crawlee 1.0.1b9__py3-none-any.whl → 1.3.1b3__py3-none-any.whl

crawlee/statistics/_models.py

@@ -1,9 +1,10 @@
 from __future__ import annotations
 
 import json
+import warnings
 from dataclasses import asdict, dataclass
 from datetime import datetime, timedelta, timezone
-from typing import Annotated, Any
+from typing import TYPE_CHECKING, Annotated, Any
 
 from pydantic import BaseModel, ConfigDict, Field, PlainSerializer, PlainValidator, computed_field
 from typing_extensions import override
@@ -76,10 +77,20 @@ class StatisticsState(BaseModel):
     crawler_started_at: Annotated[datetime | None, Field(alias='crawlerStartedAt')] = None
     crawler_last_started_at: Annotated[datetime | None, Field(alias='crawlerLastStartTimestamp')] = None
     crawler_finished_at: Annotated[datetime | None, Field(alias='crawlerFinishedAt')] = None
-    crawler_runtime: Annotated[timedelta_ms, Field(alias='crawlerRuntimeMillis')] = timedelta()
-    errors: dict[str, Any] = Field(default_factory=dict)
-    retry_errors: dict[str, Any] = Field(alias='retryErrors', default_factory=dict)
-    requests_with_status_code: dict[str, int] = Field(alias='requestsWithStatusCode', default_factory=dict)
+
+    # Workaround for Pydantic and type checkers when using Annotated with default_factory
+    if TYPE_CHECKING:
+        errors: dict[str, Any] = {}
+        retry_errors: dict[str, Any] = {}
+        requests_with_status_code: dict[str, int] = {}
+    else:
+        errors: Annotated[dict[str, Any], Field(default_factory=dict)]
+        retry_errors: Annotated[dict[str, Any], Field(alias='retryErrors', default_factory=dict)]
+        requests_with_status_code: Annotated[
+            dict[str, int],
+            Field(alias='requestsWithStatusCode', default_factory=dict),
+        ]
+
     stats_persisted_at: Annotated[
         datetime | None, Field(alias='statsPersistedAt'), PlainSerializer(lambda _: datetime.now(timezone.utc))
     ] = None
@@ -93,22 +104,53 @@ class StatisticsState(BaseModel):
         ),
     ] = {}
 
-    @computed_field(alias='requestTotalDurationMillis', return_type=timedelta_ms)  # type: ignore[prop-decorator]
+    # Used to track the crawler runtime, that had already been persisted. This is the runtime from previous runs.
+    _runtime_offset: Annotated[timedelta, Field(exclude=True)] = timedelta()
+
+    def model_post_init(self, /, __context: Any) -> None:
+        self._runtime_offset = self.crawler_runtime or self._runtime_offset
+
+    @property
+    def crawler_runtime(self) -> timedelta:
+        if self.crawler_last_started_at:
+            finished_at = self.crawler_finished_at or datetime.now(timezone.utc)
+            return self._runtime_offset + finished_at - self.crawler_last_started_at
+        return self._runtime_offset
+
+    @crawler_runtime.setter
+    def crawler_runtime(self, value: timedelta) -> None:
+        # Setter for backwards compatibility only, the crawler_runtime is now computed_field, and cant be set manually.
+        # To be removed in v2 release https://github.com/apify/crawlee-python/issues/1567
+        warnings.warn(
+            f"Setting 'crawler_runtime' is deprecated and will be removed in a future version."
+            f' Value {value} will not be used.',
+            DeprecationWarning,
+            stacklevel=2,
+        )
+
+    @computed_field(alias='crawlerRuntimeMillis')
+    def crawler_runtime_for_serialization(self) -> timedelta:
+        if self.crawler_last_started_at:
+            finished_at = self.crawler_finished_at or datetime.now(timezone.utc)
+            return self._runtime_offset + finished_at - self.crawler_last_started_at
+        return self._runtime_offset
+
+    @computed_field(alias='requestTotalDurationMillis', return_type=timedelta_ms)
     @property
     def request_total_duration(self) -> timedelta:
         return self.request_total_finished_duration + self.request_total_failed_duration
 
-    @computed_field(alias='requestAvgFailedDurationMillis', return_type=timedelta_ms | None)  # type: ignore[prop-decorator]
+    @computed_field(alias='requestAvgFailedDurationMillis', return_type=timedelta_ms | None)
     @property
     def request_avg_failed_duration(self) -> timedelta | None:
         return (self.request_total_failed_duration / self.requests_failed) if self.requests_failed else None
 
-    @computed_field(alias='requestAvgFinishedDurationMillis', return_type=timedelta_ms | None)  # type: ignore[prop-decorator]
+    @computed_field(alias='requestAvgFinishedDurationMillis', return_type=timedelta_ms | None)
     @property
     def request_avg_finished_duration(self) -> timedelta | None:
         return (self.request_total_finished_duration / self.requests_finished) if self.requests_finished else None
 
-    @computed_field(alias='requestsTotal')  # type: ignore[prop-decorator]
+    @computed_field(alias='requestsTotal')
     @property
     def requests_total(self) -> int:
         return self.requests_failed + self.requests_finished

crawlee/statistics/_statistics.py

@@ -1,6 +1,7 @@
 # Inspiration: https://github.com/apify/crawlee/blob/v3.9.2/packages/core/src/crawlers/statistics.ts
 from __future__ import annotations
 
+import asyncio
 import math
 import time
 from datetime import datetime, timedelta, timezone
@@ -17,8 +18,11 @@ from crawlee.statistics import FinalStatistics, StatisticsState
 from crawlee.statistics._error_tracker import ErrorTracker
 
 if TYPE_CHECKING:
+    from collections.abc import Callable, Coroutine
     from types import TracebackType
 
+    from crawlee.storages import KeyValueStore
+
 TStatisticsState = TypeVar('TStatisticsState', bound=StatisticsState, default=StatisticsState)
 TNewStatisticsState = TypeVar('TNewStatisticsState', bound=StatisticsState, default=StatisticsState)
 logger = getLogger(__name__)
@@ -70,6 +74,7 @@ class Statistics(Generic[TStatisticsState]):
         persistence_enabled: bool | Literal['explicit_only'] = False,
         persist_state_kvs_name: str | None = None,
         persist_state_key: str | None = None,
+        persist_state_kvs_factory: Callable[[], Coroutine[None, None, KeyValueStore]] | None = None,
         log_message: str = 'Statistics',
         periodic_message_logger: Logger | None = None,
         log_interval: timedelta = timedelta(minutes=1),
@@ -80,8 +85,6 @@ class Statistics(Generic[TStatisticsState]):
         self._id = Statistics.__next_id
         Statistics.__next_id += 1
 
-        self._instance_start: datetime | None = None
-
         self.error_tracker = ErrorTracker(
             save_error_snapshots=save_error_snapshots,
             snapshot_kvs_name=persist_state_kvs_name,
@@ -92,9 +95,10 @@ class Statistics(Generic[TStatisticsState]):
 
         self._state = RecoverableState(
             default_state=state_model(stats_id=self._id),
-            persist_state_key=persist_state_key or f'SDK_CRAWLER_STATISTICS_{self._id}',
+            persist_state_key=persist_state_key or f'__CRAWLER_STATISTICS_{self._id}',
             persistence_enabled=persistence_enabled,
             persist_state_kvs_name=persist_state_kvs_name,
+            persist_state_kvs_factory=persist_state_kvs_factory,
             logger=logger,
         )
 
@@ -110,8 +114,8 @@ class Statistics(Generic[TStatisticsState]):
         """Create near copy of the `Statistics` with replaced `state_model`."""
         new_statistics: Statistics[TNewStatisticsState] = Statistics(
             persistence_enabled=self._state._persistence_enabled,  # noqa: SLF001
-            persist_state_kvs_name=self._state._persist_state_kvs_name,  # noqa: SLF001
             persist_state_key=self._state._persist_state_key,  # noqa: SLF001
+            persist_state_kvs_factory=self._state._persist_state_kvs_factory,  # noqa: SLF001
             log_message=self._log_message,
             periodic_message_logger=self._periodic_message_logger,
             state_model=state_model,
@@ -125,6 +129,7 @@ class Statistics(Generic[TStatisticsState]):
         persistence_enabled: bool = False,
         persist_state_kvs_name: str | None = None,
         persist_state_key: str | None = None,
+        persist_state_kvs_factory: Callable[[], Coroutine[None, None, KeyValueStore]] | None = None,
         log_message: str = 'Statistics',
         periodic_message_logger: Logger | None = None,
         log_interval: timedelta = timedelta(minutes=1),
@@ -136,6 +141,7 @@ class Statistics(Generic[TStatisticsState]):
             persistence_enabled=persistence_enabled,
             persist_state_kvs_name=persist_state_kvs_name,
             persist_state_key=persist_state_key,
+            persist_state_kvs_factory=persist_state_kvs_factory,
             log_message=log_message,
             periodic_message_logger=periodic_message_logger,
             log_interval=log_interval,
@@ -158,14 +164,17 @@ class Statistics(Generic[TStatisticsState]):
         if self._active:
             raise RuntimeError(f'The {self.__class__.__name__} is already active.')
 
-        self._active = True
-        self._instance_start = datetime.now(timezone.utc)
-
         await self._state.initialize()
-        self._after_initialize()
+        # Reset `crawler_finished_at` to indicate a new run in progress.
+        self.state.crawler_finished_at = None
 
+        # Start periodic logging and let it print initial state before activation.
         self._periodic_logger.start()
+        await asyncio.sleep(0.01)
+        self._active = True
 
+        self.state.crawler_last_started_at = datetime.now(timezone.utc)
+        self.state.crawler_started_at = self.state.crawler_started_at or self.state.crawler_last_started_at
         return self
 
     async def __aexit__(
@@ -182,13 +191,14 @@ class Statistics(Generic[TStatisticsState]):
         if not self._active:
             raise RuntimeError(f'The {self.__class__.__name__} is not active.')
 
-        self._state.current_value.crawler_finished_at = datetime.now(timezone.utc)
-
-        await self._state.teardown()
+        if not self.state.crawler_last_started_at:
+            raise RuntimeError('Statistics.state.crawler_last_started_at not set.')
 
+        # Stop logging and deactivate the statistics to prevent further changes to crawler_runtime
         await self._periodic_logger.stop()
-
+        self.state.crawler_finished_at = datetime.now(timezone.utc)
         self._active = False
+        await self._state.teardown()
 
     @property
     def state(self) -> TStatisticsState:
@@ -247,11 +257,7 @@ class Statistics(Generic[TStatisticsState]):
 
     def calculate(self) -> FinalStatistics:
         """Calculate the current statistics."""
-        if self._instance_start is None:
-            raise RuntimeError('The Statistics object is not initialized')
-
-        crawler_runtime = datetime.now(timezone.utc) - self._instance_start
-        total_minutes = crawler_runtime.total_seconds() / 60
+        total_minutes = self.state.crawler_runtime.total_seconds() / 60
         state = self._state.current_value
         serialized_state = state.model_dump(by_alias=False)
 
@@ -262,7 +268,7 @@ class Statistics(Generic[TStatisticsState]):
             requests_failed_per_minute=math.floor(state.requests_failed / total_minutes) if total_minutes else 0,
             request_total_duration=state.request_total_finished_duration + state.request_total_failed_duration,
             requests_total=state.requests_failed + state.requests_finished,
-            crawler_runtime=crawler_runtime,
+            crawler_runtime=state.crawler_runtime,
             requests_finished=state.requests_finished,
             requests_failed=state.requests_failed,
             retry_histogram=serialized_state['request_retry_histogram'],
@@ -282,21 +288,6 @@ class Statistics(Generic[TStatisticsState]):
         else:
             self._periodic_message_logger.info(self._log_message, extra=stats.to_dict())
 
-    def _after_initialize(self) -> None:
-        state = self._state.current_value
-
-        if state.crawler_started_at is None:
-            state.crawler_started_at = datetime.now(timezone.utc)
-
-        if state.stats_persisted_at is not None and state.crawler_last_started_at:
-            self._instance_start = datetime.now(timezone.utc) - (
-                state.stats_persisted_at - state.crawler_last_started_at
-            )
-        elif state.crawler_last_started_at:
-            self._instance_start = state.crawler_last_started_at
-
-        state.crawler_last_started_at = self._instance_start
-
     def _save_retry_count_for_request(self, record: RequestProcessingRecord) -> None:
         retry_count = record.retry_count
         state = self._state.current_value

crawlee/storage_clients/__init__.py

@@ -13,9 +13,13 @@ _install_import_hook(__name__)
 with _try_import(__name__, 'SqlStorageClient'):
     from ._sql import SqlStorageClient
 
+with _try_import(__name__, 'RedisStorageClient'):
+    from ._redis import RedisStorageClient
+
 __all__ = [
     'FileSystemStorageClient',
     'MemoryStorageClient',
+    'RedisStorageClient',
     'SqlStorageClient',
     'StorageClient',
 ]

crawlee/storage_clients/_base/_dataset_client.py

@@ -87,8 +87,8 @@ class DatasetClient(ABC):
 
         The backend method for the `Dataset.iterate_items` call.
         """
-        # This syntax is to make mypy properly work with abstract AsyncIterator.
+        # This syntax is to make type checker properly work with abstract AsyncIterator.
         # https://mypy.readthedocs.io/en/stable/more_types.html#asynchronous-iterators
         raise NotImplementedError
-        if False:  # type: ignore[unreachable]
+        if False:
             yield 0

crawlee/storage_clients/_base/_key_value_store_client.py

@@ -72,10 +72,10 @@ class KeyValueStoreClient(ABC):
 
         The backend method for the `KeyValueStore.iterate_keys` call.
         """
-        # This syntax is to make mypy properly work with abstract AsyncIterator.
+        # This syntax is to make type checker properly work with abstract AsyncIterator.
         # https://mypy.readthedocs.io/en/stable/more_types.html#asynchronous-iterators
         raise NotImplementedError
-        if False:  # type: ignore[unreachable]
+        if False:
             yield 0
 
     @abstractmethod

crawlee/storage_clients/_file_system/_dataset_client.py

@@ -9,7 +9,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING, Any
 
 from pydantic import ValidationError
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._consts import METADATA_FILENAME
 from crawlee._utils.crypto import crypto_random_object_id
@@ -94,7 +94,7 @@ class FileSystemDatasetClient(DatasetClient):
         name: str | None,
         alias: str | None,
         configuration: Configuration,
-    ) -> FileSystemDatasetClient:
+    ) -> Self:
         """Open or create a file system dataset client.
 
         This method attempts to open an existing dataset from the file system. If a dataset with the specified ID
@@ -134,7 +134,7 @@ class FileSystemDatasetClient(DatasetClient):
                     continue
 
                 try:
-                    file = await asyncio.to_thread(path_to_metadata.open)
+                    file = await asyncio.to_thread(path_to_metadata.open, mode='r', encoding='utf-8')
                     try:
                         file_content = json.load(file)
                         metadata = DatasetMetadata(**file_content)
@@ -163,7 +163,7 @@ class FileSystemDatasetClient(DatasetClient):
 
             # If the dataset directory exists, reconstruct the client from the metadata file.
             if path_to_dataset.exists() and path_to_metadata.exists():
-                file = await asyncio.to_thread(open, path_to_metadata)
+                file = await asyncio.to_thread(path_to_metadata.open, mode='r', encoding='utf-8')
                 try:
                     file_content = json.load(file)
                 finally:
@@ -473,9 +473,10 @@ class FileSystemDatasetClient(DatasetClient):
         """
         # Retrieve and sort all JSON files in the dataset directory numerically.
         files = await asyncio.to_thread(
-            sorted,
-            self.path_to_dataset.glob('*.json'),
-            key=lambda f: int(f.stem) if f.stem.isdigit() else 0,
+            lambda: sorted(
+                self.path_to_dataset.glob('*.json'),
+                key=lambda f: int(f.stem) if f.stem.isdigit() else 0,
+            )
         )
 
         # Remove the metadata file from the list if present.

crawlee/storage_clients/_file_system/_key_value_store_client.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import asyncio
+import functools
 import json
 import shutil
 import urllib.parse
@@ -10,7 +11,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING, Any
 
 from pydantic import ValidationError
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._consts import METADATA_FILENAME
 from crawlee._utils.crypto import crypto_random_object_id
@@ -93,7 +94,7 @@ class FileSystemKeyValueStoreClient(KeyValueStoreClient):
         name: str | None,
         alias: str | None,
         configuration: Configuration,
-    ) -> FileSystemKeyValueStoreClient:
+    ) -> Self:
         """Open or create a file system key-value store client.
 
         This method attempts to open an existing key-value store from the file system. If a KVS with the specified
@@ -133,7 +134,7 @@ class FileSystemKeyValueStoreClient(KeyValueStoreClient):
                     continue
 
                 try:
-                    file = await asyncio.to_thread(path_to_metadata.open)
+                    file = await asyncio.to_thread(path_to_metadata.open, mode='r', encoding='utf-8')
                     try:
                         file_content = json.load(file)
                         metadata = KeyValueStoreMetadata(**file_content)
@@ -162,7 +163,7 @@ class FileSystemKeyValueStoreClient(KeyValueStoreClient):
 
             # If the key-value store directory exists, reconstruct the client from the metadata file.
             if path_to_kvs.exists() and path_to_metadata.exists():
-                file = await asyncio.to_thread(open, path_to_metadata)
+                file = await asyncio.to_thread(path_to_metadata.open, mode='r', encoding='utf-8')
                 try:
                     file_content = json.load(file)
                 finally:
@@ -239,7 +240,9 @@ class FileSystemKeyValueStoreClient(KeyValueStoreClient):
         # Read the metadata file
         async with self._lock:
             try:
-                file = await asyncio.to_thread(open, record_metadata_filepath)
+                file = await asyncio.to_thread(
+                    functools.partial(record_metadata_filepath.open, mode='r', encoding='utf-8'),
+                )
             except FileNotFoundError:
                 logger.warning(f'Metadata file disappeared for key "{key}", aborting get_value')
                 return None
@@ -373,7 +376,7 @@ class FileSystemKeyValueStoreClient(KeyValueStoreClient):
 
         # List and sort all files *inside* a brief lock, then release it immediately:
         async with self._lock:
-            files = sorted(await asyncio.to_thread(list, self.path_to_kvs.glob('*')))
+            files = sorted(await asyncio.to_thread(lambda: list(self.path_to_kvs.glob('*'))))
 
         count = 0
 

crawlee/storage_clients/_file_system/_request_queue_client.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import asyncio
+import functools
 import json
 import shutil
 from collections import deque
@@ -11,7 +12,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING
 
 from pydantic import BaseModel, ValidationError
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee import Request
 from crawlee._consts import METADATA_FILENAME
@@ -31,6 +32,7 @@ if TYPE_CHECKING:
     from collections.abc import Sequence
 
     from crawlee.configuration import Configuration
+    from crawlee.storages import KeyValueStore
 
 logger = getLogger(__name__)
 
@@ -92,6 +94,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
         metadata: RequestQueueMetadata,
         path_to_rq: Path,
         lock: asyncio.Lock,
+        recoverable_state: RecoverableState[RequestQueueState],
     ) -> None:
         """Initialize a new instance.
 
@@ -114,12 +117,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
         self._is_empty_cache: bool | None = None
         """Cache for is_empty result: None means unknown, True/False is cached state."""
 
-        self._state = RecoverableState[RequestQueueState](
-            default_state=RequestQueueState(),
-            persist_state_key=f'__RQ_STATE_{self._metadata.id}',
-            persistence_enabled=True,
-            logger=logger,
-        )
+        self._state = recoverable_state
         """Recoverable state to maintain request ordering, in-progress status, and handled status."""
 
     @override
@@ -136,6 +134,22 @@ class FileSystemRequestQueueClient(RequestQueueClient):
         """The full path to the request queue metadata file."""
         return self.path_to_rq / METADATA_FILENAME
 
+    @classmethod
+    async def _create_recoverable_state(cls, id: str, configuration: Configuration) -> RecoverableState:
+        async def kvs_factory() -> KeyValueStore:
+            from crawlee.storage_clients import FileSystemStorageClient  # noqa: PLC0415 avoid circular import
+            from crawlee.storages import KeyValueStore  # noqa: PLC0415 avoid circular import
+
+            return await KeyValueStore.open(storage_client=FileSystemStorageClient(), configuration=configuration)
+
+        return RecoverableState[RequestQueueState](
+            default_state=RequestQueueState(),
+            persist_state_key=f'__RQ_STATE_{id}',
+            persist_state_kvs_factory=kvs_factory,
+            persistence_enabled=True,
+            logger=logger,
+        )
+
     @classmethod
     async def open(
         cls,
@@ -144,7 +158,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
         name: str | None,
         alias: str | None,
         configuration: Configuration,
-    ) -> FileSystemRequestQueueClient:
+    ) -> Self:
         """Open or create a file system request queue client.
 
         This method attempts to open an existing request queue from the file system. If a queue with the specified
@@ -184,7 +198,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
                     continue
 
                 try:
-                    file = await asyncio.to_thread(path_to_metadata.open)
+                    file = await asyncio.to_thread(path_to_metadata.open, mode='r', encoding='utf-8')
                     try:
                         file_content = json.load(file)
                         metadata = RequestQueueMetadata(**file_content)
@@ -194,6 +208,9 @@ class FileSystemRequestQueueClient(RequestQueueClient):
                                 metadata=metadata,
                                 path_to_rq=rq_base_path / rq_dir,
                                 lock=asyncio.Lock(),
+                                recoverable_state=await cls._create_recoverable_state(
+                                    id=id, configuration=configuration
+                                ),
                             )
                             await client._state.initialize()
                             await client._discover_existing_requests()
@@ -216,7 +233,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
 
             # If the RQ directory exists, reconstruct the client from the metadata file.
             if path_to_rq.exists() and path_to_metadata.exists():
-                file = await asyncio.to_thread(open, path_to_metadata)
+                file = await asyncio.to_thread(path_to_metadata.open, encoding='utf-8')
                 try:
                     file_content = json.load(file)
                 finally:
@@ -230,6 +247,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
                     metadata=metadata,
                     path_to_rq=path_to_rq,
                     lock=asyncio.Lock(),
+                    recoverable_state=await cls._create_recoverable_state(id=metadata.id, configuration=configuration),
                 )
 
                 await client._state.initialize()
@@ -254,6 +272,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
                     metadata=metadata,
                     path_to_rq=path_to_rq,
                     lock=asyncio.Lock(),
+                    recoverable_state=await cls._create_recoverable_state(id=metadata.id, configuration=configuration),
                 )
                 await client._state.initialize()
                 await client._update_metadata()
@@ -738,7 +757,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
         await asyncio.to_thread(path_to_rq.mkdir, parents=True, exist_ok=True)
 
         # List all the json files.
-        files = await asyncio.to_thread(list, path_to_rq.glob('*.json'))
+        files = await asyncio.to_thread(lambda: list(path_to_rq.glob('*.json')))
 
         # Filter out metadata file and non-file entries.
         filtered = filter(lambda request_file: request_file.is_file() and request_file.name != METADATA_FILENAME, files)
@@ -757,7 +776,7 @@ class FileSystemRequestQueueClient(RequestQueueClient):
         """
         # Open the request file.
         try:
-            file = await asyncio.to_thread(open, file_path)
+            file = await asyncio.to_thread(functools.partial(file_path.open, mode='r', encoding='utf-8'))
         except FileNotFoundError:
             logger.warning(f'Request file "{file_path}" not found.')
             return None

crawlee/storage_clients/_memory/_dataset_client.py

@@ -4,7 +4,7 @@ from datetime import datetime, timezone
 from logging import getLogger
 from typing import TYPE_CHECKING, Any
 
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._utils.crypto import crypto_random_object_id
 from crawlee._utils.raise_if_too_many_kwargs import raise_if_too_many_kwargs
@@ -55,7 +55,7 @@ class MemoryDatasetClient(DatasetClient):
         id: str | None,
         name: str | None,
         alias: str | None,
-    ) -> MemoryDatasetClient:
+    ) -> Self:
         """Open or create a new memory dataset client.
 
         This method creates a new in-memory dataset instance. Unlike persistent storage implementations, memory

crawlee/storage_clients/_memory/_key_value_store_client.py

@@ -4,7 +4,7 @@ import sys
 from datetime import datetime, timezone
 from typing import TYPE_CHECKING, Any
 
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee._utils.crypto import crypto_random_object_id
 from crawlee._utils.file import infer_mime_type
@@ -53,7 +53,7 @@ class MemoryKeyValueStoreClient(KeyValueStoreClient):
         id: str | None,
         name: str | None,
         alias: str | None,
-    ) -> MemoryKeyValueStoreClient:
+    ) -> Self:
         """Open or create a new memory key-value store client.
 
         This method creates a new in-memory key-value store instance. Unlike persistent storage implementations,

crawlee/storage_clients/_memory/_request_queue_client.py

@@ -6,7 +6,7 @@ from datetime import datetime, timezone
 from logging import getLogger
 from typing import TYPE_CHECKING
 
-from typing_extensions import override
+from typing_extensions import Self, override
 
 from crawlee import Request
 from crawlee._utils.crypto import crypto_random_object_id
@@ -65,7 +65,7 @@ class MemoryRequestQueueClient(RequestQueueClient):
         id: str | None,
         name: str | None,
         alias: str | None,
-    ) -> MemoryRequestQueueClient:
+    ) -> Self:
         """Open or create a new memory request queue client.
 
         This method creates a new in-memory request queue instance. Unlike persistent storage implementations,

crawlee/storage_clients/_redis/__init__.py

@@ -0,0 +1,6 @@
+from ._dataset_client import RedisDatasetClient
+from ._key_value_store_client import RedisKeyValueStoreClient
+from ._request_queue_client import RedisRequestQueueClient
+from ._storage_client import RedisStorageClient
+
+__all__ = ['RedisDatasetClient', 'RedisKeyValueStoreClient', 'RedisRequestQueueClient', 'RedisStorageClient']