apify 2.3.1b2__py3-none-any.whl → 2.4.0__py3-none-any.whl

apify/_actor.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 import asyncio
 import os
 import sys
+from contextlib import suppress
 from datetime import timedelta
 from typing import TYPE_CHECKING, Any, Callable, Literal, TypeVar, cast, overload
 
@@ -43,7 +44,7 @@ if TYPE_CHECKING:
     from typing_extensions import Self
 
     from crawlee.proxy_configuration import _NewUrlFunction
-    from crawlee.storage_clients import BaseStorageClient
+    from crawlee.storage_clients import StorageClient
 
     from apify._models import Webhook
 
@@ -64,6 +65,7 @@ class _ActorType:
         configuration: Configuration | None = None,
         *,
         configure_logging: bool = True,
+        exit_process: bool | None = None,
     ) -> None:
         """Create an Actor instance.
 
@@ -74,7 +76,10 @@ class _ActorType:
             configuration: The Actor configuration to be used. If not passed, a new Configuration instance will
                 be created.
             configure_logging: Should the default logging configuration be configured?
+            exit_process: Whether the Actor should call `sys.exit` when the context manager exits. The default is
+                True except for the IPython, Pytest and Scrapy environments.
         """
+        self._exit_process = self._get_default_exit_process() if exit_process is None else exit_process
         self._is_exiting = False
 
         self._configuration = configuration or Configuration.get_global_configuration()
@@ -141,9 +146,19 @@ class _ActorType:
 
         return super().__repr__()
 
-    def __call__(self, configuration: Configuration | None = None, *, configure_logging: bool = True) -> Self:
+    def __call__(
+        self,
+        configuration: Configuration | None = None,
+        *,
+        configure_logging: bool = True,
+        exit_process: bool | None = None,
+    ) -> Self:
         """Make a new Actor instance with a non-default configuration."""
-        return self.__class__(configuration=configuration, configure_logging=configure_logging)
+        return self.__class__(
+            configuration=configuration,
+            configure_logging=configure_logging,
+            exit_process=exit_process,
+        )
 
     @property
     def apify_client(self) -> ApifyClientAsync:
@@ -171,7 +186,7 @@ class _ActorType:
         return logger
 
     @property
-    def _local_storage_client(self) -> BaseStorageClient:
+    def _local_storage_client(self) -> StorageClient:
         """The local storage client the Actor instance uses."""
         return service_locator.get_storage_client()
 
@@ -281,13 +296,7 @@ class _ActorType:
         await asyncio.wait_for(finalize(), cleanup_timeout.total_seconds())
         self._is_initialized = False
 
-        if is_running_in_ipython():
-            self.log.debug(f'Not calling sys.exit({exit_code}) because Actor is running in IPython')
-        elif os.getenv('PYTEST_CURRENT_TEST', default=False):  # noqa: PLW1508
-            self.log.debug(f'Not calling sys.exit({exit_code}) because Actor is running in an unit test')
-        elif os.getenv('SCRAPY_SETTINGS_MODULE'):
-            self.log.debug(f'Not calling sys.exit({exit_code}) because Actor is running with Scrapy')
-        else:
+        if self._exit_process:
             sys.exit(exit_code)
 
     async def fail(
@@ -1128,6 +1137,26 @@ class _ActorType:
 
         return proxy_configuration
 
+    def _get_default_exit_process(self) -> bool:
+        """Returns False for IPython, Pytest, and Scrapy environments, True otherwise."""
+        if is_running_in_ipython():
+            self.log.debug('Running in IPython, setting default `exit_process` to False.')
+            return False
+
+        # Check if running in Pytest by detecting the relevant environment variable.
+        if os.getenv('PYTEST_CURRENT_TEST'):
+            self.log.debug('Running in Pytest, setting default `exit_process` to False.')
+            return False
+
+        # Check if running in Scrapy by attempting to import it.
+        with suppress(ImportError):
+            import scrapy  # noqa: F401
+
+            self.log.debug('Running in Scrapy, setting default `exit_process` to False.')
+            return False
+
+        return True
+
 
 Actor = cast(_ActorType, Proxy(_ActorType))
 """The entry point of the SDK, through which all the Actor operations should be done."""

apify/_crypto.py

@@ -1,6 +1,9 @@
 from __future__ import annotations
 
 import base64
+import hashlib
+import hmac
+import string
 from typing import Any
 
 from cryptography.exceptions import InvalidTag as InvalidTagException
@@ -153,3 +156,38 @@ def decrypt_input_secrets(private_key: rsa.RSAPrivateKey, input_data: Any) -> An
                 )
 
     return input_data
+
+
+CHARSET = string.digits + string.ascii_letters
+
+
+def encode_base62(num: int) -> str:
+    """Encode the given number to base62."""
+    if num == 0:
+        return CHARSET[0]
+
+    res = ''
+    while num > 0:
+        num, remainder = divmod(num, 62)
+        res = CHARSET[remainder] + res
+    return res
+
+
+@ignore_docs
+def create_hmac_signature(secret_key: str, message: str) -> str:
+    """Generate an HMAC signature and encodes it using Base62. Base62 encoding reduces the signature length.
+
+    HMAC signature is truncated to 30 characters to make it shorter.
+
+    Args:
+        secret_key: Secret key used for signing signatures.
+        message: Message to be signed.
+
+    Returns:
+        Base62 encoded signature.
+    """
+    signature = hmac.new(secret_key.encode('utf-8'), message.encode('utf-8'), hashlib.sha256).hexdigest()[:30]
+
+    decimal_signature = int(signature, 16)
+
+    return encode_base62(decimal_signature)

apify/_platform_event_manager.py

@@ -4,7 +4,7 @@ import asyncio
 from datetime import datetime
 from typing import TYPE_CHECKING, Annotated, Any, Literal, Union
 
-import websockets.client
+import websockets.asyncio.client
 from pydantic import BaseModel, Discriminator, Field, TypeAdapter
 from typing_extensions import Self, Unpack, override
 
@@ -143,7 +143,7 @@ class PlatformEventManager(EventManager):
     but instead use it via the `Actor.on()` and `Actor.off()` methods.
     """
 
-    _platform_events_websocket: websockets.client.WebSocketClientProtocol | None = None
+    _platform_events_websocket: websockets.asyncio.client.ClientConnection | None = None
     _process_platform_messages_task: asyncio.Task | None = None
     _send_system_info_interval_task: asyncio.Task | None = None
     _connected_to_platform_websocket: asyncio.Future = asyncio.Future()
@@ -196,7 +196,7 @@ class PlatformEventManager(EventManager):
 
     async def _process_platform_messages(self, ws_url: str) -> None:
         try:
-            async with websockets.client.connect(ws_url) as websocket:
+            async with websockets.asyncio.client.connect(ws_url) as websocket:
                 self._platform_events_websocket = websocket
                 self._connected_to_platform_websocket.set_result(True)
 

apify/apify_storage_client/_apify_storage_client.py

@@ -6,7 +6,7 @@ from typing_extensions import override
 
 from apify_client import ApifyClientAsync
 from crawlee._utils.crypto import crypto_random_object_id
-from crawlee.storage_clients import BaseStorageClient
+from crawlee.storage_clients import StorageClient
 
 from apify._utils import docs_group
 from apify.apify_storage_client._dataset_client import DatasetClient
@@ -21,7 +21,7 @@ if TYPE_CHECKING:
 
 
 @docs_group('Classes')
-class ApifyStorageClient(BaseStorageClient):
+class ApifyStorageClient(StorageClient):
     """A storage client implementation based on the Apify platform storage."""
 
     def __init__(self, *, configuration: Configuration) -> None:
@@ -68,5 +68,5 @@ class ApifyStorageClient(BaseStorageClient):
         pass
 
     @override
-    def get_rate_limit_errors(self) -> dict[int, int]:  # type: ignore[misc]
+    def get_rate_limit_errors(self) -> dict[int, int]:
         return self._apify_client.stats.rate_limit_errors

apify/apify_storage_client/_dataset_client.py

@@ -4,7 +4,7 @@ from typing import TYPE_CHECKING
 
 from typing_extensions import override
 
-from crawlee.storage_clients._base import BaseDatasetClient
+from crawlee.storage_clients._base import DatasetClient as BaseDatasetClient
 from crawlee.storage_clients.models import DatasetItemsListPage, DatasetMetadata
 
 if TYPE_CHECKING:

apify/apify_storage_client/_dataset_collection_client.py

@@ -4,7 +4,7 @@ from typing import TYPE_CHECKING
 
 from typing_extensions import override
 
-from crawlee.storage_clients._base import BaseDatasetCollectionClient
+from crawlee.storage_clients._base import DatasetCollectionClient as BaseDatasetCollectionClient
 from crawlee.storage_clients.models import DatasetListPage, DatasetMetadata
 
 if TYPE_CHECKING:

apify/apify_storage_client/_key_value_store_client.py

@@ -4,10 +4,13 @@ from contextlib import asynccontextmanager
 from typing import TYPE_CHECKING, Any
 
 from typing_extensions import override
+from yarl import URL
 
-from crawlee.storage_clients._base import BaseKeyValueStoreClient
+from crawlee.storage_clients._base import KeyValueStoreClient as BaseKeyValueStoreClient
 from crawlee.storage_clients.models import KeyValueStoreListKeysPage, KeyValueStoreMetadata, KeyValueStoreRecord
 
+from apify._crypto import create_hmac_signature
+
 if TYPE_CHECKING:
     from collections.abc import AsyncIterator
     from contextlib import AbstractAsyncContextManager
@@ -89,6 +92,18 @@ class KeyValueStoreClient(BaseKeyValueStoreClient):
         Args:
             key: The key for which the URL should be generated.
         """
-        public_api_url = self._api_public_base_url
+        if self._client.resource_id is None:
+            raise ValueError('resource_id cannot be None when generating a public URL')
+
+        public_url = (
+            URL(self._api_public_base_url) / 'v2' / 'key-value-stores' / self._client.resource_id / 'records' / key
+        )
+
+        key_value_store = await self.get()
+
+        if key_value_store is not None and isinstance(key_value_store.model_extra, dict):
+            url_signing_secret_key = key_value_store.model_extra.get('urlSigningSecretKey')
+            if url_signing_secret_key:
+                public_url = public_url.with_query(signature=create_hmac_signature(url_signing_secret_key, key))
 
-        return f'{public_api_url}/v2/key-value-stores/{self._client.resource_id}/records/{key}'
+        return str(public_url)

apify/apify_storage_client/_key_value_store_collection_client.py

@@ -4,7 +4,7 @@ from typing import TYPE_CHECKING
 
 from typing_extensions import override
 
-from crawlee.storage_clients._base import BaseKeyValueStoreCollectionClient
+from crawlee.storage_clients._base import KeyValueStoreCollectionClient as BaseKeyValueStoreCollectionClient
 from crawlee.storage_clients.models import KeyValueStoreListPage, KeyValueStoreMetadata
 
 if TYPE_CHECKING:

apify/apify_storage_client/_request_queue_client.py

@@ -5,7 +5,7 @@ from typing import TYPE_CHECKING
 from typing_extensions import override
 
 from crawlee import Request
-from crawlee.storage_clients._base import BaseRequestQueueClient
+from crawlee.storage_clients._base import RequestQueueClient as BaseRequestQueueClient
 from crawlee.storage_clients.models import (
     BatchRequestsOperationResponse,
     ProcessedRequest,

apify/apify_storage_client/_request_queue_collection_client.py

@@ -4,7 +4,7 @@ from typing import TYPE_CHECKING
 
 from typing_extensions import override
 
-from crawlee.storage_clients._base import BaseRequestQueueCollectionClient
+from crawlee.storage_clients._base import RequestQueueCollectionClient as BaseRequestQueueCollectionClient
 from crawlee.storage_clients.models import RequestQueueListPage, RequestQueueMetadata
 
 if TYPE_CHECKING:

apify/storages/_request_list.py

@@ -10,7 +10,7 @@ from pydantic import BaseModel, Field, TypeAdapter
 
 from crawlee import Request
 from crawlee._types import HttpMethod
-from crawlee.http_clients import BaseHttpClient, HttpxHttpClient
+from crawlee.http_clients import HttpClient, HttpxHttpClient
 from crawlee.request_loaders import RequestList as CrawleeRequestList
 
 from apify._utils import docs_group
@@ -49,7 +49,7 @@ class RequestList(CrawleeRequestList):
     async def open(
         name: str | None = None,
         request_list_sources_input: list[dict[str, Any]] | None = None,
-        http_client: BaseHttpClient | None = None,
+        http_client: HttpClient | None = None,
     ) -> RequestList:
         """Creates RequestList from Actor input requestListSources.
 
@@ -78,7 +78,7 @@ class RequestList(CrawleeRequestList):
 
     @staticmethod
     async def _create_request_list(
-        name: str | None, request_list_sources_input: list[dict[str, Any]], http_client: BaseHttpClient | None
+        name: str | None, request_list_sources_input: list[dict[str, Any]], http_client: HttpClient | None
     ) -> RequestList:
         if not http_client:
             http_client = HttpxHttpClient()
@@ -108,7 +108,7 @@ class RequestList(CrawleeRequestList):
 
     @staticmethod
     async def _fetch_requests_from_url(
-        remote_url_requests_inputs: list[_RequestsFromUrlInput], http_client: BaseHttpClient
+        remote_url_requests_inputs: list[_RequestsFromUrlInput], http_client: HttpClient
     ) -> list[Request]:
         """Crete list of requests from url.
 

{apify-2.3.1b2.dist-info → apify-2.4.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: apify
-Version: 2.3.1b2
+Version: 2.4.0
 Summary: Apify SDK for Python
 Project-URL: Homepage, https://docs.apify.com/sdk/python/
 Project-URL: Apify homepage, https://apify.com
@@ -224,14 +224,14 @@ Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Software Development :: Libraries
 Requires-Python: >=3.9
 Requires-Dist: apify-client>=1.9.2
-Requires-Dist: apify-shared>=1.2.1
-Requires-Dist: crawlee~=0.5.0
+Requires-Dist: apify-shared>=1.3.0
+Requires-Dist: crawlee~=0.6.0
 Requires-Dist: cryptography>=42.0.0
 Requires-Dist: httpx>=0.27.0
 Requires-Dist: lazy-object-proxy>=1.10.0
 Requires-Dist: more-itertools>=10.2.0
 Requires-Dist: typing-extensions>=4.1.0
-Requires-Dist: websockets<14.0.0,>=10.0
+Requires-Dist: websockets>=14.0
 Provides-Extra: scrapy
 Requires-Dist: scrapy>=2.11.0; extra == 'scrapy'
 Description-Content-Type: text/markdown

{apify-2.3.1b2.dist-info → apify-2.4.0.dist-info}/RECORD

@@ -1,23 +1,23 @@
 apify/__init__.py,sha256=HpgKg2FZWJuSPfDygzJ62psylhw4NN4tKFnoYUIhcd4,838
-apify/_actor.py,sha256=EB3gGjASV0PbPJ6BtgOq45HN23vM-9ceNCNRfeh2BkQ,48821
+apify/_actor.py,sha256=PQqFDpAqSbh_aP3EjD8yLGYOmLZMo1qLzqrbVT2KjWE,49697
 apify/_charging.py,sha256=m7hJIQde4M7vS4g_4hsNRP5xHNXjYQ8MyqOEGeNb7VY,12267
 apify/_configuration.py,sha256=yidcWHsu-IJ2mmLmXStKq_HHcdfQxZq7koYjlZfRnQ8,11128
 apify/_consts.py,sha256=_Xq4hOfOA1iZ3n1P967YWdyncKivpbX6RTlp_qanUoE,330
-apify/_crypto.py,sha256=e0_aM3l9_5Osk-jszYOOjrAKK60OggSHbiw5c30QnsU,5638
+apify/_crypto.py,sha256=8BgeQC0ZhYP5KdmLxxLQAW87Gq-Z4HlREbYGXr46w0U,6607
 apify/_models.py,sha256=-Y0rljBJWxMMCp8iDCTG4UV3bEvNZzp-kx2SYbPfeIY,7919
-apify/_platform_event_manager.py,sha256=44xyV0Lpzf4h4VZ0rkyYg_nhbQkEONNor8_Z9gIKO40,7899
+apify/_platform_event_manager.py,sha256=k1e5ruSJdcMKr6j-_XIF-gfhrgyMzdSenYW0QoJROu8,7916
 apify/_proxy_configuration.py,sha256=c-O6_PZ9pUD-i4J0RFEKTtfyJPP2rTRJJA1TH8NVsV8,13189
 apify/_utils.py,sha256=92byxeXTpDFwhBq7ZS-obeXKtKWvVzCZMV0Drg3EjhQ,1634
 apify/log.py,sha256=j-E4t-WeA93bc1NCQRG8sTntehQCiiN8ia-MdQe3_Ts,1291
 apify/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 apify/apify_storage_client/__init__.py,sha256=-UbR68bFsDR6ln8OFs4t50eqcnY36hujO-SeOt-KmcA,114
-apify/apify_storage_client/_apify_storage_client.py,sha256=jTX5vd-K9mnFTyZu2V2dUg7oyWogvmNIDUlEXnvIlOw,2766
-apify/apify_storage_client/_dataset_client.py,sha256=UUodnR_MQBg5RkURrfegkGJWR5OmdPPgPfGepvkdQoU,5580
-apify/apify_storage_client/_dataset_collection_client.py,sha256=qCcKZlA0bkO-sL7xED0Yose85NlrRa9AKr4oCSrYX6k,1489
-apify/apify_storage_client/_key_value_store_client.py,sha256=MSuoIeqEHLu92WfUU7kyB3Cc_gKUlm8TghnU3_xkPtE,3363
-apify/apify_storage_client/_key_value_store_collection_client.py,sha256=NxD-3XDJP6JGMDyIa6ib0gl8op7rQjSQ0vlToCiV190,1563
-apify/apify_storage_client/_request_queue_client.py,sha256=n-CR-hA5LM6_8IwiMwQ9tT2juavq7X2zC3ZNlrtv-2s,5156
-apify/apify_storage_client/_request_queue_collection_client.py,sha256=MdzgbQb2D8rHWpUlPCrQSHRlAi0fI0PSZ9bYagr-MhY,1571
+apify/apify_storage_client/_apify_storage_client.py,sha256=qeWYsEQGeyyhJzS9TZTQFNqdSl8JzHz_4_HDKGY4I_Y,2736
+apify/apify_storage_client/_dataset_client.py,sha256=9RxxhrJMic5QRJn2Vl4J-FnSlEigIpYW5Z_2B1dcRzM,5597
+apify/apify_storage_client/_dataset_collection_client.py,sha256=gf5skMTkfpGhEscRy5bgo13vznxGZrSd7w9Ivh3Usyc,1516
+apify/apify_storage_client/_key_value_store_client.py,sha256=OCFUAW0o-8KQvUpL8zmlZrpU3yRmDKdsO2529H2v40I,4002
+apify/apify_storage_client/_key_value_store_collection_client.py,sha256=zjsbRW4zjme6dIzxxlHyCW3voBA5489MUhdjl5YMaro,1596
+apify/apify_storage_client/_request_queue_client.py,sha256=cNMhXz85s1ZtjLpVqkduYl1y6o9QyNdcIGoy6ccD-h0,5178
+apify/apify_storage_client/_request_queue_collection_client.py,sha256=MTLM2cG0txAe3cSjkGbXyq2Ek0R7wlsMbGGULmQGD3I,1603
 apify/apify_storage_client/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 apify/scrapy/__init__.py,sha256=m2a0ts_JY9xJkBy4JU5mV8PJqjA3GGKLXBFu4nl-n-A,1048
 apify/scrapy/_actor_runner.py,sha256=rXWSnlQWGskDUH8PtLCv5SkOIx4AiVa4QbCYeCett5c,938
@@ -34,9 +34,9 @@ apify/scrapy/pipelines/__init__.py,sha256=GWPeLN_Zwj8vRBWtXW6DaxdB7mvyQ7Jw5Tz1cc
 apify/scrapy/pipelines/actor_dataset_push.py,sha256=XUUyznQTD-E3wYUUFt2WAOnWhbnRrY0WuedlfYfYhDI,846
 apify/scrapy/pipelines/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 apify/storages/__init__.py,sha256=FW-z6ubuPnHGM-Wp15T8mR5q6lnpDGrCW-IkgZd5L30,177
-apify/storages/_request_list.py,sha256=-lZJcE5nq69aJhGFJ7Sh2ctqgAWUDyOwYm5_0y1hdAE,5865
+apify/storages/_request_list.py,sha256=7WpcdWvT3QxEBthynBpTVCSNDLXq6UbpQQmfUVyJ1jE,5849
 apify/storages/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-apify-2.3.1b2.dist-info/METADATA,sha256=OYF8JJ4fpejdIurNXYEWXEa1Zrkmljyaq_ITIXeynU4,21566
-apify-2.3.1b2.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-apify-2.3.1b2.dist-info/licenses/LICENSE,sha256=AsFjHssKjj4LGd2ZCqXn6FBzMqcWdjQre1byPPSypVw,11355
-apify-2.3.1b2.dist-info/RECORD,,
+apify-2.4.0.dist-info/METADATA,sha256=u-SDIKoBk3as6apTeCKo4zCMdB3jvH26c9QauKoCdCk,21556
+apify-2.4.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+apify-2.4.0.dist-info/licenses/LICENSE,sha256=AsFjHssKjj4LGd2ZCqXn6FBzMqcWdjQre1byPPSypVw,11355
+apify-2.4.0.dist-info/RECORD,,