fix: Move tests under ldclient namespace (#29)

Previous distributions of this package included two packages -- ldclient
and testing. This top level testing namespace can conflict with other
packages. In fact, it conflicts with our own eventsource library.

In general this doesn't matter, but it may if:

1. You are using a build process that warns about conflicts (see [this
   issue][1])
2. You want to install the sdist on an unsupported platform and would
   like to be able to verify the tests.

To resolve this issue, we are moving the testing folder into the
ldclient package. These testing files will only be included in the sdist
format. This allows for a smaller wheel size while also allowing for
flexibility with consumers.

[1]: #281

Makefile

@@ -33,7 +33,7 @@ test: install
 .PHONY: lint
 lint: #! Run type analysis and linting checks
 lint: install
- @poetry run mypy ldclient testing
+ @poetry run mypy ldclient
 
 #
 # Documentation generation

testing/feature_store_test_base.py → ldclient/testing/feature_store_test_base.py

@@ -1,7 +1,7 @@
 from ldclient.interfaces import FeatureStore
 from ldclient.versioned_data_kind import FEATURES
 
-from testing.builders import *
+from ldclient.testing.builders import *
 
 from abc import abstractmethod
 import pytest
@@ -21,11 +21,11 @@ def create_feature_store(self) -> FeatureStore:
 class StoreTestScope:
  def __init__(self, store: FeatureStore):
  self.__store = store
- 
+
  @property
  def store(self) -> FeatureStore:
  return self.__store
- 
+
  # These magic methods allow the scope to be automatically cleaned up in a "with" block
  def __enter__(self):
  return self.__store
@@ -59,7 +59,7 @@ def inited_store(self, tester):
  }
  })
  return scope
- 
+
  @staticmethod
  def make_feature(key, ver):
  return FlagBuilder(key).version(ver).on(True).variations(True, False).salt('abc').build()

testing/http_util.py → ldclient/testing/http_util.py

@@ -48,7 +48,7 @@ def __init__(self, port, secure):
  self.server = HTTPServer(('localhost', port), MockServerRequestHandler)
  if secure:
  context = SSLContext(PROTOCOL_TLSv1_2)
- context.load_cert_chain('./testing/selfsigned.pem', './testing/selfsigned.key')
+ context.load_cert_chain('./ldclient/testing/selfsigned.pem', './ldclient/testing/selfsigned.key')
  self.server.socket = context.wrap_socket(
  self.server.socket,
  server_side=True

testing/impl/datasource/test_feature_requester.py → ldclient/testing/impl/datasource/test_feature_requester.py

@@ -2,8 +2,8 @@
 from ldclient.impl.datasource.feature_requester import FeatureRequesterImpl
 from ldclient.version import VERSION
 from ldclient.versioned_data_kind import FEATURES, SEGMENTS
-from testing.http_util import start_server, BasicResponse, JsonResponse
-from testing.proxy_test_util import do_proxy_tests
+from ldclient.testing.http_util import start_server, BasicResponse, JsonResponse
+from ldclient.testing.proxy_test_util import do_proxy_tests
 
 def test_get_all_data_returns_data():
  with start_server() as server:

testing/impl/datasource/test_polling_processor.py → ldclient/testing/impl/datasource/test_polling_processor.py

@@ -11,9 +11,9 @@
 from ldclient.interfaces import DataSourceStatus, DataSourceState, DataSourceErrorKind
 from ldclient.versioned_data_kind import FEATURES, SEGMENTS
 
-from testing.builders import *
-from testing.stub_util import MockFeatureRequester, MockResponse
-from testing.test_util import SpyListener
+from ldclient.testing.builders import *
+from ldclient.testing.stub_util import MockFeatureRequester, MockResponse
+from ldclient.testing.test_util import SpyListener
 
 pp = None
 mock_requester = None

testing/impl/datasource/test_streaming.py → ldclient/testing/impl/datasource/test_streaming.py

@@ -13,11 +13,11 @@
 from ldclient.interfaces import DataSourceStatus, DataSourceState, DataSourceErrorKind
 from ldclient.impl.datasource.status import DataSourceUpdateSinkImpl
 
-from testing.builders import *
-from testing.http_util import start_server, BasicResponse, CauseNetworkError, SequentialHandler
-from testing.proxy_test_util import do_proxy_tests
-from testing.stub_util import make_delete_event, make_patch_event, make_put_event, make_invalid_put_event, stream_content
-from testing.test_util import SpyListener
+from ldclient.testing.builders import *
+from ldclient.testing.http_util import start_server, BasicResponse, CauseNetworkError, SequentialHandler
+from ldclient.testing.proxy_test_util import do_proxy_tests
+from ldclient.testing.stub_util import make_delete_event, make_patch_event, make_put_event, make_invalid_put_event, stream_content
+from ldclient.testing.test_util import SpyListener
 
 brief_delay = 0.001
 

testing/impl/evaluator_util.py → ldclient/testing/impl/evaluator_util.py

@@ -3,7 +3,7 @@
 from ldclient.impl.evaluator import Evaluator, _make_big_segment_ref
 from ldclient.impl.events.types import EventFactory
 from ldclient.impl.model import *
-from testing.builders import *
+from ldclient.testing.builders import *
 
 from typing import Any, Optional, Tuple, Union
 

testing/impl/events/test_event_factory.py → ldclient/testing/impl/events/test_event_factory.py

@@ -2,7 +2,7 @@
 from ldclient.evaluation import EvaluationDetail
 from ldclient.impl.events.types import EventFactory
 
-from testing.builders import *
+from ldclient.testing.builders import *
 
 _event_factory_default = EventFactory(False)
 _user = Context.create('x')

testing/impl/events/test_event_processor.py → ldclient/testing/impl/events/test_event_processor.py

@@ -17,9 +17,9 @@
 from ldclient.impl.util import timedelta_millis
 from ldclient.impl.events.event_context_formatter import EventContextFormatter
 
-from testing.builders import *
-from testing.proxy_test_util import do_proxy_tests
-from testing.stub_util import MockHttp
+from ldclient.testing.builders import *
+from ldclient.testing.proxy_test_util import do_proxy_tests
+from ldclient.testing.stub_util import MockHttp
 
 
 default_config = Config("fake_sdk_key")

testing/impl/events/test_event_summarizer.py → ldclient/testing/impl/events/test_event_summarizer.py

@@ -2,7 +2,7 @@
 from ldclient.impl.events.event_summarizer import EventSummarizer, EventSummaryCounter, EventSummaryFlag
 from ldclient.impl.events.types import *
 
-from testing.builders import *
+from ldclient.testing.builders import *
 
 
 user = Context.create('user1')

testing/impl/test_big_segments.py → ldclient/testing/impl/test_big_segments.py

@@ -2,7 +2,7 @@
 from ldclient.evaluation import BigSegmentsStatus
 from ldclient.impl.big_segments import BigSegmentStoreManager, _hash_for_user_key
 from ldclient.interfaces import BigSegmentStoreMetadata
-from testing.mock_components import MockBigSegmentStore
+from ldclient.testing.mock_components import MockBigSegmentStore
 
 from queue import Queue
 import time
@@ -108,20 +108,20 @@ def test_membership_query_least_recent_context_evicted_from_cache():
  result3 = manager.get_user_membership(user_key_3)
 
  assert store.membership_queries == [user_hash_1, user_hash_2, user_hash_3]
- 
+
  # Since the capacity is only 2 and user_key_1 was the least recently used, that key should be
  # evicted by the user_key_3 query. Now only user_key_2 and user_key_3 are in the cache, and
  # querying them again should not cause a new query to the store.
  result2a = manager.get_user_membership(user_key_2)
  result3a = manager.get_user_membership(user_key_3)
  assert result2a == result2
  assert result3a == result3
- 
+
  assert store.membership_queries == [user_hash_1, user_hash_2, user_hash_3]
- 
+
  result1a = manager.get_user_membership(user_key_1)
  assert result1a == result1
- 
+
  assert store.membership_queries == [user_hash_1, user_hash_2, user_hash_3, user_hash_1]
  finally:
  manager.stop()
@@ -130,7 +130,7 @@ def test_status_polling_detects_store_unavailability():
  store = MockBigSegmentStore()
  store.setup_metadata_always_up_to_date()
  statuses = Queue()
- 
+
  manager = BigSegmentStoreManager(BigSegmentsConfig(store=store, status_poll_interval=0.01))
 
  try:
@@ -155,7 +155,7 @@ def test_status_polling_detects_stale_status():
  store = MockBigSegmentStore()
  store.setup_metadata_always_up_to_date()
  statuses = Queue()
- 
+
  manager = BigSegmentStoreManager(BigSegmentsConfig(store=store, status_poll_interval=0.01))
 
  try:

testing/impl/test_data_sink.py → ldclient/testing/impl/test_data_sink.py

@@ -9,8 +9,8 @@
 from ldclient.impl.listeners import Listeners
 from ldclient.versioned_data_kind import FEATURES, SEGMENTS
 
-from testing.test_util import SpyListener
-from testing.builders import FlagBuilder, FlagRuleBuilder, make_clause, SegmentBuilder, SegmentRuleBuilder
+from ldclient.testing.test_util import SpyListener
+from ldclient.testing.builders import FlagBuilder, FlagRuleBuilder, make_clause, SegmentBuilder, SegmentRuleBuilder
 
 
 @pytest.fixture

testing/impl/test_evaluator.py → ldclient/testing/impl/test_evaluator.py

@@ -1,7 +1,7 @@
 from ldclient.client import Context
 from ldclient.evaluation import EvaluationDetail
-from testing.builders import *
-from testing.impl.evaluator_util import *
+from ldclient.testing.builders import *
+from ldclient.testing.impl.evaluator_util import *
 
 
 def test_flag_returns_off_variation_if_flag_is_off():
@@ -100,5 +100,5 @@ def test_segment_match_clause_falls_through_with_no_errors_if_segment_not_found(
  user = Context.create('foo')
  flag = make_boolean_flag_with_clauses(make_clause_matching_segment_key('segkey'))
  evaluator = EvaluatorBuilder().with_unknown_segment('segkey').build()
- 
+
  assert evaluator.evaluate(flag, user, event_factory).detail.value == False

testing/impl/test_evaluator_big_segment.py → ldclient/testing/impl/test_evaluator_big_segment.py

@@ -1,8 +1,8 @@
 import pytest
 
 from ldclient.evaluation import BigSegmentsStatus
-from testing.builders import *
-from testing.impl.evaluator_util import *
+from ldclient.testing.builders import *
+from ldclient.testing.impl.evaluator_util import *
 
 
 def test_big_segment_with_no_generation_is_not_matched():
@@ -30,7 +30,7 @@ def _test_matched_with_include(non_default_kind: bool, multi_kind_context: bool)
  single_kind_context = Context.create(target_key, 'kind1') if non_default_kind else Context.create(target_key)
  eval_context = Context.create_multi(single_kind_context, Context.create('key2', 'kind2')) if multi_kind_context \
  else single_kind_context
- 
+
  segment = SegmentBuilder('key').version(1) \
  .unbounded(True) \
  .unbounded_context_kind('kind1' if non_default_kind else None) \

testing/impl/test_evaluator_bucketing.py → ldclient/testing/impl/test_evaluator_bucketing.py

@@ -2,8 +2,8 @@
 from ldclient.impl.evaluator import _bucket_context, _variation_index_for_context
 from ldclient.impl.model import *
 
-from testing.builders import *
-from testing.impl.evaluator_util import *
+from ldclient.testing.builders import *
+from ldclient.testing.impl.evaluator_util import *
 
 import math
 import pytest
@@ -22,7 +22,7 @@ def test_variation_index_is_returned_for_bucket(self):
  # so we can construct a rollout whose second bucket just barely contains that value
  bucket_value = math.trunc(_bucket_context(None, user, None, flag.key, flag.salt, None) * 100000)
  assert bucket_value > 0 and bucket_value < 100000
- 
+
  bad_variation_a = 0
  matched_variation = 1
  bad_variation_b = 2
@@ -44,7 +44,7 @@ def test_last_bucket_is_used_if_bucket_value_equals_total_weight(self):
 
  # We'll construct a list of variations that stops right at the target bucket value
  bucket_value = math.trunc(_bucket_context(None, user, None, flag.key, flag.salt, None) * 100000)
- 
+
  rule = VariationOrRollout({
  'rollout': {
  'variations': [
@@ -54,7 +54,7 @@ def test_last_bucket_is_used_if_bucket_value_equals_total_weight(self):
  })
  result_variation = _variation_index_for_context(flag, rule, user)
  assert result_variation == (0, False)
- 
+
  def test_bucket_by_user_key(self):
  user = Context.create('userKeyA')
  bucket = _bucket_context(None, user, None, 'hashKey', 'saltyA', None)

testing/impl/test_evaluator_clause.py → ldclient/testing/impl/test_evaluator_clause.py

@@ -1,6 +1,6 @@
 from ldclient.client import Context
-from testing.builders import *
-from testing.impl.evaluator_util import *
+from ldclient.testing.builders import *
+from ldclient.testing.impl.evaluator_util import *
 
 
 def assert_match_clause(clause: dict, context: Context, should_match: bool):

testing/impl/test_evaluator_prerequisites.py → ldclient/testing/impl/test_evaluator_prerequisites.py

@@ -4,8 +4,8 @@
 from ldclient.evaluation import EvaluationDetail
 from ldclient.impl.events.types import EventInputEvaluation
 
-from testing.builders import *
-from testing.impl.evaluator_util import *
+from ldclient.testing.builders import *
+from ldclient.testing.impl.evaluator_util import *
 
 
 def test_flag_returns_off_variation_if_prerequisite_not_found():

testing/impl/test_evaluator_segment.py → ldclient/testing/impl/test_evaluator_segment.py

@@ -2,8 +2,8 @@
 
 from ldclient import Context
 from ldclient.impl.evaluator import _bucket_context
-from testing.builders import *
-from testing.impl.evaluator_util import *
+from ldclient.testing.builders import *
+from ldclient.testing.impl.evaluator_util import *
 
 
 def _segment_matches_context(segment: Segment, context: Context) -> bool:

testing/impl/test_evaluator_target.py → ldclient/testing/impl/test_evaluator_target.py

@@ -1,6 +1,6 @@
 from ldclient.client import Context
-from testing.builders import *
-from testing.impl.evaluator_util import *
+from ldclient.testing.builders import *
+from ldclient.testing.impl.evaluator_util import *
 
 
 FALLTHROUGH_VAR = 0
@@ -34,7 +34,7 @@ def test_user_targets_only(self):
  .target(MATCH_VAR_1, 'c') \
  .target(MATCH_VAR_2, 'b', 'a') \
  .build()
- 
+
  expect_match(flag, Context.create('a'), MATCH_VAR_2)
  expect_match(flag, Context.create('b'), MATCH_VAR_2)
  expect_match(flag, Context.create('c'), MATCH_VAR_1)
@@ -61,7 +61,7 @@ def test_user_targets_and_context_targets(self):
  .context_target(Context.DEFAULT_KIND, MATCH_VAR_1) \
  .context_target(Context.DEFAULT_KIND, MATCH_VAR_2) \
  .build()
- 
+
  expect_match(flag, Context.create('a'), MATCH_VAR_2)
  expect_match(flag, Context.create('b'), MATCH_VAR_2)
  expect_match(flag, Context.create('c'), MATCH_VAR_1)

testing/impl/test_flag_tracker.py → ldclient/testing/impl/test_flag_tracker.py

@@ -1,5 +1,5 @@
 from ldclient.impl.flag_tracker import FlagTrackerImpl
-from testing.test_util import SpyListener
+from ldclient.testing.test_util import SpyListener
 from ldclient.impl.listeners import Listeners
 from ldclient.interfaces import FlagChange
 

testing/impl/test_model_decode.py → ldclient/testing/impl/test_model_decode.py

@@ -4,7 +4,7 @@
 
 from ldclient.impl.model import *
 
-from testing.builders import *
+from ldclient.testing.builders import *
 
 
 def test_flag_targets_are_stored_as_sets():

testing/impl/test_operators.py → ldclient/testing/impl/test_operators.py

@@ -2,7 +2,7 @@
 
 from ldclient.impl import operators
 
-from testing.builders import *
+from ldclient.testing.builders import *
 
 
 @pytest.mark.parametrize("op,context_value,clause_value,expected", [

testing/integrations/persistent_feature_store_test_base.py → ldclient/testing/integrations/persistent_feature_store_test_base.py

@@ -5,12 +5,12 @@
 from ldclient.interfaces import FeatureStore
 from ldclient.versioned_data_kind import FEATURES
 
-from testing.feature_store_test_base import FeatureStoreTestBase, FeatureStoreTester, StoreTestScope
-from testing.test_util import skip_database_tests
+from ldclient.testing.feature_store_test_base import FeatureStoreTestBase, FeatureStoreTester, StoreTestScope
+from ldclient.testing.test_util import skip_database_tests
 
 
 # The standard test suite to be run against all persistent feature store implementations. See
-# testing.feature_store_test_base for the basic model being used here. For each database integration,
+# ldclient.testing.feature_store_test_base for the basic model being used here. For each database integration,
 # we must define a subclass of PersistentFeatureStoreTester which overrides its abstract methods as
 # appropriate for that database, and then define a subclass of PersistentFeatureStoreTestBase which
 # simply specifies what tester subclass to use.
@@ -29,7 +29,7 @@ def create_persistent_feature_store(self, prefix: str, caching: CacheConfig) ->
  :param caching: caching parameters for the store constructor
  """
  pass
- 
+
  @abstractmethod
  def clear_data(self, prefix: str):
  """
@@ -74,12 +74,12 @@ def test_stores_with_different_prefixes_are_independent(self):
  tester_b = self.tester_class()
  tester_b.prefix = "b"
  tester_b.clear_data(tester_b.prefix)
- 
+
  flag_a1 = { 'key': 'flagA1', 'version': 1 }
  flag_a2 = { 'key': 'flagA2', 'version': 1 }
  flag_b1 = { 'key': 'flagB1', 'version': 1 }
  flag_b2 = { 'key': 'flagB2', 'version': 1 }
- 
+
  with StoreTestScope(tester_a.create_feature_store()) as store_a:
  with StoreTestScope(tester_b.create_feature_store()) as store_b:
  store_a.init({ FEATURES: { 'flagA1': flag_a1 } })