apify 3.4.2b26__tar.gz → 3.4.2b28__tar.gz

{apify-3.4.2b26 → apify-3.4.2b28}/CHANGELOG.md

@@ -8,6 +8,7 @@ All notable changes to this project will be documented in this file.
 ### 🚀 Features
 
 - Expose missing platform env vars via Actor.get_env() ([#984](https://github.com/apify/apify-sdk-python/pull/984)) ([f6e9b3b](https://github.com/apify/apify-sdk-python/commit/f6e9b3b328768aefb51257a258c4b6f03980fc72)) by [@vdusek](https://github.com/vdusek)
+- Re-export apify-client errors from apify.errors ([#990](https://github.com/apify/apify-sdk-python/pull/990)) ([165a3f6](https://github.com/apify/apify-sdk-python/commit/165a3f698dac6ace7394fc6b9f15ab7bb3a48050)) by [@vdusek](https://github.com/vdusek), closes [#988](https://github.com/apify/apify-sdk-python/issues/988)
 
 ### 🐛 Bug Fixes
 
@@ -37,6 +38,7 @@ All notable changes to this project will be documented in this file.
 - [**breaking**] Mark secondary arguments as keyword-only ([#917](https://github.com/apify/apify-sdk-python/pull/917)) ([eb94c99](https://github.com/apify/apify-sdk-python/commit/eb94c992ec4aba1cd7cf4dfd7a98731cb304651b)) by [@vdusek](https://github.com/vdusek), closes [#881](https://github.com/apify/apify-sdk-python/issues/881)
 - [**breaking**] Adapt to apify-client v3 ([#719](https://github.com/apify/apify-sdk-python/pull/719)) ([10203bc](https://github.com/apify/apify-sdk-python/commit/10203bc51e67590c97938b37d81614376bc3d29a)) by [@vdusek](https://github.com/vdusek), closes [#697](https://github.com/apify/apify-sdk-python/issues/697), [#736](https://github.com/apify/apify-sdk-python/issues/736), [#770](https://github.com/apify/apify-sdk-python/issues/770), [#853](https://github.com/apify/apify-sdk-python/issues/853)
 - [**breaking**] Remove wait_for_finish from Actor.start ([#962](https://github.com/apify/apify-sdk-python/pull/962)) ([0daca28](https://github.com/apify/apify-sdk-python/commit/0daca2818860cbcbd0778efa4d73622783ce63cf)) by [@vdusek](https://github.com/vdusek)
+- [**breaking**] Polish logging across the codebase ([#1005](https://github.com/apify/apify-sdk-python/pull/1005)) ([f949c24](https://github.com/apify/apify-sdk-python/commit/f949c24529e424b9151c8600f8057d7518c9b6ff)) by [@vdusek](https://github.com/vdusek)
 
 ### ⚙️ Miscellaneous Tasks
 

{apify-3.4.2b26 → apify-3.4.2b28}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: apify
-Version: 3.4.2b26
+Version: 3.4.2b28
 Summary: Apify SDK for Python
 Project-URL: Apify Homepage, https://apify.com
 Project-URL: Changelog, https://docs.apify.com/sdk/python/docs/changelog

{apify-3.4.2b26 → apify-3.4.2b28}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "apify"
-version = "3.4.2b26"
+version = "3.4.2b28"
 description = "Apify SDK for Python"
 authors = [{ name = "Apify Technologies s.r.o.", email = "support@apify.com" }]
 license = { file = "LICENSE" }

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/_actor.py

@@ -1162,7 +1162,7 @@ class _ActorType:
             custom_after_sleep: How long to sleep for after the metamorph, to wait for the container to be stopped.
         """
         if not self.is_at_home():
-            self.log.error('Actor.metamorph() is only supported when running on the Apify platform.')
+            self.log.warning('Actor.metamorph() is only supported when running on the Apify platform.')
             return
 
         if custom_after_sleep is None:
@@ -1172,6 +1172,7 @@ class _ActorType:
         if not self.configuration.actor_run_id:
             raise RuntimeError('actor_run_id cannot be None when running on the Apify platform.')
 
+        self.log.debug(f'Metamorphing the Actor run into target Actor {target_actor_id!r}.')
         await self.apify_client.run(self.configuration.actor_run_id).metamorph(
             target_actor_id=target_actor_id,
             run_input=run_input,
@@ -1198,7 +1199,7 @@ class _ActorType:
             custom_after_sleep: How long to sleep for after the reboot, to wait for the container to be stopped.
         """
         if not self.is_at_home():
-            self.log.error('Actor.reboot() is only supported when running on the Apify platform.')
+            self.log.warning('Actor.reboot() is only supported when running on the Apify platform.')
             return
 
         if self._is_rebooting:
@@ -1243,6 +1244,7 @@ class _ActorType:
             except TimeoutError:
                 self.log.warning('Pre-reboot event listeners did not finish within timeout; proceeding with reboot')
 
+            self.log.debug('Rebooting the Actor run.')
             await self.apify_client.run(self.configuration.actor_run_id).reboot()
         except BaseException:
             # Reset the flag so that a failed or cancelled reboot can be retried.
@@ -1277,7 +1279,7 @@ class _ActorType:
             )
 
         if not self.is_at_home():
-            self.log.error('Actor.add_webhook() is only supported when running on the Apify platform.')
+            self.log.warning('Actor.add_webhook() is only supported when running on the Apify platform.')
             return
 
         # If is_at_home() is True, config.actor_run_id is always set
@@ -1454,9 +1456,9 @@ class _ActorType:
             return max(self.configuration.timeout_at - datetime.now(tz=UTC), timedelta(0))
 
         self.log.warning(
-            'Using `inherit` argument is only possible when the Actor'
-            ' is running on the Apify platform and when the timeout for the Actor run is set. '
-            f'{self.is_at_home()=}, {self.configuration.timeout_at=}'
+            'Using the `inherit` argument is only possible when the Actor is running on the Apify platform and '
+            'the timeout for the Actor run is set.',
+            extra={'is_at_home': self.is_at_home(), 'timeout_at': self.configuration.timeout_at},
         )
         return None
 

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/_charging.py

@@ -5,6 +5,7 @@ from contextvars import ContextVar
 from dataclasses import dataclass
 from datetime import UTC, datetime
 from decimal import Decimal
+from logging import getLogger
 from typing import TYPE_CHECKING, Literal, Protocol, TypedDict
 
 from pydantic import ConfigDict
@@ -19,7 +20,6 @@ from apify_client._models import PricePerDatasetItemActorPricingInfo as ClientPr
 from apify_client._models import PricingPerEvent as ClientPricingPerEvent
 
 from apify._utils import ReentrantLock, docs_group, ensure_context
-from apify.log import logger
 from apify.storages import Dataset
 
 if TYPE_CHECKING:
@@ -29,6 +29,8 @@ if TYPE_CHECKING:
 
     from apify._configuration import Configuration
 
+logger = getLogger(__name__)
+
 charging_manager_ctx: ContextVar[ChargingManager | None] = ContextVar('charging_manager_ctx', default=None)
 """Holds the current `ChargingManager` instance, if any.
 
@@ -450,6 +452,7 @@ class ChargingManagerImplementation(ChargingManager):
                     pass
                 elif event_name in self._pricing_info:
                     await self._client.run(self._actor_run_id).charge(event_name, count=charged_count)
+                    logger.debug(f"Charged {charged_count} occurrence(s) of event '{event_name}'.")
                 elif event_name in self._tier_priced_events:
                     logger.warning(
                         f"Event '{event_name}' is tier-priced and is not chargeable via the pay-per-event API."

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/_proxy_configuration.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 import json
 import re
 from dataclasses import dataclass, field
+from logging import getLogger
 from re import Pattern
 from typing import TYPE_CHECKING, Any
 
@@ -22,13 +23,14 @@ from crawlee.proxy_configuration import (
 from apify._configuration import Configuration
 from apify._consts import ApifyEnvVars
 from apify._utils import docs_group
-from apify.log import logger
 
 if TYPE_CHECKING:
     from apify_client import ApifyClientAsync
 
     from apify import Request
 
+logger = getLogger(__name__)
+
 APIFY_PROXY_VALUE_REGEX = re.compile(r'^[\w._~]+$')
 COUNTRY_CODE_REGEX = re.compile(r'^[A-Z]{2}$')
 # ISO 3166-2 subdivision codes are 1-3 uppercase alphanumeric characters, e.g. 'CA', 'NSW', '9' (Wien, AT-9)

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/_utils.py

@@ -74,6 +74,7 @@ GroupName = Literal[
     'Actor',
     'Charging',
     'Configuration',
+    'Errors',
     'Event data',
     'Event managers',
     'Events',

apify-3.4.2b28/src/apify/errors.py

@@ -0,0 +1,34 @@
+"""`apify.errors` re-exports the Apify API client's error hierarchy.
+
+Callers get a single import location for every error raised by an operation that talks to the Apify API. The SDK
+raises these client exceptions as-is and does not wrap them in its own types. See
+https://docs.apify.com/api/client/python for the full client error reference.
+"""
+
+from __future__ import annotations
+
+from apify_client.errors import (
+    ApifyApiError,
+    ApifyClientError,
+    ConflictError,
+    ForbiddenError,
+    InvalidRequestError,
+    InvalidResponseBodyError,
+    NotFoundError,
+    RateLimitError,
+    ServerError,
+    UnauthorizedError,
+)
+
+__all__ = [
+    'ApifyApiError',
+    'ApifyClientError',
+    'ConflictError',
+    'ForbiddenError',
+    'InvalidRequestError',
+    'InvalidResponseBodyError',
+    'NotFoundError',
+    'RateLimitError',
+    'ServerError',
+    'UnauthorizedError',
+]

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/events/_apify_event_manager.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 import asyncio
 import contextlib
 import time
+from logging import getLogger
 from typing import TYPE_CHECKING, Annotated, Self
 
 import websockets.asyncio.client
@@ -17,7 +18,6 @@ from crawlee.events._types import Event, EventPersistStateData
 
 from apify._utils import docs_group
 from apify.events._types import DeprecatedEvent, EventMessage, SystemInfoEventData, UnknownEvent
-from apify.log import logger
 
 if TYPE_CHECKING:
     from collections.abc import Generator
@@ -28,6 +28,8 @@ if TYPE_CHECKING:
     from apify._configuration import Configuration
 
 
+logger = getLogger(__name__)
+
 event_data_adapter = TypeAdapter[EventMessage | DeprecatedEvent | UnknownEvent](
     Annotated[EventMessage, Discriminator('name')] | DeprecatedEvent | UnknownEvent
 )
@@ -195,7 +197,7 @@ class ApifyEventManager(EventManager):
                 return
 
             if isinstance(parsed_message, UnknownEvent):
-                logger.info(
+                logger.debug(
                     f'Unknown message received: event_name={parsed_message.name}, event_data={parsed_message.data}'
                 )
                 return

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/scrapy/middlewares/apify_proxy.py

@@ -72,7 +72,6 @@ class ApifyHttpProxyMiddleware:
         Raises:
             ValueError: If username and password are not provided in the proxy URL.
         """
-        Actor.log.debug(f'ApifyHttpProxyMiddleware.process_request: request={request}')
         url = await self._get_new_proxy_url()
 
         if not (url.username and url.password):
@@ -82,8 +81,6 @@ class ApifyHttpProxyMiddleware:
         basic_auth_header = get_basic_auth_header(url.username, url.password)
         request.headers[b'Proxy-Authorization'] = basic_auth_header
 
-        Actor.log.debug(f'ApifyHttpProxyMiddleware.process_request: updated request.meta={request.meta}')
-
     def process_exception(
         self,
         request: Request,
@@ -100,10 +97,6 @@ class ApifyHttpProxyMiddleware:
             process_exception() methods of installed middleware, until no middleware is left and the default
             exception handling kicks in.
         """
-        Actor.log.debug(
-            f'ApifyHttpProxyMiddleware.process_exception: request={request}, exception={exception}',
-        )
-
         if isinstance(exception, TunnelError):
             Actor.log.warning(
                 f'ApifyHttpProxyMiddleware: TunnelError occurred for request="{request}", '

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/scrapy/pipelines/actor_dataset_push.py

@@ -1,6 +1,5 @@
 from __future__ import annotations
 
-from logging import getLogger
 from typing import TYPE_CHECKING
 
 from itemadapter.adapter import ItemAdapter
@@ -10,8 +9,6 @@ from apify import Actor
 if TYPE_CHECKING:
     from scrapy import Item
 
-logger = getLogger(__name__)
-
 
 class ActorDatasetPushPipeline:
     """A Scrapy pipeline for pushing items to an Actor's default dataset.
@@ -25,6 +22,5 @@ class ActorDatasetPushPipeline:
     ) -> Item:
         """Pushes the provided Scrapy item to the Actor's default dataset."""
         item_dict = ItemAdapter(item).asdict()
-        logger.debug(f'Pushing item={item_dict} to the dataset.')
         await Actor.push_data(item_dict)
         return item

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/scrapy/requests.py

@@ -61,8 +61,6 @@ def to_apify_request(scrapy_request: ScrapyRequest, spider: Spider) -> ApifyRequ
         logger.warning('Failed to convert to Apify request: Scrapy request must be a ScrapyRequest instance.')
         return None
 
-    logger.debug(f'to_apify_request was called (scrapy_request={scrapy_request})...')
-
     # Configuration to behave as similarly as possible to Scrapy's default RFPDupeFilter.
     #
     # The body is stored twice on purpose: as `payload` (used for the extended unique key) and inside the serialized
@@ -150,7 +148,6 @@ def to_apify_request(scrapy_request: ScrapyRequest, spider: Spider) -> ApifyRequ
     # a second base64 layer would only add ~33% overhead on the enqueue path.
     apify_request.user_data['scrapy_request'] = scrapy_request_json
 
-    logger.debug(f'scrapy_request was converted to the apify_request={apify_request}')
     return apify_request
 
 
@@ -171,14 +168,10 @@ def to_scrapy_request(apify_request: ApifyRequest, spider: Spider) -> ScrapyRequ
     if not isinstance(cast('Any', apify_request), ApifyRequest):
         raise TypeError('apify_request must be an apify.Request instance')
 
-    logger.debug(f'to_scrapy_request was called (apify_request={apify_request})...')
-
     # If the apify_request comes from the Scrapy
     if 'scrapy_request' in apify_request.user_data:
         # Deserialize the Scrapy ScrapyRequest from the apify_request by parsing the stored JSON and reconstructing
         # the Scrapy ScrapyRequest object from its dictionary representation.
-        logger.debug('Restoring the Scrapy ScrapyRequest from the apify_request...')
-
         scrapy_request_json = apify_request.user_data['scrapy_request']
         if not isinstance(scrapy_request_json, str):
             raise TypeError('the stored scrapy_request must be a string')
@@ -194,8 +187,6 @@ def to_scrapy_request(apify_request: ApifyRequest, spider: Spider) -> ScrapyRequ
         if not isinstance(scrapy_request, ScrapyRequest):
             raise TypeError('scrapy_request must be an instance of the ScrapyRequest class')
 
-        logger.debug(f'Scrapy ScrapyRequest successfully reconstructed (scrapy_request={scrapy_request})...')
-
         # Update the meta field with the meta field from the apify_request
         meta = scrapy_request.meta or {}
         meta.update({'apify_request_unique_key': apify_request.unique_key})
@@ -204,8 +195,6 @@ def to_scrapy_request(apify_request: ApifyRequest, spider: Spider) -> ScrapyRequ
 
     # If the apify_request comes directly from the Scrapy, typically start URLs.
     else:
-        logger.debug('Gonna create a new Scrapy ScrapyRequest (cannot be restored)')
-
         scrapy_request = ScrapyRequest(
             url=apify_request.url,
             method=apify_request.method,
@@ -222,5 +211,4 @@ def to_scrapy_request(apify_request: ApifyRequest, spider: Spider) -> ScrapyRequ
     if apify_request.user_data:
         scrapy_request.meta['userData'] = apify_request.user_data
 
-    logger.debug(f'an apify_request was converted to the scrapy_request={scrapy_request}')
     return scrapy_request

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/scrapy/scheduler.py

@@ -134,17 +134,14 @@ class ApifyScheduler(BaseScheduler):
         Returns:
             True if the request was successfully enqueued, False otherwise.
         """
-        logger.debug(f'ApifyScheduler.enqueue_request was called (scrapy_request={request})...')
-
         if not isinstance(self.spider, Spider):
             raise TypeError('self.spider must be an instance of the Spider class')
 
         apify_request = to_apify_request(request, spider=self.spider)
         if apify_request is None:
-            logger.error(f'Request {request} could not be converted to Apify request.')
+            logger.warning(f'Request {request} could not be converted to Apify request; skipping it.')
             return False
 
-        logger.debug(f'Converted to apify_request: {apify_request}')
         if not isinstance(self._rq, RequestQueue):
             raise TypeError('self._rq must be an instance of the RequestQueue class')
 
@@ -156,7 +153,6 @@ class ApifyScheduler(BaseScheduler):
             logger.exception('Failed to enqueue the request to the request queue.')
             raise
 
-        logger.debug(f'rq.add_request result: {result}')
         return not bool(result.was_already_present)
 
     def next_request(self) -> Request | None:
@@ -165,7 +161,6 @@ class ApifyScheduler(BaseScheduler):
         Returns:
             The next request, or None if there are no more requests.
         """
-        logger.debug('next_request called...')
         if not isinstance(self._rq, RequestQueue):
             raise TypeError('self._rq must be an instance of the RequestQueue class')
 
@@ -177,7 +172,6 @@ class ApifyScheduler(BaseScheduler):
             logger.exception('Failed to fetch the next request from the request queue.')
             raise
 
-        logger.debug(f'Fetched apify_request: {apify_request}')
         if apify_request is None:
             return None
 
@@ -206,5 +200,4 @@ class ApifyScheduler(BaseScheduler):
         if scrapy_request is None:
             return None
 
-        logger.debug(f'Converted to scrapy_request: {scrapy_request}')
         return scrapy_request

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/storage_clients/_apify/_alias_resolving.py

@@ -1,6 +1,5 @@
 from __future__ import annotations
 
-import logging
 from asyncio import Lock
 from functools import cached_property
 from logging import getLogger
@@ -225,7 +224,7 @@ class AliasResolver:
         alias_map[self._storage_key] = storage_id
 
         if not self._configuration.is_at_home:
-            logging.getLogger(__name__).debug(
+            logger.debug(
                 'AliasResolver storage limited retention is only supported on Apify platform. Storage is not exported.'
             )
             return

{apify-3.4.2b26 → apify-3.4.2b28}/src/apify/storage_clients/_file_system/_key_value_store_client.py

@@ -122,7 +122,7 @@ class ApifyFileSystemKeyValueStoreClient(FileSystemKeyValueStoreClient):
         try:
             content = await asyncio.to_thread(record_path.read_bytes)
         except FileNotFoundError:
-            logger.warning(f'Input file disparaged on path: "{record_path}"')
+            logger.warning(f'Input file disappeared from path: "{record_path}"')
             return
 
         # Figure out the metadata from the file content