match.func(req)
@instrument_url
def client_get(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self,
url: str,
info: Mapping[str, Any] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
skip_user_agent: bool = False,
follow: bool = False,
secure: bool = False,
headers: Optional[Mapping[str, Any]] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented: bool = False,
**extra: str,
) -> "TestHttpResponse":
django_client = self.client # see WRAPPER_COMMENT
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self.set_http_headers(extra, skip_user_agent)
return django_client.get(
url,
info,
follow=follow,
secure=secure,
headers=headers,
intentionally_undocumented=intentionally_undocumented,
**extra,
)
example_user_map = dict(
hamlet="hamlet@zulip.com",
cordelia="cordelia@zulip.com",
iago="iago@zulip.com",
prospero="prospero@zulip.com",
othello="othello@zulip.com",
AARON="AARON@zulip.com",
aaron="aaron@zulip.com",
ZOE="ZOE@zulip.com",
polonius="polonius@zulip.com",
desdemona="desdemona@zulip.com",
shiva="shiva@zulip.com",
webhook_bot="webhook-bot@zulip.com",
outgoing_webhook_bot="outgoing-webhook@zulip.com",
default_bot="default-bot@zulip.com",
)
2017-05-23 01:26:38 +02:00
mit_user_map = dict(
sipbtest="sipbtest@mit.edu",
starnine="starnine@mit.edu",
espuser="espuser@mit.edu",
2017-05-23 01:26:38 +02:00
)
lear_user_map = dict(
cordelia="cordelia@zulip.com",
king="king@lear.org",
)
# Non-registered test users
nonreg_user_map = dict(
test="test@zulip.com",
test1="test1@zulip.com",
alice="alice@zulip.com",
newuser="newuser@zulip.com",
bob="bob@zulip.com",
cordelia="cordelia@zulip.com",
newguy="newguy@zulip.com",
me="me@zulip.com",
)
example_user_ldap_username_map = dict(
hamlet="hamlet",
cordelia="cordelia",
# aaron's uid in our test directory is "letham".
aaron="letham",
)
def nonreg_user(self, name: str) -> UserProfile:
email = self.nonreg_user_map[name]
return get_user_by_delivery_email(email, get_realm("zulip"))
def example_user(self, name: str) -> UserProfile:
email = self.example_user_map[name]
return get_user_by_delivery_email(email, get_realm("zulip"))
2017-05-23 01:26:38 +02:00
def mit_user(self, name: str) -> UserProfile:
2017-05-23 01:26:38 +02:00
email = self.mit_user_map[name]
return self.get_user_from_email(email, get_realm("zephyr"))
def lear_user(self, name: str) -> UserProfile:
email = self.lear_user_map[name]
return self.get_user_from_email(email, get_realm("lear"))
def nonreg_email(self, name: str) -> str:
return self.nonreg_user_map[name]
def example_email(self, name: str) -> str:
return self.example_user_map[name]
def mit_email(self, name: str) -> str:
return self.mit_user_map[name]
def notification_bot(self, realm: Realm) -> UserProfile:
return get_system_bot(settings.NOTIFICATION_BOT, realm.id)
def create_test_bot(
self, short_name: str, user_profile: UserProfile, full_name: str = "Foo Bot", **extras: Any
) -> UserProfile:
self.login_user(user_profile)
bot_info = {
"short_name": short_name,
"full_name": full_name,
}
bot_info.update(extras)
result = self.client_post("/json/bots", bot_info)
self.assert_json_success(result)
bot_email = f"{short_name}-bot@zulip.testserver"
bot_profile = self.get_user_from_email(bot_email, user_profile.realm)
return bot_profile
def fail_to_create_test_bot(
self,
short_name: str,
user_profile: UserProfile,
full_name: str = "Foo Bot",
*,
assert_json_error_msg: str,
**extras: Any,
) -> None:
self.login_user(user_profile)
bot_info = {
"short_name": short_name,
"full_name": full_name,
}
bot_info.update(extras)
result = self.client_post("/json/bots", bot_info)
self.assert_json_error(result, assert_json_error_msg)
def _get_page_params(self, result: "TestHttpResponse") -> Dict[str, Any]:
"""Helper for parsing page_params after fetching the web app's home view."""
doc = lxml.html.document_fromstring(result.content)
div = cast(lxml.html.HtmlMixin, doc).get_element_by_id("page-params")
assert div is not None
page_params_json = div.get("data-params")
assert page_params_json is not None
page_params = orjson.loads(page_params_json)
return page_params
def check_rendered_logged_in_app(self, result: "TestHttpResponse") -> None:
"""Verifies that a visit of / was a 200 that rendered page_params
and not for a (logged-out) spectator."""
self.assertEqual(result.status_code, 200)
page_params = self._get_page_params(result)
# It is important to check `is_spectator` to verify
# that we treated this request as a normal logged-in session,
# not as a spectator.
self.assertEqual(page_params["is_spectator"], False)
def login_with_return(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
self, email: str, password: Optional[str] = None, **extra: str
) -> "TestHttpResponse":
if password is None:
password = initial_password(email)
result = self.client_post(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
"/accounts/login/",
{"username": email, "password": password},
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
self.assertNotEqual(result.status_code, 500)
return result
def login(self, name: str) -> None:
"""
Use this for really simple tests where you just need
to be logged in as some user, but don't need the actual
user object for anything else. Try to use 'hamlet' for
non-admins and 'iago' for admins:
self.login('hamlet')
Try to use 'cordelia' or 'othello' as "other" users.
"""
assert "@" not in name, "use login_by_email for email logins"
user = self.example_user(name)
self.login_user(user)
def login_by_email(self, email: str, password: str) -> None:
realm = get_realm("zulip")
request = HttpRequest()
request.session = self.client.session
self.assertTrue(
self.client.login(
request=request,
username=email,
password=password,
realm=realm,
),
)
def assert_login_failure(self, email: str, password: str) -> None:
realm = get_realm("zulip")
request = HttpRequest()
request.session = self.client.session
self.assertFalse(
self.client.login(
request=request,
username=email,
password=password,
realm=realm,
),
)
def login_user(self, user_profile: UserProfile) -> None:
email = user_profile.delivery_email
realm = user_profile.realm
password = initial_password(email)
request = HttpRequest()
request.session = self.client.session
self.assertTrue(
self.client.login(request=request, username=email, password=password, realm=realm)
)
2017-07-13 13:42:57 +02:00
def login_2fa(self, user_profile: UserProfile) -> None:
"""
We need this function to call request.session.save().
do_two_factor_login doesn't save session; in normal request-response
cycle this doesn't matter because middleware will save the session
when it finds it dirty; however,in tests we will have to do that
explicitly.
"""
request = HttpRequest()
request.session = self.client.session
request.user = user_profile
do_two_factor_login(request, user_profile)
request.session.save()
def logout(self) -> None:
self.client.logout()
def register(self, email: str, password: str, subdomain: str = DEFAULT_SUBDOMAIN) -> None:
response = self.client_post("/accounts/home/", {"email": email}, subdomain=subdomain)
self.assertEqual(response.status_code, 302)
self.assertEqual(response["Location"], f"/accounts/send_confirm/?email={quote(email)}")
response = self.submit_reg_form_for_user(email, password, subdomain=subdomain)
self.assertEqual(response.status_code, 302)
self.assertEqual(response["Location"], f"http://{Realm.host_for_subdomain(subdomain)}/")
def submit_reg_form_for_user(
self,
email: str,
password: Optional[str],
realm_name: str = "Zulip Test",
realm_subdomain: str = "zuliptest",
from_confirmation: str = "",
full_name: Optional[str] = None,
timezone: str = "",
realm_in_root_domain: Optional[str] = None,
default_stream_groups: Sequence[str] = [],
source_realm_id: str = "",
key: Optional[str] = None,
realm_type: int = Realm.ORG_TYPES["business"]["id"],
realm_default_language: str = "en",
enable_marketing_emails: Optional[bool] = None,
email_address_visibility: Optional[int] = None,
is_demo_organization: bool = False,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
"""
Stage two of the two-step registration process.
If things are working correctly the account should be fully
registered after this call.
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
You can pass the HTTP_HOST variable for subdomains via extra.
"""
if full_name is None:
full_name = email.replace("@", "_")
payload = {
"full_name": full_name,
"realm_name": realm_name,
"realm_subdomain": realm_subdomain,
"realm_type": realm_type,
"realm_default_language": realm_default_language,
"key": key if key is not None else find_key_by_email(email),
"timezone": timezone,
"terms": True,
"from_confirmation": from_confirmation,
"default_stream_group": default_stream_groups,
"source_realm_id": source_realm_id,
"is_demo_organization": is_demo_organization,
"how_realm_creator_found_zulip": "other",
"how_realm_creator_found_zulip_extra_context": "I found it on the internet.",
}
if enable_marketing_emails is not None:
payload["enable_marketing_emails"] = enable_marketing_emails
if email_address_visibility is not None:
payload["email_address_visibility"] = email_address_visibility
if password is not None:
payload["password"] = password
if realm_in_root_domain is not None:
payload["realm_in_root_domain"] = realm_in_root_domain
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
return self.client_post(
"/accounts/register/",
payload,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
def submit_realm_creation_form(
self,
email: str,
*,
realm_subdomain: str,
realm_name: str,
realm_type: int = Realm.ORG_TYPES["business"]["id"],
realm_default_language: str = "en",
realm_in_root_domain: Optional[str] = None,
) -> "TestHttpResponse":
payload = {
"email": email,
"realm_name": realm_name,
"realm_type": realm_type,
"realm_default_language": realm_default_language,
"realm_subdomain": realm_subdomain,
}
if realm_in_root_domain is not None:
payload["realm_in_root_domain"] = realm_in_root_domain
return self.client_post(
"/new/",
payload,
)
def get_confirmation_url_from_outbox(
self,
email_address: str,
*,
url_pattern: Optional[str] = None,
email_subject_contains: Optional[str] = None,
email_body_contains: Optional[str] = None,
) -> str:
from django.core.mail import outbox
if url_pattern is None:
# This is a bit of a crude heuristic, but good enough for most tests.
url_pattern = settings.EXTERNAL_HOST + r"(\S+)>"
for message in reversed(outbox):
if any(
addr == email_address or addr.endswith(f" <{email_address}>") for addr in message.to
):
match = re.search(url_pattern, str(message.body))
assert match is not None
if email_subject_contains:
self.assertIn(email_subject_contains, message.subject)
if email_body_contains:
self.assertIn(email_body_contains, message.body)
[confirmation_url] = match.groups()
return confirmation_url
raise AssertionError("Couldn't find a confirmation email.")
def encode_uuid(self, uuid: str) -> str:
"""
identifier: Can be an email or a remote server uuid.
"""
if uuid in self.API_KEYS:
api_key = self.API_KEYS[uuid]
else:
api_key = get_remote_server_by_uuid(uuid).api_key
self.API_KEYS[uuid] = api_key
return self.encode_credentials(uuid, api_key)
def encode_user(self, user: UserProfile) -> str:
email = user.delivery_email
api_key = user.api_key
return self.encode_credentials(email, api_key)
def encode_email(self, email: str, realm: str = "zulip") -> str:
# TODO: use encode_user where possible
assert "@" in email
user = get_user_by_delivery_email(email, get_realm(realm))
api_key = get_api_key(user)
return self.encode_credentials(email, api_key)
def encode_credentials(self, identifier: str, api_key: str) -> str:
"""
identifier: Can be an email or a remote server uuid.
"""
credentials = f"{identifier}:{api_key}"
return "Basic " + base64.b64encode(credentials.encode()).decode()
def uuid_get(
self, identifier: str, url: str, info: Mapping[str, Any] = {}, **extra: str
) -> "TestHttpResponse":
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["HTTP_AUTHORIZATION"] = self.encode_uuid(identifier)
return self.client_get(
url,
info,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
def uuid_post(
self,
identifier: str,
url: str,
info: Union[str, bytes, Mapping[str, Any]] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["HTTP_AUTHORIZATION"] = self.encode_uuid(identifier)
return self.client_post(
url,
info,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
def api_get(
self, user: UserProfile, url: str, info: Mapping[str, Any] = {}, **extra: str
) -> "TestHttpResponse":
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["HTTP_AUTHORIZATION"] = self.encode_user(user)
return self.client_get(
url,
info,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
def api_post(
self,
user: UserProfile,
url: str,
info: Union[str, bytes, Mapping[str, Any]] = {},
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented: bool = False,
**extra: str,
) -> "TestHttpResponse":
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["HTTP_AUTHORIZATION"] = self.encode_user(user)
return self.client_post(
url,
info,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=intentionally_undocumented,
**extra,
)
def api_patch(
self, user: UserProfile, url: str, info: Mapping[str, Any] = {}, **extra: str
) -> "TestHttpResponse":
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["HTTP_AUTHORIZATION"] = self.encode_user(user)
return self.client_patch(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
url,
info,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
def api_delete(
self, user: UserProfile, url: str, info: Mapping[str, Any] = {}, **extra: str
) -> "TestHttpResponse":
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["HTTP_AUTHORIZATION"] = self.encode_user(user)
return self.client_delete(
url,
info,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
def get_streams(self, user_profile: UserProfile) -> List[str]:
"""
Helper function to get the active stream names for a user
"""
return list(
Stream.objects.filter(
id__in=get_subscribed_stream_ids_for_user(user_profile)
).values_list("name", flat=True)
)
def send_personal_message(
self,
from_user: UserProfile,
to_user: UserProfile,
content: str = "test content",
*,
read_by_sender: bool = True,
) -> int:
recipient_list = [to_user.id]
(sending_client, _) = Client.objects.get_or_create(name="test suite")
sent_message_result = check_send_message(
from_user,
sending_client,
"private",
recipient_list,
None,
content,
read_by_sender=read_by_sender,
)
return sent_message_result.message_id
def send_huddle_message(
self,
from_user: UserProfile,
to_users: List[UserProfile],
content: str = "test content",
*,
read_by_sender: bool = True,
) -> int:
to_user_ids = [u.id for u in to_users]
assert len(to_user_ids) >= 2
(sending_client, _) = Client.objects.get_or_create(name="test suite")
sent_message_result = check_send_message(
from_user,
sending_client,
"private",
to_user_ids,
None,
content,
read_by_sender=read_by_sender,
)
return sent_message_result.message_id
def send_stream_message(
self,
sender: UserProfile,
stream_name: str,
content: str = "test content",
topic_name: str = "test",
recipient_realm: Optional[Realm] = None,
*,
tests: Ensure stream senders get a UserMessage row. We now complain if a test author sends a stream message that does not result in the sender getting a UserMessage row for the message. This is basically 100% equivalent to complaining that the author failed to subscribe the sender to the stream as part of the test setup, as far as I can tell, so the AssertionError instructs the author to subscribe the sender to the stream. We exempt bots from this check, although it is plausible we should only exempt the system bots like the notification bot. I considered auto-subscribing the sender to the stream, but that can be a little more expensive than the current check, and we generally want test setup to be explicit. If there is some legitimate way than a subscribed human sender can't get a UserMessage, then we probably want an explicit test for that, or we may want to change the backend to just write a UserMessage row in that hypothetical situation. For most tests, including almost all the ones fixed here, the author just wants their test setup to realistically reflect normal operation, and often devs may not realize that Cordelia is not subscribed to Denmark or not realize that Hamlet is not subscribed to Scotland. Some of us don't remember our Shakespeare from high school, and our stream subscriptions don't even necessarily reflect which countries the Bard placed his characters in. There may also be some legitimate use case where an author wants to simulate sending a message to an unsubscribed stream, but for those edge cases, they can always set allow_unsubscribed_sender to True.
2021-12-10 13:55:48 +01:00
allow_unsubscribed_sender: bool = False,
read_by_sender: bool = True,
) -> int:
(sending_client, _) = Client.objects.get_or_create(name="test suite")
message_id = check_send_stream_message(
sender=sender,
client=sending_client,
stream_name=stream_name,
topic_name=topic_name,
body=content,
realm=recipient_realm,
read_by_sender=read_by_sender,
)
if (
not UserMessage.objects.filter(user_profile=sender, message_id=message_id).exists()
and not sender.is_bot
and not allow_unsubscribed_sender
):
raise AssertionError(
f"""
tests: Ensure stream senders get a UserMessage row. We now complain if a test author sends a stream message that does not result in the sender getting a UserMessage row for the message. This is basically 100% equivalent to complaining that the author failed to subscribe the sender to the stream as part of the test setup, as far as I can tell, so the AssertionError instructs the author to subscribe the sender to the stream. We exempt bots from this check, although it is plausible we should only exempt the system bots like the notification bot. I considered auto-subscribing the sender to the stream, but that can be a little more expensive than the current check, and we generally want test setup to be explicit. If there is some legitimate way than a subscribed human sender can't get a UserMessage, then we probably want an explicit test for that, or we may want to change the backend to just write a UserMessage row in that hypothetical situation. For most tests, including almost all the ones fixed here, the author just wants their test setup to realistically reflect normal operation, and often devs may not realize that Cordelia is not subscribed to Denmark or not realize that Hamlet is not subscribed to Scotland. Some of us don't remember our Shakespeare from high school, and our stream subscriptions don't even necessarily reflect which countries the Bard placed his characters in. There may also be some legitimate use case where an author wants to simulate sending a message to an unsubscribed stream, but for those edge cases, they can always set allow_unsubscribed_sender to True.
2021-12-10 13:55:48 +01:00
It appears that the sender did not get a UserMessage row, which is
almost certainly an artificial symptom that in your test setup you
have decided to send a message to a stream without the sender being
subscribed.
Please do self.subscribe(<user for {sender.full_name}>, {stream_name!r}) first.
tests: Ensure stream senders get a UserMessage row. We now complain if a test author sends a stream message that does not result in the sender getting a UserMessage row for the message. This is basically 100% equivalent to complaining that the author failed to subscribe the sender to the stream as part of the test setup, as far as I can tell, so the AssertionError instructs the author to subscribe the sender to the stream. We exempt bots from this check, although it is plausible we should only exempt the system bots like the notification bot. I considered auto-subscribing the sender to the stream, but that can be a little more expensive than the current check, and we generally want test setup to be explicit. If there is some legitimate way than a subscribed human sender can't get a UserMessage, then we probably want an explicit test for that, or we may want to change the backend to just write a UserMessage row in that hypothetical situation. For most tests, including almost all the ones fixed here, the author just wants their test setup to realistically reflect normal operation, and often devs may not realize that Cordelia is not subscribed to Denmark or not realize that Hamlet is not subscribed to Scotland. Some of us don't remember our Shakespeare from high school, and our stream subscriptions don't even necessarily reflect which countries the Bard placed his characters in. There may also be some legitimate use case where an author wants to simulate sending a message to an unsubscribed stream, but for those edge cases, they can always set allow_unsubscribed_sender to True.
2021-12-10 13:55:48 +01:00
Or choose a stream that the user is already subscribed to:
{self.subscribed_stream_name_list(sender)}
"""
)
tests: Ensure stream senders get a UserMessage row. We now complain if a test author sends a stream message that does not result in the sender getting a UserMessage row for the message. This is basically 100% equivalent to complaining that the author failed to subscribe the sender to the stream as part of the test setup, as far as I can tell, so the AssertionError instructs the author to subscribe the sender to the stream. We exempt bots from this check, although it is plausible we should only exempt the system bots like the notification bot. I considered auto-subscribing the sender to the stream, but that can be a little more expensive than the current check, and we generally want test setup to be explicit. If there is some legitimate way than a subscribed human sender can't get a UserMessage, then we probably want an explicit test for that, or we may want to change the backend to just write a UserMessage row in that hypothetical situation. For most tests, including almost all the ones fixed here, the author just wants their test setup to realistically reflect normal operation, and often devs may not realize that Cordelia is not subscribed to Denmark or not realize that Hamlet is not subscribed to Scotland. Some of us don't remember our Shakespeare from high school, and our stream subscriptions don't even necessarily reflect which countries the Bard placed his characters in. There may also be some legitimate use case where an author wants to simulate sending a message to an unsubscribed stream, but for those edge cases, they can always set allow_unsubscribed_sender to True.
2021-12-10 13:55:48 +01:00
return message_id
def get_messages_response(
self,
anchor: Union[int, str] = 1,
num_before: int = 100,
num_after: int = 100,
use_first_unread_anchor: bool = False,
include_anchor: bool = True,
) -> Dict[str, List[Dict[str, Any]]]:
post_params = {
"anchor": anchor,
"num_before": num_before,
"num_after": num_after,
"use_first_unread_anchor": orjson.dumps(use_first_unread_anchor).decode(),
"include_anchor": orjson.dumps(include_anchor).decode(),
}
result = self.client_get("/json/messages", dict(post_params))
2017-08-17 08:46:39 +02:00
data = result.json()
return data
def get_messages(
self,
anchor: Union[str, int] = 1,
num_before: int = 100,
num_after: int = 100,
use_first_unread_anchor: bool = False,
) -> List[Dict[str, Any]]:
data = self.get_messages_response(anchor, num_before, num_after, use_first_unread_anchor)
return data["messages"]
def users_subscribed_to_stream(self, stream_name: str, realm: Realm) -> List[UserProfile]:
stream = Stream.objects.get(name=stream_name, realm=realm)
recipient = Recipient.objects.get(type_id=stream.id, type=Recipient.STREAM)
subscriptions = Subscription.objects.filter(recipient=recipient, active=True)
return [subscription.user_profile for subscription in subscriptions]
def not_long_term_idle_subscriber_ids(self, stream_name: str, realm: Realm) -> Set[int]:
stream = Stream.objects.get(name=stream_name, realm=realm)
recipient = Recipient.objects.get(type_id=stream.id, type=Recipient.STREAM)
subscriptions = Subscription.objects.filter(
recipient=recipient, active=True, is_user_active=True
).exclude(user_profile__long_term_idle=True)
user_profile_ids = set(subscriptions.values_list("user_profile_id", flat=True))
return user_profile_ids
def assert_json_success(
self,
result: Union["TestHttpResponse", HttpResponse],
*,
ignored_parameters: Optional[List[str]] = None,
) -> Dict[str, Any]:
"""
Successful POSTs return a 200 and JSON of the form {"result": "success",
"msg": ""}.
"""
try:
json = orjson.loads(result.content)
except orjson.JSONDecodeError: # nocoverage
json = {"msg": "Error parsing JSON in response!"}
self.assertEqual(result.status_code, 200, json["msg"])
self.assertEqual(json.get("result"), "success")
# We have a msg key for consistency with errors, but it typically has an
# empty value.
self.assertIn("msg", json)
self.assertNotEqual(json["msg"], "Error parsing JSON in response!")
# Check ignored parameters.
if ignored_parameters is None:
self.assertNotIn("ignored_parameters_unsupported", json)
else:
self.assertIn("ignored_parameters_unsupported", json)
self.assert_length(json["ignored_parameters_unsupported"], len(ignored_parameters))
for param in ignored_parameters:
self.assertTrue(param in json["ignored_parameters_unsupported"])
return json
def get_json_error(self, result: "TestHttpResponse", status_code: int = 400) -> str:
try:
json = orjson.loads(result.content)
except orjson.JSONDecodeError: # nocoverage
json = {"msg": "Error parsing JSON in response!"}
self.assertEqual(result.status_code, status_code, msg=json.get("msg"))
self.assertEqual(json.get("result"), "error")
return json["msg"]
def assert_json_error(
self, result: "TestHttpResponse", msg: str, status_code: int = 400
) -> None:
"""
Invalid POSTs return an error status code and JSON of the form
{"result": "error", "msg": "reason"}.
"""
self.assertEqual(self.get_json_error(result, status_code=status_code), msg)
def assert_length(self, items: Collection[Any] | ValuesQuerySet[Any, Any], count: int) -> None:
actual_count = len(items)
if actual_count != count: # nocoverage
print("\nITEMS:\n")
for item in items:
print(item)
print(f"\nexpected length: {count}\nactual length: {actual_count}")
raise AssertionError(f"{type(items)} is of unexpected size!")
@contextmanager
def assert_memcached_count(self, count: int) -> Iterator[None]:
with cache_tries_captured() as cache_tries:
yield
self.assert_length(cache_tries, count)
@contextmanager
def assert_database_query_count(
self, count: int, include_savepoints: bool = False, keep_cache_warm: bool = False
) -> Iterator[None]:
"""
This captures the queries executed and check the total number of queries.
Useful when minimizing unnecessary roundtrips to the database is important.
"""
with queries_captured(
include_savepoints=include_savepoints, keep_cache_warm=keep_cache_warm
) as queries:
yield
actual_count = len(queries)
if actual_count != count: # nocoverage
print("\nITEMS:\n")
for index, query in enumerate(queries):
print(f"#{index + 1}\nsql: {query.sql}\ntime: {query.time}\n")
print(f"expected count: {count}\nactual count: {actual_count}")
raise AssertionError(
f"""
{count} queries expected, got {actual_count}.
This is a performance-critical code path, where we check
the number of database queries used in order to avoid accidental regressions.
If an unnecessary query was removed or the new query is necessary, you should
update this test, and explain what queries we added/removed in the pull request
and why any new queries can't be avoided."""
)
def assert_json_error_contains(
self, result: "TestHttpResponse", msg_substring: str, status_code: int = 400
) -> None:
self.assertIn(msg_substring, self.get_json_error(result, status_code=status_code))
def assert_in_response(
self, substring: str, response: Union["TestHttpResponse", HttpResponse]
) -> None:
self.assertIn(substring, response.content.decode())
def assert_in_success_response(
self, substrings: List[str], response: Union["TestHttpResponse", HttpResponse]
) -> None:
self.assertEqual(response.status_code, 200)
decoded = response.content.decode()
for substring in substrings:
self.assertIn(substring, decoded)
def assert_not_in_success_response(
self, substrings: List[str], response: Union["TestHttpResponse", HttpResponse]
) -> None:
self.assertEqual(response.status_code, 200)
decoded = response.content.decode()
for substring in substrings:
self.assertNotIn(substring, decoded)
def assert_logged_in_user_id(self, user_id: Optional[int]) -> None:
"""
Verifies the user currently logged in for the test client has the provided user_id.
Pass None to verify no user is logged in.
"""
self.assertEqual(get_session_dict_user(self.client.session), user_id)
def assert_message_stream_name(self, message: Message, stream_name: str) -> None:
self.assertEqual(message.recipient.type, Recipient.STREAM)
stream_id = message.recipient.type_id
stream = Stream.objects.get(id=stream_id)
self.assertEqual(stream.recipient_id, message.recipient_id)
self.assertEqual(stream.name, stream_name)
def webhook_fixture_data(self, type: str, action: str, file_type: str = "json") -> str:
fn = os.path.join(
os.path.dirname(__file__),
f"../webhooks/{type}/fixtures/{action}.{file_type}",
)
with open(fn) as f:
return f.read()
def fixture_file_name(self, file_name: str, type: str = "") -> str:
return os.path.join(
os.path.dirname(__file__),
f"../tests/fixtures/{type}/{file_name}",
)
def fixture_data(self, file_name: str, type: str = "") -> str:
fn = self.fixture_file_name(file_name, type)
with open(fn) as f:
return f.read()
def make_stream(
self,
stream_name: str,
realm: Optional[Realm] = None,
invite_only: bool = False,
is_web_public: bool = False,
history_public_to_subscribers: Optional[bool] = None,
) -> Stream:
if realm is None:
realm = get_realm("zulip")
history_public_to_subscribers = get_default_value_for_history_public_to_subscribers(
realm, invite_only, history_public_to_subscribers
)
administrators_user_group = NamedUserGroup.objects.get(
name=SystemGroups.ADMINISTRATORS, realm=realm, is_system_group=True
)
try:
stream = Stream.objects.create(
realm=realm,
name=stream_name,
invite_only=invite_only,
is_web_public=is_web_public,
history_public_to_subscribers=history_public_to_subscribers,
can_remove_subscribers_group=administrators_user_group,
)
except IntegrityError: # nocoverage -- this is for bugs in the tests
raise Exception(
f"""
{stream_name} already exists
Please call make_stream with a stream name
that is not already in use."""
)
recipient = Recipient.objects.create(type_id=stream.id, type=Recipient.STREAM)
stream.recipient = recipient
stream.save(update_fields=["recipient"])
return stream
INVALID_STREAM_ID = 999999
def get_stream_id(self, name: str, realm: Optional[Realm] = None) -> int:
if not realm:
realm = get_realm("zulip")
try:
stream = get_realm_stream(name, realm.id)
except Stream.DoesNotExist:
return self.INVALID_STREAM_ID
return stream.id
# Subscribe to a stream directly
def subscribe(
self,
user_profile: UserProfile,
stream_name: str,
invite_only: bool = False,
is_web_public: bool = False,
) -> Stream:
realm = user_profile.realm
try:
stream = get_stream(stream_name, user_profile.realm)
except Stream.DoesNotExist:
stream, from_stream_creation = create_stream_if_needed(
realm, stream_name, invite_only=invite_only, is_web_public=is_web_public
)
bulk_add_subscriptions(realm, [stream], [user_profile], acting_user=None)
return stream
def unsubscribe(self, user_profile: UserProfile, stream_name: str) -> None:
realm = user_profile.realm
stream = get_stream(stream_name, user_profile.realm)
bulk_remove_subscriptions(realm, [user_profile], [stream], acting_user=None)
# Subscribe to a stream by making an API request
def common_subscribe_to_streams(
self,
user: UserProfile,
streams: Iterable[str],
extra_post_data: Mapping[str, Any] = {},
invite_only: bool = False,
is_web_public: bool = False,
allow_fail: bool = False,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> "TestHttpResponse":
post_data = {
"subscriptions": orjson.dumps([{"name": stream} for stream in streams]).decode(),
"is_web_public": orjson.dumps(is_web_public).decode(),
"invite_only": orjson.dumps(invite_only).decode(),
}
post_data.update(extra_post_data)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
result = self.api_post(
user,
"/api/v1/users/me/subscriptions",
post_data,
intentionally_undocumented=False,
**extra,
)
if not allow_fail:
self.assert_json_success(result)
return result
tests: Ensure stream senders get a UserMessage row. We now complain if a test author sends a stream message that does not result in the sender getting a UserMessage row for the message. This is basically 100% equivalent to complaining that the author failed to subscribe the sender to the stream as part of the test setup, as far as I can tell, so the AssertionError instructs the author to subscribe the sender to the stream. We exempt bots from this check, although it is plausible we should only exempt the system bots like the notification bot. I considered auto-subscribing the sender to the stream, but that can be a little more expensive than the current check, and we generally want test setup to be explicit. If there is some legitimate way than a subscribed human sender can't get a UserMessage, then we probably want an explicit test for that, or we may want to change the backend to just write a UserMessage row in that hypothetical situation. For most tests, including almost all the ones fixed here, the author just wants their test setup to realistically reflect normal operation, and often devs may not realize that Cordelia is not subscribed to Denmark or not realize that Hamlet is not subscribed to Scotland. Some of us don't remember our Shakespeare from high school, and our stream subscriptions don't even necessarily reflect which countries the Bard placed his characters in. There may also be some legitimate use case where an author wants to simulate sending a message to an unsubscribed stream, but for those edge cases, they can always set allow_unsubscribed_sender to True.
2021-12-10 13:55:48 +01:00
def subscribed_stream_name_list(self, user: UserProfile) -> str:
# This is currently only used for producing error messages.
subscribed_streams = gather_subscriptions(user)[0]
return "".join(sorted(f" * {stream['name']}\n" for stream in subscribed_streams))
def check_user_subscribed_only_to_streams(self, user_name: str, streams: List[Stream]) -> None:
streams = sorted(streams, key=lambda x: x.name)
subscribed_streams = gather_subscriptions(self.nonreg_user(user_name))[0]
self.assert_length(subscribed_streams, len(streams))
for x, y in zip(subscribed_streams, streams):
self.assertEqual(x["name"], y.name)
def resolve_topic_containing_message(
self,
acting_user: UserProfile,
target_message_id: int,
**extra: str,
) -> "TestHttpResponse":
"""
Mark all messages within the topic associated with message `target_message_id` as resolved.
"""
message = access_message(acting_user, target_message_id)
return self.api_patch(
acting_user,
f"/api/v1/messages/{target_message_id}",
{
"topic": RESOLVED_TOPIC_PREFIX + message.topic_name(),
"propagate_mode": "change_all",
},
**extra,
)
def send_webhook_payload(
self,
user_profile: UserProfile,
url: str,
payload: Union[str, Dict[str, Any]],
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> Message:
"""
Send a webhook payload to the server, and verify that the
post is successful.
This is a pretty low-level function. For most use cases
see the helpers that call this function, which do additional
checks.
Occasionally tests will call this directly, for unique
situations like having multiple messages go to a stream,
where the other helper functions are a bit too rigid,
and you'll want the test itself do various assertions.
Even in those cases, you're often better to simply
call client_post and assert_json_success.
If the caller expects a message to be sent to a stream,
the caller should make sure the user is subscribed.
"""
prior_msg = self.get_last_message()
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
result = self.client_post(
url,
payload,
skip_user_agent=False,
follow=False,
secure=False,
headers=None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
intentionally_undocumented=False,
**extra,
)
self.assert_json_success(result)
# Check the correct message was sent
msg = self.get_last_message()
if msg.id == prior_msg.id:
raise EmptyResponseError(
"""
Your test code called an endpoint that did
not write any new messages. It is probably
broken (but still returns 200 due to exception
handling).
One possible gotcha is that you forgot to
subscribe the test user to the stream that
the webhook sends to.
"""
) # nocoverage
self.assertEqual(msg.sender.email, user_profile.email)
return msg
def get_last_message(self) -> Message:
return Message.objects.latest("id")
def get_second_to_last_message(self) -> Message:
return Message.objects.all().order_by("-id")[1]
@contextmanager
def simulated_markdown_failure(self) -> Iterator[None]:
"""
This raises a failure inside of the try/except block of
markdown.__init__.do_convert.
"""
with mock.patch(
"zerver.lib.markdown.unsafe_timeout", side_effect=subprocess.CalledProcessError(1, [])
), self.assertLogs(level="ERROR"): # For markdown_logger.exception
yield
def create_default_device(
self, user_profile: UserProfile, number: str = "+12125550100"
) -> None:
phone_device = PhoneDevice(
user=user_profile,
name="default",
confirmed=True,
number=number,
key="abcd",
method="sms",
)
2017-07-13 13:42:57 +02:00
phone_device.save()
def rm_tree(self, path: str) -> None:
if os.path.exists(path):
shutil.rmtree(path)
def make_import_output_dir(self, exported_from: str) -> str:
output_dir = tempfile.mkdtemp(
dir=settings.TEST_WORKER_DIR, prefix="test-" + exported_from + "-import-"
)
os.makedirs(output_dir, exist_ok=True)
return output_dir
2019-05-21 12:21:32 +02:00
def get_set(self, data: List[Dict[str, Any]], field: str) -> Set[str]:
values = {r[field] for r in data}
2019-05-21 12:21:32 +02:00
return values
def find_by_id(self, data: List[Dict[str, Any]], db_id: int) -> Dict[str, Any]:
[r] = (r for r in data if r["id"] == db_id)
return r
def init_default_ldap_database(self) -> None:
"""
Takes care of the mock_ldap setup, loads
a directory from zerver/tests/fixtures/ldap/directory.json with various entries
to be used by tests.
If a test wants to specify its own directory, it can just replace
self.mock_ldap.directory with its own content, but in most cases it should be
enough to use change_user_attr to make simple modifications to the pre-loaded
directory. If new user entries are needed to test for some additional unusual
scenario, it's most likely best to add that to directory.json.
"""
directory = orjson.loads(self.fixture_data("directory.json", type="ldap"))
for attrs in directory.values():
if "uid" in attrs:
# Generate a password for the LDAP account:
attrs["userPassword"] = [self.ldap_password(attrs["uid"][0])]
# Load binary attributes. If in "directory", an attribute as its value
# has a string starting with "file:", the rest of the string is assumed
# to be a path to the file from which binary data should be loaded,
# as the actual value of the attribute in LDAP.
for attr, value in attrs.items():
if isinstance(value, str) and value.startswith("file:"):
with open(value[5:], "rb") as f:
attrs[attr] = [f.read()]
ldap_patcher = mock.patch("django_auth_ldap.config.ldap.initialize")
self.mock_initialize = ldap_patcher.start()
self.mock_ldap = MockLDAP(directory)
self.mock_initialize.return_value = self.mock_ldap
def change_ldap_user_attr(
self, username: str, attr_name: str, attr_value: Union[str, bytes], binary: bool = False
) -> None:
"""
Method for changing the value of an attribute of a user entry in the mock
directory. Use option binary=True if you want binary data to be loaded
into the attribute from a file specified at attr_value. This changes
the attribute only for the specific test function that calls this method,
and is isolated from other tests.
"""
dn = f"uid={username},ou=users,dc=zulip,dc=com"
if binary:
with open(attr_value, "rb") as f:
# attr_value should be a path to the file with the binary data
python: Convert assignment type annotations to Python 3.6 style. This commit was split by tabbott; this piece covers the vast majority of files in Zulip, but excludes scripts/, tools/, and puppet/ to help ensure we at least show the right error messages for Xenial systems. We can likely further refine the remaining pieces with some testing. Generated by com2ann, with whitespace fixes and various manual fixes for runtime issues: - invoiced_through: Optional[LicenseLedger] = models.ForeignKey( + invoiced_through: Optional["LicenseLedger"] = models.ForeignKey( -_apns_client: Optional[APNsClient] = None +_apns_client: Optional["APNsClient"] = None - notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - signup_notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + signup_notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - author: Optional[UserProfile] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) + author: Optional["UserProfile"] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) - bot_owner: Optional[UserProfile] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) + bot_owner: Optional["UserProfile"] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) - default_sending_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) - default_events_register_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_sending_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_events_register_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) -descriptors_by_handler_id: Dict[int, ClientDescriptor] = {} +descriptors_by_handler_id: Dict[int, "ClientDescriptor"] = {} -worker_classes: Dict[str, Type[QueueProcessingWorker]] = {} -queues: Dict[str, Dict[str, Type[QueueProcessingWorker]]] = {} +worker_classes: Dict[str, Type["QueueProcessingWorker"]] = {} +queues: Dict[str, Dict[str, Type["QueueProcessingWorker"]]] = {} -AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional[LDAPSearch] = None +AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional["LDAPSearch"] = None Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
2020-04-22 01:09:50 +02:00
data: Union[str, bytes] = f.read()
else:
data = attr_value
self.mock_ldap.directory[dn][attr_name] = [data]
def remove_ldap_user_attr(self, username: str, attr_name: str) -> None:
"""
Method for removing the value of an attribute of a user entry in the mock
directory. This changes the attribute only for the specific test function
that calls this method, and is isolated from other tests.
"""
dn = f"uid={username},ou=users,dc=zulip,dc=com"
self.mock_ldap.directory[dn].pop(attr_name, None)
def ldap_username(self, username: str) -> str:
"""
Maps Zulip username to the name of the corresponding LDAP user
in our test directory at zerver/tests/fixtures/ldap/directory.json,
if the LDAP user exists.
"""
return self.example_user_ldap_username_map[username]
def ldap_password(self, uid: str) -> str:
return f"{uid}_ldap_password"
def email_display_from(self, email_message: EmailMessage) -> str:
"""
Returns the email address that will show in email clients as the
"From" field.
"""
# The extra_headers field may contain a "From" which is used
# for display in email clients, and appears in the RFC822
# header as `From`. The `.from_email` accessor is the
# "envelope from" address, used by mail transfer agents if
# the email bounces.
return email_message.extra_headers.get("From", email_message.from_email)
def email_envelope_from(self, email_message: EmailMessage) -> str:
"""
Returns the email address that will be used if the email bounces.
"""
# See email_display_from, above.
return email_message.from_email
def check_has_permission_policies(
self, policy: str, validation_func: Callable[[UserProfile], bool]
) -> None:
realm = get_realm("zulip")
owner = "desdemona"
admin = "iago"
moderator = "shiva"
member = "hamlet"
new_member = "othello"
guest = "polonius"
def set_age(user_name: str, age: int) -> None:
user = self.example_user(user_name)
user.date_joined = timezone_now() - timedelta(days=age)
user.save()
do_set_realm_property(realm, "waiting_period_threshold", 1000, acting_user=None)
set_age(member, age=realm.waiting_period_threshold + 1)
set_age(new_member, age=realm.waiting_period_threshold - 1)
def allow(user_name: str) -> None:
# Fetch a clean object for the user.
user = self.example_user(user_name)
with self.assert_database_query_count(0):
self.assertTrue(validation_func(user))
def prevent(user_name: str) -> None:
# Fetch a clean object for the user.
user = self.example_user(user_name)
with self.assert_database_query_count(0):
self.assertFalse(validation_func(user))
def set_policy(level: int) -> None:
do_set_realm_property(realm, policy, level, acting_user=None)
set_policy(Realm.POLICY_NOBODY)
prevent(owner)
prevent(admin)
prevent(moderator)
prevent(member)
prevent(new_member)
prevent(guest)
set_policy(Realm.POLICY_OWNERS_ONLY)
allow(owner)
prevent(admin)
prevent(moderator)
prevent(member)
prevent(new_member)
prevent(guest)
set_policy(Realm.POLICY_ADMINS_ONLY)
allow(owner)
allow(admin)
prevent(moderator)
prevent(member)
prevent(new_member)
prevent(guest)
set_policy(Realm.POLICY_MODERATORS_ONLY)
allow(owner)
allow(admin)
allow(moderator)
prevent(member)
prevent(new_member)
prevent(guest)
set_policy(Realm.POLICY_FULL_MEMBERS_ONLY)
allow(owner)
allow(admin)
allow(moderator)
allow(member)
prevent(new_member)
prevent(guest)
set_policy(Realm.POLICY_MEMBERS_ONLY)
allow(owner)
allow(admin)
allow(moderator)
allow(member)
allow(new_member)
prevent(guest)
set_policy(Realm.POLICY_EVERYONE)
allow(owner)
allow(admin)
allow(moderator)
allow(member)
allow(new_member)
allow(guest)
def subscribe_realm_to_manual_license_management_plan(
self, realm: Realm, licenses: int, licenses_at_next_renewal: int, billing_schedule: int
) -> Tuple[CustomerPlan, LicenseLedger]:
customer, _ = Customer.objects.get_or_create(realm=realm)
plan = CustomerPlan.objects.create(
customer=customer,
automanage_licenses=False,
billing_cycle_anchor=timezone_now(),
billing_schedule=billing_schedule,
tier=CustomerPlan.TIER_CLOUD_STANDARD,
)
ledger = LicenseLedger.objects.create(
plan=plan,
is_renewal=True,
event_time=timezone_now(),
licenses=licenses,
licenses_at_next_renewal=licenses_at_next_renewal,
)
realm.plan_type = Realm.PLAN_TYPE_STANDARD
realm.save(update_fields=["plan_type"])
return plan, ledger
def subscribe_realm_to_monthly_plan_on_manual_license_management(
self, realm: Realm, licenses: int, licenses_at_next_renewal: int
) -> Tuple[CustomerPlan, LicenseLedger]:
return self.subscribe_realm_to_manual_license_management_plan(
realm, licenses, licenses_at_next_renewal, CustomerPlan.BILLING_SCHEDULE_MONTHLY
)
def create_user_notifications_data_object(
self, *, user_id: int, **kwargs: Any
) -> UserMessageNotificationsData:
return UserMessageNotificationsData(
user_id=user_id,
online_push_enabled=kwargs.get("online_push_enabled", False),
dm_email_notify=kwargs.get("dm_email_notify", False),
dm_push_notify=kwargs.get("dm_push_notify", False),
notifications: Calculate PMs/mentions settings like other settings. Previously, we checked for the `enable_offline_email_notifications` and `enable_offline_push_notifications` settings (which determine whether the user will receive notifications for PMs and mentions) just before sending notifications. This has a few problem: 1. We do not have access to all the user settings in the notification handlers (`handle_missedmessage_emails` and `handle_push_notifications`), and therefore, we cannot correctly determine whether the notification should be sent. Checks like the following which existed previously, will, for example, incorrectly not send notifications even when stream email notifications are enabled- ``` if not receives_offline_email_notifications(user_profile): return ``` With this commit, we simply do not enqueue notifications if the "offline" settings are disabled, which fixes that bug. Additionally, this also fixes a bug with the "online push notifications" feature, which was, if someone were to: * turn off notifications for PMs and mentions (`enable_offline_push_notifications`) * turn on stream push notifications (`enable_stream_push_notifications`) * turn on "online push" (`enable_online_push_notifications`) then, they would still receive notifications for PMs when online. This isn't how the "online push enabled" feature is supposed to work; it should only act as a wrapper around the other notification settings. The buggy code was this in `handle_push_notifications`: ``` if not ( receives_offline_push_notifications(user_profile) or receives_online_push_notifications(user_profile) ): return // send notifications ``` This commit removes that code, and extends our `notification_data.py` logic to cover this case, along with tests. 2. The name for these settings is slightly misleading. They essentially talk about "what to send notifications for" (PMs and mentions), and not "when to send notifications" (offline). This commit improves this condition by restricting the use of this term only to the database field, and using clearer names everywhere else. This distinction will be important to have non-confusing code when we implement multiple options for notifications in the future as dropdown (never/when offline/when offline or online, etc). 3. We should ideally re-check all notification settings just before the notifications are sent. This is especially important for email notifications, which may be sent after a long time after the message was sent. We will in the future add code to thoroughly re-check settings before sending notifications in a clean manner, but temporarily not re-checking isn't a terrible scenario either.
2021-07-14 15:34:01 +02:00
mention_email_notify=kwargs.get("mention_email_notify", False),
mention_push_notify=kwargs.get("mention_push_notify", False),
topic_wildcard_mention_email_notify=kwargs.get(
"topic_wildcard_mention_email_notify", False
),
topic_wildcard_mention_push_notify=kwargs.get(
"topic_wildcard_mention_push_notify", False
),
stream_wildcard_mention_email_notify=kwargs.get(
"stream_wildcard_mention_email_notify", False
),
stream_wildcard_mention_push_notify=kwargs.get(
"stream_wildcard_mention_push_notify", False
),
stream_email_notify=kwargs.get("stream_email_notify", False),
stream_push_notify=kwargs.get("stream_push_notify", False),
followed_topic_email_notify=kwargs.get("followed_topic_email_notify", False),
followed_topic_push_notify=kwargs.get("followed_topic_push_notify", False),
topic_wildcard_mention_in_followed_topic_email_notify=kwargs.get(
"topic_wildcard_mention_in_followed_topic_email_notify", False
),
topic_wildcard_mention_in_followed_topic_push_notify=kwargs.get(
"topic_wildcard_mention_in_followed_topic_push_notify", False
),
stream_wildcard_mention_in_followed_topic_email_notify=kwargs.get(
"stream_wildcard_mention_in_followed_topic_email_notify", False
),
stream_wildcard_mention_in_followed_topic_push_notify=kwargs.get(
"stream_wildcard_mention_in_followed_topic_push_notify", False
),
sender_is_muted=kwargs.get("sender_is_muted", False),
disable_external_notifications=kwargs.get("disable_external_notifications", False),
)
maybe_enqueue_notifications: Take in notification_data dataclass. * Modify `maybe_enqueue_notifications` to take in an instance of the dataclass introduced in 951b49c048ba3464e74ad7965da3453fe36d0a96. * The `check_notify` tests tested the "when to notify" logic in a way which involved `maybe_enqueue_notifications`. To simplify things, we've earlier extracted this logic in 8182632d7e9f8490b9b9295e01b5912dcf173fd5. So, we just kill off the `check_notify` test, and keep only those parts which verify the queueing and return value behavior of that funtion. * We retain the the missedmessage_hook and message message_edit_notifications since they are more integration-style. * There's a slightly subtle change with the missedmessage_hook tests. Before this commit, we short-circuited the hook if the sender was muted (5a642cea115be159175d1189f83ba25d2c5c7632). With this commit, we delegate the check to our dataclass methods. So, `maybe_enqueue_notifications` will be called even if the sender was muted, and the test needs to be updated. * In our test helper `get_maybe_enqueue_notifications_parameters` which generates default values for testing `maybe_enqueue_notifications` calls, we keep `message_id`, `sender_id`, and `user_id` as required arguments, so that the tests are super-clear and avoid accidental false positives. * Because `do_update_embedded_data` also sends `update_message` events, we deal with that case with some hacky code for now. See the comment there. This mostly completes the extraction of the "when to notify" logic into our new `notification_data` module.
2021-06-23 14:12:32 +02:00
def get_maybe_enqueue_notifications_parameters(
self, *, message_id: int, user_id: int, acting_user_id: int, **kwargs: Any
maybe_enqueue_notifications: Take in notification_data dataclass. * Modify `maybe_enqueue_notifications` to take in an instance of the dataclass introduced in 951b49c048ba3464e74ad7965da3453fe36d0a96. * The `check_notify` tests tested the "when to notify" logic in a way which involved `maybe_enqueue_notifications`. To simplify things, we've earlier extracted this logic in 8182632d7e9f8490b9b9295e01b5912dcf173fd5. So, we just kill off the `check_notify` test, and keep only those parts which verify the queueing and return value behavior of that funtion. * We retain the the missedmessage_hook and message message_edit_notifications since they are more integration-style. * There's a slightly subtle change with the missedmessage_hook tests. Before this commit, we short-circuited the hook if the sender was muted (5a642cea115be159175d1189f83ba25d2c5c7632). With this commit, we delegate the check to our dataclass methods. So, `maybe_enqueue_notifications` will be called even if the sender was muted, and the test needs to be updated. * In our test helper `get_maybe_enqueue_notifications_parameters` which generates default values for testing `maybe_enqueue_notifications` calls, we keep `message_id`, `sender_id`, and `user_id` as required arguments, so that the tests are super-clear and avoid accidental false positives. * Because `do_update_embedded_data` also sends `update_message` events, we deal with that case with some hacky code for now. See the comment there. This mostly completes the extraction of the "when to notify" logic into our new `notification_data` module.
2021-06-23 14:12:32 +02:00
) -> Dict[str, Any]:
"""
Returns a dictionary with the passed parameters, after filling up the
missing data with default values, for testing what was passed to the
`maybe_enqueue_notifications` method.
"""
maybe_enqueue_notifications: Take in notification_data dataclass. * Modify `maybe_enqueue_notifications` to take in an instance of the dataclass introduced in 951b49c048ba3464e74ad7965da3453fe36d0a96. * The `check_notify` tests tested the "when to notify" logic in a way which involved `maybe_enqueue_notifications`. To simplify things, we've earlier extracted this logic in 8182632d7e9f8490b9b9295e01b5912dcf173fd5. So, we just kill off the `check_notify` test, and keep only those parts which verify the queueing and return value behavior of that funtion. * We retain the the missedmessage_hook and message message_edit_notifications since they are more integration-style. * There's a slightly subtle change with the missedmessage_hook tests. Before this commit, we short-circuited the hook if the sender was muted (5a642cea115be159175d1189f83ba25d2c5c7632). With this commit, we delegate the check to our dataclass methods. So, `maybe_enqueue_notifications` will be called even if the sender was muted, and the test needs to be updated. * In our test helper `get_maybe_enqueue_notifications_parameters` which generates default values for testing `maybe_enqueue_notifications` calls, we keep `message_id`, `sender_id`, and `user_id` as required arguments, so that the tests are super-clear and avoid accidental false positives. * Because `do_update_embedded_data` also sends `update_message` events, we deal with that case with some hacky code for now. See the comment there. This mostly completes the extraction of the "when to notify" logic into our new `notification_data` module.
2021-06-23 14:12:32 +02:00
user_notifications_data = self.create_user_notifications_data_object(
user_id=user_id, **kwargs
)
return dict(
user_notifications_data=user_notifications_data,
maybe_enqueue_notifications: Take in notification_data dataclass. * Modify `maybe_enqueue_notifications` to take in an instance of the dataclass introduced in 951b49c048ba3464e74ad7965da3453fe36d0a96. * The `check_notify` tests tested the "when to notify" logic in a way which involved `maybe_enqueue_notifications`. To simplify things, we've earlier extracted this logic in 8182632d7e9f8490b9b9295e01b5912dcf173fd5. So, we just kill off the `check_notify` test, and keep only those parts which verify the queueing and return value behavior of that funtion. * We retain the the missedmessage_hook and message message_edit_notifications since they are more integration-style. * There's a slightly subtle change with the missedmessage_hook tests. Before this commit, we short-circuited the hook if the sender was muted (5a642cea115be159175d1189f83ba25d2c5c7632). With this commit, we delegate the check to our dataclass methods. So, `maybe_enqueue_notifications` will be called even if the sender was muted, and the test needs to be updated. * In our test helper `get_maybe_enqueue_notifications_parameters` which generates default values for testing `maybe_enqueue_notifications` calls, we keep `message_id`, `sender_id`, and `user_id` as required arguments, so that the tests are super-clear and avoid accidental false positives. * Because `do_update_embedded_data` also sends `update_message` events, we deal with that case with some hacky code for now. See the comment there. This mostly completes the extraction of the "when to notify" logic into our new `notification_data` module.
2021-06-23 14:12:32 +02:00
message_id=message_id,
acting_user_id=acting_user_id,
mentioned_user_group_id=kwargs.get("mentioned_user_group_id", None),
maybe_enqueue_notifications: Take in notification_data dataclass. * Modify `maybe_enqueue_notifications` to take in an instance of the dataclass introduced in 951b49c048ba3464e74ad7965da3453fe36d0a96. * The `check_notify` tests tested the "when to notify" logic in a way which involved `maybe_enqueue_notifications`. To simplify things, we've earlier extracted this logic in 8182632d7e9f8490b9b9295e01b5912dcf173fd5. So, we just kill off the `check_notify` test, and keep only those parts which verify the queueing and return value behavior of that funtion. * We retain the the missedmessage_hook and message message_edit_notifications since they are more integration-style. * There's a slightly subtle change with the missedmessage_hook tests. Before this commit, we short-circuited the hook if the sender was muted (5a642cea115be159175d1189f83ba25d2c5c7632). With this commit, we delegate the check to our dataclass methods. So, `maybe_enqueue_notifications` will be called even if the sender was muted, and the test needs to be updated. * In our test helper `get_maybe_enqueue_notifications_parameters` which generates default values for testing `maybe_enqueue_notifications` calls, we keep `message_id`, `sender_id`, and `user_id` as required arguments, so that the tests are super-clear and avoid accidental false positives. * Because `do_update_embedded_data` also sends `update_message` events, we deal with that case with some hacky code for now. See the comment there. This mostly completes the extraction of the "when to notify" logic into our new `notification_data` module.
2021-06-23 14:12:32 +02:00
idle=kwargs.get("idle", True),
already_notified=kwargs.get(
"already_notified", {"email_notified": False, "push_notified": False}
),
)
2021-12-05 11:43:34 +01:00
def verify_emoji_code_foreign_keys(self) -> None:
"""
DB tables that refer to RealmEmoji use int(emoji_code) as the
foreign key. Those tables tend to de-normalize emoji_name due
to our inheritance-based setup. This helper makes sure those
2021-12-05 11:43:34 +01:00
invariants are intact, which is particularly tricky during
the import/export process (or during conversions from things
like Slack/RocketChat/MatterMost/etc.).
"""
dct = {}
for realm_emoji in RealmEmoji.objects.all():
dct[realm_emoji.id] = realm_emoji
2021-12-05 11:43:34 +01:00
if not dct:
raise AssertionError("test needs RealmEmoji rows")
count = 0
for reaction in Reaction.objects.filter(reaction_type=Reaction.REALM_EMOJI):
realm_emoji_id = int(reaction.emoji_code)
2021-12-05 11:43:34 +01:00
assert realm_emoji_id in dct
self.assertEqual(dct[realm_emoji_id].name, reaction.emoji_name)
self.assertEqual(dct[realm_emoji_id].realm_id, reaction.user_profile.realm_id)
2021-12-05 11:43:34 +01:00
count += 1
for user_status in UserStatus.objects.filter(reaction_type=UserStatus.REALM_EMOJI):
realm_emoji_id = int(user_status.emoji_code)
assert realm_emoji_id in dct
self.assertEqual(dct[realm_emoji_id].name, user_status.emoji_name)
self.assertEqual(dct[realm_emoji_id].realm_id, user_status.user_profile.realm_id)
count += 1
2021-12-05 11:43:34 +01:00
if count == 0:
raise AssertionError("test is meaningless without any pertinent rows")
def check_user_added_in_system_group(self, user: UserProfile) -> None:
user_group = get_system_user_group_for_user(user)
self.assertTrue(
UserGroupMembership.objects.filter(user_profile=user, user_group=user_group).exists()
)
def _assert_long_term_idle(self, user: UserProfile) -> None:
if not user.long_term_idle:
raise AssertionError(
"""
We expect you to explicitly call self.soft_deactivate_user
if your user is not already soft-deactivated.
"""
)
def expect_soft_reactivation(self, user: UserProfile, action: Callable[[], None]) -> None:
self._assert_long_term_idle(user)
action()
# Prevent from using the old user object
user.refresh_from_db()
self.assertEqual(user.long_term_idle, False)
def expect_to_stay_long_term_idle(self, user: UserProfile, action: Callable[[], None]) -> None:
self._assert_long_term_idle(user)
action()
# Prevent from using the old user object
user.refresh_from_db()
self.assertEqual(user.long_term_idle, True)
def soft_deactivate_user(self, user: UserProfile) -> None:
do_soft_deactivate_users([user])
assert user.long_term_idle
def set_up_db_for_testing_user_access(self) -> None:
polonius = self.example_user("polonius")
hamlet = self.example_user("hamlet")
othello = self.example_user("othello")
iago = self.example_user("iago")
prospero = self.example_user("prospero")
aaron = self.example_user("aaron")
zoe = self.example_user("ZOE")
shiva = self.example_user("shiva")
realm = get_realm("zulip")
# Polonius is subscribed to "Verona" by default, so we unsubscribe
# it so that it becomes easier to test the restricted access.
self.unsubscribe(polonius, "Verona")
self.make_stream("test_stream1")
self.make_stream("test_stream2", invite_only=True)
self.subscribe(othello, "test_stream1")
self.send_stream_message(othello, "test_stream1", content="test message", topic_name="test")
self.unsubscribe(othello, "test_stream1")
self.subscribe(polonius, "test_stream1")
self.subscribe(polonius, "test_stream2")
self.subscribe(hamlet, "test_stream1")
self.subscribe(iago, "test_stream2")
self.send_personal_message(polonius, prospero)
self.send_personal_message(shiva, polonius)
self.send_huddle_message(aaron, [polonius, zoe])
members_group = NamedUserGroup.objects.get(name="role:members", realm=realm)
do_change_realm_permission_group_setting(
realm, "can_access_all_users_group", members_group, acting_user=None
)
class ZulipTestCase(ZulipTestCaseMixin, TestCase):
@contextmanager
def capture_send_event_calls(
self, expected_num_events: int
) -> Iterator[List[Mapping[str, Any]]]:
lst: List[Mapping[str, Any]] = []
# process_notification takes a single parameter called 'notice'.
# lst.append takes a single argument called 'object'.
# Some code might call process_notification using keyword arguments,
# so mypy doesn't allow assigning lst.append to process_notification
# So explicitly change parameter name to 'notice' to work around this problem
with mock.patch("zerver.tornado.event_queue.process_notification", lst.append):
# Some `send_event` calls need to be executed only after the current transaction
# commits (using `on_commit` hooks). Because the transaction in Django tests never
# commits (rather, gets rolled back after the test completes), such events would
# never be sent in tests, and we would be unable to verify them. Hence, we use
# this helper to make sure the `send_event` calls actually run.
with self.captureOnCommitCallbacks(execute=True):
yield lst
self.assert_length(lst, expected_num_events)
user_groups: Make locks required for updating user group memberships. **Background** User groups are expected to comply with the DAG constraint for the many-to-many inter-group membership. The check for this constraint has to be performed recursively so that we can find all direct and indirect subgroups of the user group to be added. This kind of check is vulnerable to phantom reads which is possible at the default read committed isolation level because we cannot guarantee that the check is still valid when we are adding the subgroups to the user group. **Solution** To avoid having another transaction concurrently update one of the to-be-subgroup after the recursive check is done, and before the subgroup is added, we use SELECT FOR UPDATE to lock the user group rows. The lock needs to be acquired before a group membership change is about to occur before any check has been conducted. Suppose that we are adding subgroup B to supergroup A, the locking protocol is specified as follows: 1. Acquire a lock for B and all its direct and indirect subgroups. 2. Acquire a lock for A. For the removal of user groups, we acquire a lock for the user group to be removed with all its direct and indirect subgroups. This is the special case A=B, which is still complaint with the protocol. **Error handling** We currently rely on Postgres' deadlock detection to abort transactions and show an error for the users. In the future, we might need some recovery mechanism or at least better error handling. **Notes** An important note is that we need to reuse the recursive CTE query that finds the direct and indirect subgroups when applying the lock on the rows. And the lock needs to be acquired the same way for the addition and removal of direct subgroups. User membership change (as opposed to user group membership) is not affected. Read-only queries aren't either. The locks only protect critical regions where the user group dependency graph might violate the DAG constraint, where users are not participating. **Testing** We implement a transaction test case targeting some typical scenarios when an internal server error is expected to happen (this means that the user group view makes the correct decision to abort the transaction when something goes wrong with locks). To achieve this, we add a development view intended only for unit tests. It has a global BARRIER that can be shared across threads, so that we can synchronize them to consistently reproduce certain potential race conditions prevented by the database locks. The transaction test case lanuches pairs of threads initiating possibly conflicting requests at the same time. The tests are set up such that exactly N of them are expected to succeed with a certain error message (while we don't know each one). **Security notes** get_recursive_subgroups_for_groups will no longer fetch user groups from other realms. As a result, trying to add/remove a subgroup from another realm results in a UserGroup not found error response. We also implement subgroup-specific checks in has_user_group_access to keep permission managing in a single place. Do note that the API currently don't have a way to violate that check because we are only checking the realm ID now.
2023-06-17 04:39:52 +02:00
def get_row_ids_in_all_tables() -> Iterator[Tuple[str, Set[int]]]:
all_models = apps.get_models(include_auto_created=True)
ignored_tables = {"django_session"}
for model in all_models:
table_name = model._meta.db_table
if table_name in ignored_tables:
continue
ids = model._default_manager.all().values_list("id", flat=True)
yield table_name, set(ids)
class ZulipTransactionTestCase(ZulipTestCaseMixin, TransactionTestCase):
"""The default Django TestCase wraps each test in a transaction. This
is invaluable for being able to rollback the transaction and thus
efficiently do many tests containing database changes, but it
prevents testing certain transaction-related races and locking
bugs.
This test class is intended to be used (sparingly!) for tests that
need to verify transaction related behavior, like locking with
select_for_update or transaction.atomic(durable=True).
Unlike ZulipTestCase, ZulipTransactionTestCase has the following traits:
1. Does not offer isolation between tests by wrapping them inside an atomic transaction.
2. Changes are committed to the current worker's test database, so side effects carry on.
All ZulipTransactionTestCase tests must be carefully written to
avoid side effects on the database; while Django runs
TransactionTestCase after all normal TestCase tests on a given
test worker to avoid pollution, they can break other
ZulipTransactionTestCase tests if they leak state.
"""
@override
user_groups: Make locks required for updating user group memberships. **Background** User groups are expected to comply with the DAG constraint for the many-to-many inter-group membership. The check for this constraint has to be performed recursively so that we can find all direct and indirect subgroups of the user group to be added. This kind of check is vulnerable to phantom reads which is possible at the default read committed isolation level because we cannot guarantee that the check is still valid when we are adding the subgroups to the user group. **Solution** To avoid having another transaction concurrently update one of the to-be-subgroup after the recursive check is done, and before the subgroup is added, we use SELECT FOR UPDATE to lock the user group rows. The lock needs to be acquired before a group membership change is about to occur before any check has been conducted. Suppose that we are adding subgroup B to supergroup A, the locking protocol is specified as follows: 1. Acquire a lock for B and all its direct and indirect subgroups. 2. Acquire a lock for A. For the removal of user groups, we acquire a lock for the user group to be removed with all its direct and indirect subgroups. This is the special case A=B, which is still complaint with the protocol. **Error handling** We currently rely on Postgres' deadlock detection to abort transactions and show an error for the users. In the future, we might need some recovery mechanism or at least better error handling. **Notes** An important note is that we need to reuse the recursive CTE query that finds the direct and indirect subgroups when applying the lock on the rows. And the lock needs to be acquired the same way for the addition and removal of direct subgroups. User membership change (as opposed to user group membership) is not affected. Read-only queries aren't either. The locks only protect critical regions where the user group dependency graph might violate the DAG constraint, where users are not participating. **Testing** We implement a transaction test case targeting some typical scenarios when an internal server error is expected to happen (this means that the user group view makes the correct decision to abort the transaction when something goes wrong with locks). To achieve this, we add a development view intended only for unit tests. It has a global BARRIER that can be shared across threads, so that we can synchronize them to consistently reproduce certain potential race conditions prevented by the database locks. The transaction test case lanuches pairs of threads initiating possibly conflicting requests at the same time. The tests are set up such that exactly N of them are expected to succeed with a certain error message (while we don't know each one). **Security notes** get_recursive_subgroups_for_groups will no longer fetch user groups from other realms. As a result, trying to add/remove a subgroup from another realm results in a UserGroup not found error response. We also implement subgroup-specific checks in has_user_group_access to keep permission managing in a single place. Do note that the API currently don't have a way to violate that check because we are only checking the realm ID now.
2023-06-17 04:39:52 +02:00
def setUp(self) -> None:
super().setUp()
self.models_ids_set = dict(get_row_ids_in_all_tables())
@override
user_groups: Make locks required for updating user group memberships. **Background** User groups are expected to comply with the DAG constraint for the many-to-many inter-group membership. The check for this constraint has to be performed recursively so that we can find all direct and indirect subgroups of the user group to be added. This kind of check is vulnerable to phantom reads which is possible at the default read committed isolation level because we cannot guarantee that the check is still valid when we are adding the subgroups to the user group. **Solution** To avoid having another transaction concurrently update one of the to-be-subgroup after the recursive check is done, and before the subgroup is added, we use SELECT FOR UPDATE to lock the user group rows. The lock needs to be acquired before a group membership change is about to occur before any check has been conducted. Suppose that we are adding subgroup B to supergroup A, the locking protocol is specified as follows: 1. Acquire a lock for B and all its direct and indirect subgroups. 2. Acquire a lock for A. For the removal of user groups, we acquire a lock for the user group to be removed with all its direct and indirect subgroups. This is the special case A=B, which is still complaint with the protocol. **Error handling** We currently rely on Postgres' deadlock detection to abort transactions and show an error for the users. In the future, we might need some recovery mechanism or at least better error handling. **Notes** An important note is that we need to reuse the recursive CTE query that finds the direct and indirect subgroups when applying the lock on the rows. And the lock needs to be acquired the same way for the addition and removal of direct subgroups. User membership change (as opposed to user group membership) is not affected. Read-only queries aren't either. The locks only protect critical regions where the user group dependency graph might violate the DAG constraint, where users are not participating. **Testing** We implement a transaction test case targeting some typical scenarios when an internal server error is expected to happen (this means that the user group view makes the correct decision to abort the transaction when something goes wrong with locks). To achieve this, we add a development view intended only for unit tests. It has a global BARRIER that can be shared across threads, so that we can synchronize them to consistently reproduce certain potential race conditions prevented by the database locks. The transaction test case lanuches pairs of threads initiating possibly conflicting requests at the same time. The tests are set up such that exactly N of them are expected to succeed with a certain error message (while we don't know each one). **Security notes** get_recursive_subgroups_for_groups will no longer fetch user groups from other realms. As a result, trying to add/remove a subgroup from another realm results in a UserGroup not found error response. We also implement subgroup-specific checks in has_user_group_access to keep permission managing in a single place. Do note that the API currently don't have a way to violate that check because we are only checking the realm ID now.
2023-06-17 04:39:52 +02:00
def tearDown(self) -> None:
"""Verifies that the test did not adjust the set of rows in the test
database. This is a sanity check to help ensure that tests
using this class do not have unintended side effects on the
test database.
"""
super().tearDown()
for table_name, ids in get_row_ids_in_all_tables():
self.assertSetEqual(
self.models_ids_set[table_name],
ids,
f"{table_name} got a different set of ids after this test",
)
def _fixture_teardown(self) -> None:
"""We override the default _fixture_teardown method defined on
TransactionTestCase, so that the test database does not get
flushed/deleted after each test using this class.
"""
class WebhookTestCase(ZulipTestCase):
"""Shared test class for all incoming webhooks tests.
Used by configuring the below class attributes, and calling
send_and_test_message in individual tests.
* Tests can override build_webhook_url if the webhook requires a
different URL format.
* Tests can override get_body for cases where there is no
available fixture file.
* Tests should specify WEBHOOK_DIR_NAME to enforce that all event
types are declared in the @webhook_view decorator. This is
important for ensuring we document all fully supported event types.
"""
CHANNEL_NAME: Optional[str] = None
TEST_USER_EMAIL = "webhook-bot@zulip.com"
URL_TEMPLATE: str
WEBHOOK_DIR_NAME: Optional[str] = None
# This last parameter is a workaround to handle webhooks that do not
# name the main function api_{WEBHOOK_DIR_NAME}_webhook.
VIEW_FUNCTION_NAME: Optional[str] = None
@property
def test_user(self) -> UserProfile:
return self.get_user_from_email(self.TEST_USER_EMAIL, get_realm("zulip"))
@override
def setUp(self) -> None:
super().setUp()
self.url = self.build_webhook_url()
if self.WEBHOOK_DIR_NAME is not None:
# If VIEW_FUNCTION_NAME is explicitly specified and
# WEBHOOK_DIR_NAME is not None, an exception will be
# raised when a test triggers events that are not
# explicitly specified via the event_types parameter to
# the @webhook_view decorator.
if self.VIEW_FUNCTION_NAME is None:
function = import_string(
f"zerver.webhooks.{self.WEBHOOK_DIR_NAME}.view.api_{self.WEBHOOK_DIR_NAME}_webhook"
)
else:
function = import_string(
f"zerver.webhooks.{self.WEBHOOK_DIR_NAME}.view.{self.VIEW_FUNCTION_NAME}"
)
all_event_types = None
if hasattr(function, "_all_event_types"):
all_event_types = function._all_event_types
if all_event_types is None:
return # nocoverage
def side_effect(*args: Any, **kwargs: Any) -> None:
complete_event_type = (
kwargs.get("complete_event_type")
if len(args) < 5
else args[4] # complete_event_type is the argument at index 4
)
if (
complete_event_type is not None
and all_event_types is not None
and complete_event_type not in all_event_types
): # nocoverage
raise Exception(
f"""
Error: This test triggered a message using the event "{complete_event_type}", which was not properly
registered via the @webhook_view(..., event_types=[...]). These registrations are important for Zulip
self-documenting the supported event types for this integration.
You can fix this by adding "{complete_event_type}" to ALL_EVENT_TYPES for this webhook.
""".strip()
)
check_send_webhook_message(*args, **kwargs)
self.patch = mock.patch(
f"zerver.webhooks.{self.WEBHOOK_DIR_NAME}.view.check_send_webhook_message",
side_effect=side_effect,
)
self.patch.start()
self.addCleanup(self.patch.stop)
def api_channel_message(
self,
user: UserProfile,
fixture_name: str,
expected_topic: Optional[str] = None,
expected_message: Optional[str] = None,
content_type: Optional[str] = "application/json",
expect_noop: bool = False,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> None:
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["HTTP_AUTHORIZATION"] = self.encode_user(user)
self.check_webhook(
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
fixture_name,
expected_topic,
expected_message,
content_type,
expect_noop,
**extra,
)
def check_webhook(
self,
fixture_name: str,
expected_topic_name: Optional[str] = None,
expected_message: Optional[str] = None,
content_type: Optional[str] = "application/json",
expect_noop: bool = False,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> None:
"""
check_webhook is the main way to test "normal" webhooks that
work by receiving a payload from a third party and then writing
some message to a Zulip stream.
We use `fixture_name` to find the payload data in of our test
fixtures. Then we verify that a message gets sent to a stream:
self.CHANNEL_NAME: stream name
expected_topic_name: topic name
expected_message: content
We simulate the delivery of the payload with `content_type`,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
and you can pass other headers via `extra`.
For the rare cases of webhooks actually sending direct messages,
see send_and_test_private_message.
When no message is expected to be sent, set `expect_noop` to True.
"""
assert self.CHANNEL_NAME is not None
self.subscribe(self.test_user, self.CHANNEL_NAME)
payload = self.get_payload(fixture_name)
if content_type is not None:
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["content_type"] = content_type
if self.WEBHOOK_DIR_NAME is not None:
headers = get_fixture_http_headers(self.WEBHOOK_DIR_NAME, fixture_name)
headers = standardize_headers(headers)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra.update(headers)
try:
msg = self.send_webhook_payload(
self.test_user,
self.url,
payload,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra,
)
except EmptyResponseError:
if expect_noop:
return
else:
raise AssertionError(
"No message was sent. Pass expect_noop=True if this is intentional."
)
if expect_noop:
raise Exception(
"""
While no message is expected given expect_noop=True,
your test code triggered an endpoint that did write
one or more new messages.
""".strip()
)
assert expected_message is not None and expected_topic_name is not None
self.assert_channel_message(
message=msg,
channel_name=self.CHANNEL_NAME,
topic_name=expected_topic_name,
content=expected_message,
)
def assert_channel_message(
self,
message: Message,
channel_name: str,
topic_name: str,
content: str,
) -> None:
self.assert_message_stream_name(message, channel_name)
self.assertEqual(message.topic_name(), topic_name)
self.assertEqual(message.content, content)
def send_and_test_private_message(
self,
fixture_name: str,
expected_message: str,
content_type: str = "application/json",
*,
sender: Optional[UserProfile] = None,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra: str,
) -> Message:
"""
For the rare cases that you are testing a webhook that sends
direct messages, use this function.
Most webhooks send to streams, and you will want to look at
check_webhook.
"""
payload = self.get_payload(fixture_name)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra["content_type"] = content_type
if self.WEBHOOK_DIR_NAME is not None:
headers = get_fixture_http_headers(self.WEBHOOK_DIR_NAME, fixture_name)
headers = standardize_headers(headers)
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
extra.update(headers)
if sender is None:
sender = self.test_user
msg = self.send_webhook_payload(
sender,
self.url,
payload,
tests: Tighten signature of the wrapped test client helpers. We wrap methods of the django test client for the test suite, and type keyword variadic arguments as `ClientArg` as it might called with a mix of `bool` and `str`. This is problematic when we call the original methods on the test client as we attempt to unpack the dictionary of keyword arguments, which has no type guarantee that certain keys that the test client requires to be bool will certainly be bool. For example, you can call `self.client_post(url, info, follow="invalid")` without getting a mypy error while the django test client requires `follow: bool`. The unsafely typed keyword variadic arguments leads to error within the body the wrapped test client functions as we call `django_client.post` with `**kwargs` when django-stubs gets added, making it necessary to refactor these wrappers for type safety. The approach here minimizes the need to refactor callers, as we keep `kwargs` being variadic while change its type from `ClientArg` to `str` after defining all the possible `bool` arguments that might previously appear in `kwargs`. We also copy the defaults from the django test client as they are unlikely to change. The tornado test cases are also refactored due to the change of the signature of `set_http_headers` with the `skip_user_agent` being added as a keyword argument. We want to unconditionally set this flag to `True` because the `HTTP_USER_AGENT` is not supported. It also removes a unnecessary duplication of an argument. This is a part of the django-stubs refactorings. Signed-off-by: Zixuan James Li <p359101898@gmail.com>
2022-06-14 22:44:49 +02:00
**extra,
)
self.assertEqual(msg.content, expected_message)
return msg
def build_webhook_url(self, *args: str, **kwargs: str) -> str:
url = self.URL_TEMPLATE
if url.find("api_key") >= 0:
api_key = get_api_key(self.test_user)
url = self.URL_TEMPLATE.format(api_key=api_key, stream=self.CHANNEL_NAME)
else:
url = self.URL_TEMPLATE.format(stream=self.CHANNEL_NAME)
has_arguments = kwargs or args
if has_arguments and url.find("?") == -1:
url = f"{url}?" # nocoverage
else:
url = f"{url}&"
for key, value in kwargs.items():
url = f"{url}{key}={value}&"
for arg in args:
url = f"{url}{arg}&"
return url[:-1] if has_arguments else url
def get_payload(self, fixture_name: str) -> Union[str, Dict[str, str]]:
"""
Generally webhooks that override this should return dicts."""
return self.get_body(fixture_name)
def get_body(self, fixture_name: str) -> str:
assert self.WEBHOOK_DIR_NAME is not None
body = self.webhook_fixture_data(self.WEBHOOK_DIR_NAME, fixture_name)
# fail fast if we don't have valid json
orjson.loads(body)
return body
class MigrationsTestCase(ZulipTransactionTestCase): # nocoverage
"""
Test class for database migrations inspired by this blog post:
https://www.caktusgroup.com/blog/2016/02/02/writing-unit-tests-django-migrations/
Documented at https://zulip.readthedocs.io/en/latest/subsystems/schema-migrations.html
"""
@property
def app(self) -> str:
app_config = apps.get_containing_app_config(type(self).__module__)
assert app_config is not None
return app_config.name
python: Convert assignment type annotations to Python 3.6 style. This commit was split by tabbott; this piece covers the vast majority of files in Zulip, but excludes scripts/, tools/, and puppet/ to help ensure we at least show the right error messages for Xenial systems. We can likely further refine the remaining pieces with some testing. Generated by com2ann, with whitespace fixes and various manual fixes for runtime issues: - invoiced_through: Optional[LicenseLedger] = models.ForeignKey( + invoiced_through: Optional["LicenseLedger"] = models.ForeignKey( -_apns_client: Optional[APNsClient] = None +_apns_client: Optional["APNsClient"] = None - notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - signup_notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + signup_notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - author: Optional[UserProfile] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) + author: Optional["UserProfile"] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) - bot_owner: Optional[UserProfile] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) + bot_owner: Optional["UserProfile"] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) - default_sending_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) - default_events_register_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_sending_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_events_register_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) -descriptors_by_handler_id: Dict[int, ClientDescriptor] = {} +descriptors_by_handler_id: Dict[int, "ClientDescriptor"] = {} -worker_classes: Dict[str, Type[QueueProcessingWorker]] = {} -queues: Dict[str, Dict[str, Type[QueueProcessingWorker]]] = {} +worker_classes: Dict[str, Type["QueueProcessingWorker"]] = {} +queues: Dict[str, Dict[str, Type["QueueProcessingWorker"]]] = {} -AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional[LDAPSearch] = None +AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional["LDAPSearch"] = None Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
2020-04-22 01:09:50 +02:00
migrate_from: Optional[str] = None
migrate_to: Optional[str] = None
@override
def setUp(self) -> None:
super().setUp()
assert (
self.migrate_from and self.migrate_to
), f"TestCase '{type(self).__name__}' must define migrate_from and migrate_to properties"
python: Convert assignment type annotations to Python 3.6 style. This commit was split by tabbott; this piece covers the vast majority of files in Zulip, but excludes scripts/, tools/, and puppet/ to help ensure we at least show the right error messages for Xenial systems. We can likely further refine the remaining pieces with some testing. Generated by com2ann, with whitespace fixes and various manual fixes for runtime issues: - invoiced_through: Optional[LicenseLedger] = models.ForeignKey( + invoiced_through: Optional["LicenseLedger"] = models.ForeignKey( -_apns_client: Optional[APNsClient] = None +_apns_client: Optional["APNsClient"] = None - notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - signup_notifications_stream: Optional[Stream] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) + signup_notifications_stream: Optional["Stream"] = models.ForeignKey('Stream', related_name='+', null=True, blank=True, on_delete=CASCADE) - author: Optional[UserProfile] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) + author: Optional["UserProfile"] = models.ForeignKey('UserProfile', blank=True, null=True, on_delete=CASCADE) - bot_owner: Optional[UserProfile] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) + bot_owner: Optional["UserProfile"] = models.ForeignKey('self', null=True, on_delete=models.SET_NULL) - default_sending_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) - default_events_register_stream: Optional[Stream] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_sending_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) + default_events_register_stream: Optional["Stream"] = models.ForeignKey('zerver.Stream', null=True, related_name='+', on_delete=CASCADE) -descriptors_by_handler_id: Dict[int, ClientDescriptor] = {} +descriptors_by_handler_id: Dict[int, "ClientDescriptor"] = {} -worker_classes: Dict[str, Type[QueueProcessingWorker]] = {} -queues: Dict[str, Dict[str, Type[QueueProcessingWorker]]] = {} +worker_classes: Dict[str, Type["QueueProcessingWorker"]] = {} +queues: Dict[str, Dict[str, Type["QueueProcessingWorker"]]] = {} -AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional[LDAPSearch] = None +AUTH_LDAP_REVERSE_EMAIL_SEARCH: Optional["LDAPSearch"] = None Signed-off-by: Anders Kaseorg <anders@zulipchat.com>
2020-04-22 01:09:50 +02:00
migrate_from: List[Tuple[str, str]] = [(self.app, self.migrate_from)]
migrate_to: List[Tuple[str, str]] = [(self.app, self.migrate_to)]
executor = MigrationExecutor(connection)
old_apps = executor.loader.project_state(migrate_from).apps
# Reverse to the original migration
executor.migrate(migrate_from)
self.setUpBeforeMigration(old_apps)
# Run the migration to test
executor = MigrationExecutor(connection)
executor.loader.build_graph() # reload.
executor.migrate(migrate_to)
self.apps = executor.loader.project_state(migrate_to).apps
def setUpBeforeMigration(self, apps: StateApps) -> None:
pass # nocoverage
def get_topic_messages(user_profile: UserProfile, stream: Stream, topic_name: str) -> List[Message]:
query = UserMessage.objects.filter(
user_profile=user_profile,
message__recipient=stream.recipient,
).order_by("id")
return [um.message for um in filter_by_topic_name_via_message(query, topic_name)]
@skipUnless(settings.ZILENCER_ENABLED, "requires zilencer")
class BouncerTestCase(ZulipTestCase):
@override
def setUp(self) -> None:
# Set a deterministic uuid and a nice hostname for convenience.
self.server_uuid = "6cde5f7a-1f7e-4978-9716-49f69ebfc9fe"
self.server = RemoteZulipServer.objects.all().latest("id")
self.server.uuid = self.server_uuid
self.server.hostname = "demo.example.com"
self.server.save()
super().setUp()
@override
def tearDown(self) -> None:
RemoteZulipServer.objects.filter(uuid=self.server_uuid).delete()
super().tearDown()
def request_callback(self, request: PreparedRequest) -> Tuple[int, ResponseHeaders, bytes]:
kwargs = {}
if isinstance(request.body, bytes):
# send_json_to_push_bouncer sends the body as bytes containing json.
data = orjson.loads(request.body)
kwargs = dict(content_type="application/json")
else:
assert isinstance(request.body, str) or request.body is None
params: Dict[str, List[str]] = parse_qs(request.body)
# In Python 3, the values of the dict from `parse_qs` are
# in a list, because there might be multiple values.
# But since we are sending values with no same keys, hence
# we can safely pick the first value.
data = {k: v[0] for k, v in params.items()}
assert request.url is not None # allow mypy to infer url is present.
assert settings.PUSH_NOTIFICATION_BOUNCER_URL is not None
local_url = request.url.replace(settings.PUSH_NOTIFICATION_BOUNCER_URL, "")
if request.method == "POST":
result = self.uuid_post(self.server_uuid, local_url, data, subdomain="", **kwargs)
elif request.method == "GET":
result = self.uuid_get(self.server_uuid, local_url, data, subdomain="")
return (result.status_code, result.headers, result.content)
def add_mock_response(self) -> None:
# Match any endpoint with the PUSH_NOTIFICATION_BOUNCER_URL.
assert settings.PUSH_NOTIFICATION_BOUNCER_URL is not None
COMPILED_URL = re.compile(settings.PUSH_NOTIFICATION_BOUNCER_URL + r".*")
responses.add_callback(responses.POST, COMPILED_URL, callback=self.request_callback)
responses.add_callback(responses.GET, COMPILED_URL, callback=self.request_callback)
def get_generic_payload(self, method: str = "register") -> Dict[str, Any]:
user_id = 10
token = "111222"
token_kind = PushDeviceToken.GCM
return {"user_id": user_id, "token": token, "token_kind": token_kind}