crawlee 1.0.1b8__py3-none-any.whl → 1.0.5b18__py3-none-any.whl

crawlee/_request.py

@@ -185,9 +185,6 @@ class Request(BaseModel):
     method: HttpMethod = 'GET'
     """HTTP request method."""
 
-    headers: Annotated[HttpHeaders, Field(default_factory=HttpHeaders)] = HttpHeaders()
-    """HTTP request headers."""
-
     payload: Annotated[
         HttpPayload | None,
         BeforeValidator(lambda v: v.encode() if isinstance(v, str) else v),
@@ -195,23 +192,37 @@ class Request(BaseModel):
     ] = None
     """HTTP request payload."""
 
-    user_data: Annotated[
-        dict[str, JsonSerializable],  # Internally, the model contains `UserData`, this is just for convenience
-        Field(alias='userData', default_factory=lambda: UserData()),
-        PlainValidator(user_data_adapter.validate_python),
-        PlainSerializer(
-            lambda instance: user_data_adapter.dump_python(
-                instance,
-                by_alias=True,
-                exclude_none=True,
-                exclude_unset=True,
-                exclude_defaults=True,
-            )
-        ),
-    ] = {}
-    """Custom user data assigned to the request. Use this to save any request related data to the
-    request's scope, keeping them accessible on retries, failures etc.
-    """
+    # Workaround for pydantic 2.12 and mypy type checking issue for Annotated with default_factory
+    if TYPE_CHECKING:
+        headers: HttpHeaders = HttpHeaders()
+        """HTTP request headers."""
+
+        user_data: dict[str, JsonSerializable] = {}
+        """Custom user data assigned to the request. Use this to save any request related data to the
+        request's scope, keeping them accessible on retries, failures etc.
+        """
+
+    else:
+        headers: Annotated[HttpHeaders, Field(default_factory=HttpHeaders)]
+        """HTTP request headers."""
+
+        user_data: Annotated[
+            dict[str, JsonSerializable],  # Internally, the model contains `UserData`, this is just for convenience
+            Field(alias='userData', default_factory=lambda: UserData()),
+            PlainValidator(user_data_adapter.validate_python),
+            PlainSerializer(
+                lambda instance: user_data_adapter.dump_python(
+                    instance,
+                    by_alias=True,
+                    exclude_none=True,
+                    exclude_unset=True,
+                    exclude_defaults=True,
+                )
+            ),
+        ]
+        """Custom user data assigned to the request. Use this to save any request related data to the
+        request's scope, keeping them accessible on retries, failures etc.
+        """
 
     retry_count: Annotated[int, Field(alias='retryCount')] = 0
     """Number of times the request has been retried."""

crawlee/_service_locator.py

@@ -38,7 +38,7 @@ class ServiceLocator:
     def get_configuration(self) -> Configuration:
         """Get the configuration."""
         if self._configuration is None:
-            logger.warning('No configuration set, implicitly creating and using default Configuration.')
+            logger.debug('No configuration set, implicitly creating and using default Configuration.')
             self._configuration = Configuration()
 
         return self._configuration
@@ -63,9 +63,9 @@ class ServiceLocator:
     def get_event_manager(self) -> EventManager:
         """Get the event manager."""
         if self._event_manager is None:
-            logger.warning('No event manager set, implicitly creating and using default LocalEventManager.')
+            logger.debug('No event manager set, implicitly creating and using default LocalEventManager.')
             if self._configuration is None:
-                logger.warning(
+                logger.debug(
                     'Implicit creation of event manager will implicitly set configuration as side effect. '
                     'It is advised to explicitly first set the configuration instead.'
                 )
@@ -93,7 +93,7 @@ class ServiceLocator:
     def get_storage_client(self) -> StorageClient:
         """Get the storage client."""
         if self._storage_client is None:
-            logger.warning('No storage client set, implicitly creating and using default FileSystemStorageClient.')
+            logger.debug('No storage client set, implicitly creating and using default FileSystemStorageClient.')
             if self._configuration is None:
                 logger.warning(
                     'Implicit creation of storage client will implicitly set configuration as side effect. '

crawlee/_types.py

@@ -3,17 +3,7 @@ from __future__ import annotations
 import dataclasses
 from collections.abc import Callable, Iterator, Mapping
 from dataclasses import dataclass
-from typing import (
-    TYPE_CHECKING,
-    Annotated,
-    Any,
-    Literal,
-    Protocol,
-    TypedDict,
-    TypeVar,
-    cast,
-    overload,
-)
+from typing import TYPE_CHECKING, Annotated, Any, Literal, Protocol, TypedDict, TypeVar, cast, overload
 
 from pydantic import ConfigDict, Field, PlainValidator, RootModel
 
@@ -71,11 +61,15 @@ class HttpHeaders(RootModel, Mapping[str, str]):
 
     model_config = ConfigDict(validate_by_name=True, validate_by_alias=True)
 
-    root: Annotated[
-        dict[str, str],
-        PlainValidator(lambda value: _normalize_headers(value)),
-        Field(default_factory=dict),
-    ] = {}
+    # Workaround for pydantic 2.12 and mypy type checking issue for Annotated with default_factory
+    if TYPE_CHECKING:
+        root: dict[str, str] = {}
+    else:
+        root: Annotated[
+            dict[str, str],
+            PlainValidator(lambda value: _normalize_headers(value)),
+            Field(default_factory=dict),
+        ]
 
     def __getitem__(self, key: str) -> str:
         return self.root[key.lower()]

crawlee/_utils/recoverable_state.py

@@ -4,12 +4,14 @@ from typing import TYPE_CHECKING, Generic, Literal, TypeVar
 
 from pydantic import BaseModel
 
+from crawlee._utils.raise_if_too_many_kwargs import raise_if_too_many_kwargs
 from crawlee.events._types import Event, EventPersistStateData
 
 if TYPE_CHECKING:
     import logging
+    from collections.abc import Callable, Coroutine
 
-    from crawlee.storages._key_value_store import KeyValueStore
+    from crawlee.storages import KeyValueStore
 
 TStateModel = TypeVar('TStateModel', bound=BaseModel)
 
@@ -37,6 +39,7 @@ class RecoverableState(Generic[TStateModel]):
         persistence_enabled: Literal[True, False, 'explicit_only'] = False,
         persist_state_kvs_name: str | None = None,
         persist_state_kvs_id: str | None = None,
+        persist_state_kvs_factory: Callable[[], Coroutine[None, None, KeyValueStore]] | None = None,
         logger: logging.Logger,
     ) -> None:
         """Initialize a new recoverable state object.
@@ -51,16 +54,40 @@ class RecoverableState(Generic[TStateModel]):
                 If neither a name nor and id are supplied, the default store will be used.
             persist_state_kvs_id: The identifier of the KeyValueStore to use for persistence.
                 If neither a name nor and id are supplied, the default store will be used.
+            persist_state_kvs_factory: Factory that can be awaited to create KeyValueStore to use for persistence. If
+                not provided, a system-wide KeyValueStore will be used, based on service locator configuration.
             logger: A logger instance for logging operations related to state persistence
         """
+        raise_if_too_many_kwargs(
+            persist_state_kvs_name=persist_state_kvs_name,
+            persist_state_kvs_id=persist_state_kvs_id,
+            persist_state_kvs_factory=persist_state_kvs_factory,
+        )
+        if not persist_state_kvs_factory:
+            logger.debug(
+                'No explicit key_value_store set for recoverable state. Recovery will use a system-wide KeyValueStore '
+                'based on service_locator configuration, potentially calling service_locator.set_storage_client in the '
+                'process. It is recommended to initialize RecoverableState with explicit key_value_store to avoid '
+                'global side effects.'
+            )
+
         self._default_state = default_state
         self._state_type: type[TStateModel] = self._default_state.__class__
         self._state: TStateModel | None = None
         self._persistence_enabled = persistence_enabled
         self._persist_state_key = persist_state_key
-        self._persist_state_kvs_name = persist_state_kvs_name
-        self._persist_state_kvs_id = persist_state_kvs_id
-        self._key_value_store: 'KeyValueStore | None' = None  # noqa: UP037
+        if persist_state_kvs_factory is None:
+
+            async def kvs_factory() -> KeyValueStore:
+                from crawlee.storages import KeyValueStore  # noqa: PLC0415 avoid circular import
+
+                return await KeyValueStore.open(name=persist_state_kvs_name, id=persist_state_kvs_id)
+
+            self._persist_state_kvs_factory = kvs_factory
+        else:
+            self._persist_state_kvs_factory = persist_state_kvs_factory
+
+        self._key_value_store: KeyValueStore | None = None
         self._log = logger
 
     async def initialize(self) -> TStateModel:
@@ -77,11 +104,8 @@ class RecoverableState(Generic[TStateModel]):
             return self.current_value
 
         # Import here to avoid circular imports.
-        from crawlee.storages._key_value_store import KeyValueStore  # noqa: PLC0415
 
-        self._key_value_store = await KeyValueStore.open(
-            name=self._persist_state_kvs_name, id=self._persist_state_kvs_id
-        )
+        self._key_value_store = await self._persist_state_kvs_factory()
 
         await self._load_saved_state()
 

crawlee/_utils/recurring_task.py

@@ -7,6 +7,9 @@ from typing import TYPE_CHECKING
 if TYPE_CHECKING:
     from collections.abc import Callable
     from datetime import timedelta
+    from types import TracebackType
+
+    from typing_extensions import Self
 
 logger = getLogger(__name__)
 
@@ -26,6 +29,18 @@ class RecurringTask:
         self.delay = delay
         self.task: asyncio.Task | None = None
 
+    async def __aenter__(self) -> Self:
+        self.start()
+        return self
+
+    async def __aexit__(
+        self,
+        exc_type: type[BaseException] | None,
+        exc_value: BaseException | None,
+        exc_traceback: TracebackType | None,
+    ) -> None:
+        await self.stop()
+
     async def _wrapper(self) -> None:
         """Continuously execute the provided function with the specified delay.
 

crawlee/_utils/robots.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+from logging import getLogger
 from typing import TYPE_CHECKING
 
 from protego import Protego
@@ -15,6 +16,9 @@ if TYPE_CHECKING:
     from crawlee.proxy_configuration import ProxyInfo
 
 
+logger = getLogger(__name__)
+
+
 class RobotsTxtFile:
     def __init__(
         self, url: str, robots: Protego, http_client: HttpClient | None = None, proxy_info: ProxyInfo | None = None
@@ -56,12 +60,20 @@ class RobotsTxtFile:
             http_client: The `HttpClient` instance used to perform the network request for fetching the robots.txt file.
             proxy_info: Optional `ProxyInfo` to be used when fetching the robots.txt file. If None, no proxy is used.
         """
-        response = await http_client.send_request(url, proxy_info=proxy_info)
-        body = (
-            b'User-agent: *\nAllow: /' if is_status_code_client_error(response.status_code) else await response.read()
-        )
+        try:
+            response = await http_client.send_request(url, proxy_info=proxy_info)
+
+            body = (
+                b'User-agent: *\nAllow: /'
+                if is_status_code_client_error(response.status_code)
+                else await response.read()
+            )
+            robots = Protego.parse(body.decode('utf-8'))
+
+        except Exception as e:
+            logger.warning(f'Failed to fetch from robots.txt from "{url}" with error: "{e}"')
 
-        robots = Protego.parse(body.decode('utf-8'))
+            robots = Protego.parse('User-agent: *\nAllow: /')
 
         return cls(url, robots, http_client=http_client, proxy_info=proxy_info)
 

crawlee/_utils/sitemap.py

@@ -335,7 +335,7 @@ async def _fetch_and_process_sitemap(
                         # Check if the first chunk is a valid gzip header
                         if first_chunk and raw_chunk.startswith(b'\x1f\x8b'):
                             decompressor = zlib.decompressobj(zlib.MAX_WBITS | 16)
-                            first_chunk = False
+                        first_chunk = False
 
                         chunk = decompressor.decompress(raw_chunk) if decompressor else raw_chunk
                         text_chunk = decoder.decode(chunk)

crawlee/_utils/urls.py

@@ -7,6 +7,7 @@ from yarl import URL
 
 if TYPE_CHECKING:
     from collections.abc import Iterator
+    from logging import Logger
 
 
 def is_url_absolute(url: str) -> bool:
@@ -22,13 +23,19 @@ def convert_to_absolute_url(base_url: str, relative_url: str) -> str:
     return str(URL(base_url).join(URL(relative_url)))
 
 
-def to_absolute_url_iterator(base_url: str, urls: Iterator[str]) -> Iterator[str]:
+def to_absolute_url_iterator(base_url: str, urls: Iterator[str], logger: Logger | None = None) -> Iterator[str]:
     """Convert an iterator of relative URLs to absolute URLs using a base URL."""
     for url in urls:
         if is_url_absolute(url):
             yield url
         else:
-            yield convert_to_absolute_url(base_url, url)
+            converted_url = convert_to_absolute_url(base_url, url)
+            # Skip the URL if conversion fails, probably due to an incorrect format, such as 'mailto:'.
+            if not is_url_absolute(converted_url):
+                if logger:
+                    logger.debug(f'Could not convert URL "{url}" to absolute using base URL "{base_url}". Skipping it.')
+                continue
+            yield converted_url
 
 
 _http_url_adapter = TypeAdapter(AnyHttpUrl)

crawlee/browsers/_browser_pool.py

@@ -118,7 +118,10 @@ class BrowserPool:
         """Initialize a new instance with a single `PlaywrightBrowserPlugin` configured with the provided options.
 
         Args:
-            browser_type: The type of browser to launch ('chromium', 'firefox', or 'webkit').
+            browser_type: The type of browser to launch:
+                - 'chromium', 'firefox', 'webkit': Use Playwright-managed browsers
+                - 'chrome': Use your locally installed Google Chrome browser. Requires Google Chrome to be installed on
+                    the system.
             user_data_dir: Path to a user data directory, which stores browser session data like cookies
                 and local storage.
             browser_launch_options: Keyword arguments to pass to the browser launch method. These options are provided

crawlee/browsers/_playwright_browser_controller.py

@@ -216,7 +216,7 @@ class PlaywrightBrowserController(BrowserController):
         browser_new_context_options = dict(browser_new_context_options) if browser_new_context_options else {}
         if proxy_info:
             if browser_new_context_options.get('proxy'):
-                logger.warning("browser_new_context_options['proxy'] overriden by explicit `proxy_info` argument.")
+                logger.warning("browser_new_context_options['proxy'] overridden by explicit `proxy_info` argument.")
 
             browser_new_context_options['proxy'] = ProxySettings(
                 server=f'{proxy_info.scheme}://{proxy_info.hostname}:{proxy_info.port}',

crawlee/browsers/_playwright_browser_plugin.py

@@ -34,8 +34,8 @@ class PlaywrightBrowserPlugin(BrowserPlugin):
 
     It is a plugin designed to manage browser instances using the Playwright automation library. It acts as a factory
     for creating new browser instances and provides a unified interface for interacting with different browser types
-    (chromium, firefox, and webkit). This class integrates configuration options for browser launches (headless mode,
-    executable paths, sandboxing, ...). It also manages browser contexts and the number of pages open within each
+    (chromium, firefox, webkit and chrome). This class integrates configuration options for browser launches (headless
+    mode, executable paths, sandboxing, ...). It also manages browser contexts and the number of pages open within each
     browser instance, ensuring that resource limits are respected.
     """
 
@@ -55,7 +55,10 @@ class PlaywrightBrowserPlugin(BrowserPlugin):
         """Initialize a new instance.
 
         Args:
-            browser_type: The type of browser to launch ('chromium', 'firefox', or 'webkit').
+            browser_type: The type of browser to launch:
+                - 'chromium', 'firefox', 'webkit': Use Playwright-managed browsers
+                - 'chrome': Use your locally installed Google Chrome browser. Requires Google Chrome to be installed on
+                    the system.
             user_data_dir: Path to a User Data Directory, which stores browser session data like cookies and local
                 storage.
             browser_launch_options: Keyword arguments to pass to the browser launch method. These options are provided
@@ -80,6 +83,17 @@ class PlaywrightBrowserPlugin(BrowserPlugin):
             'chromium_sandbox': not config.disable_browser_sandbox,
         }
 
+        if browser_type == 'chrome' and default_launch_browser_options['executable_path']:
+            raise ValueError(
+                'Cannot use browser_type `chrome` with `Configuration.default_browser_path` or `executable_path` set.'
+            )
+
+        # Map 'chrome' to 'chromium' with the 'chrome' channel.
+        if browser_type == 'chrome':
+            browser_type = 'chromium'
+            # Chromium parameter 'channel' set to 'chrome' enables using installed Google Chrome.
+            default_launch_browser_options['channel'] = 'chrome'
+
         self._browser_type: BrowserType = browser_type
         self._browser_launch_options: dict[str, Any] = default_launch_browser_options | (browser_launch_options or {})
         self._browser_new_context_options = browser_new_context_options or {}

crawlee/browsers/_types.py

@@ -6,7 +6,7 @@ from typing import TYPE_CHECKING, Literal
 if TYPE_CHECKING:
     from playwright.async_api import Page
 
-BrowserType = Literal['chromium', 'firefox', 'webkit']
+BrowserType = Literal['chromium', 'firefox', 'webkit', 'chrome']
 
 
 @dataclass

crawlee/configuration.py

@@ -28,7 +28,9 @@ class Configuration(BaseSettings):
     Settings can also be configured via environment variables, prefixed with `CRAWLEE_`.
     """
 
-    model_config = SettingsConfigDict(validate_by_name=True, validate_by_alias=True)
+    # TODO: https://github.com/pydantic/pydantic-settings/issues/706
+    # Use `SettingsConfigDict(validate_by_name=True, validate_by_alias=True)` when issue is resolved.
+    model_config = SettingsConfigDict(populate_by_name=True)
 
     internal_timeout: Annotated[timedelta | None, Field(alias='crawlee_internal_timeout')] = None
     """Timeout for the internal asynchronous operations."""

crawlee/crawlers/_abstract_http/_abstract_http_crawler.py

@@ -167,7 +167,9 @@ class AbstractHttpCrawler(
             kwargs.setdefault('strategy', 'same-hostname')
 
             links_iterator: Iterator[str] = iter(self._parser.find_links(parsed_content, selector=selector))
-            links_iterator = to_absolute_url_iterator(context.request.loaded_url or context.request.url, links_iterator)
+            links_iterator = to_absolute_url_iterator(
+                context.request.loaded_url or context.request.url, links_iterator, logger=context.log
+            )
 
             if robots_txt_file:
                 skipped, links_iterator = partition(lambda url: robots_txt_file.is_allowed(url), links_iterator)

crawlee/crawlers/_adaptive_playwright/_adaptive_playwright_crawler.py

@@ -71,7 +71,6 @@ class _NonPersistentStatistics(Statistics):
     async def __aenter__(self) -> Self:
         self._active = True
         await self._state.initialize()
-        self._after_initialize()
         return self
 
     async def __aexit__(
@@ -149,10 +148,6 @@ class AdaptivePlaywrightCrawler(
                 non-default configuration.
             kwargs: Additional keyword arguments to pass to the underlying `BasicCrawler`.
         """
-        # Some sub crawler kwargs are internally modified. Prepare copies.
-        basic_crawler_kwargs_for_static_crawler = deepcopy(kwargs)
-        basic_crawler_kwargs_for_pw_crawler = deepcopy(kwargs)
-
         # Adaptive crawling related.
         self.rendering_type_predictor = rendering_type_predictor or DefaultRenderingTypePredictor()
         self.result_checker = result_checker or (lambda _: True)
@@ -170,11 +165,11 @@ class AdaptivePlaywrightCrawler(
         # Each sub crawler will use custom logger .
         static_logger = getLogger('Subcrawler_static')
         static_logger.setLevel(logging.ERROR)
-        basic_crawler_kwargs_for_static_crawler['_logger'] = static_logger
+        basic_crawler_kwargs_for_static_crawler: _BasicCrawlerOptions = {'_logger': static_logger, **kwargs}
 
         pw_logger = getLogger('Subcrawler_playwright')
         pw_logger.setLevel(logging.ERROR)
-        basic_crawler_kwargs_for_pw_crawler['_logger'] = pw_logger
+        basic_crawler_kwargs_for_pw_crawler: _BasicCrawlerOptions = {'_logger': pw_logger, **kwargs}
 
         # Initialize sub crawlers to create their pipelines.
         static_crawler_class = AbstractHttpCrawler.create_parsed_http_crawler_class(static_parser=static_parser)
@@ -319,7 +314,7 @@ class AdaptivePlaywrightCrawler(
                 ),
                 logger=self._logger,
             )
-            return SubCrawlerRun(result=result)
+            return SubCrawlerRun(result=result, run_context=context_linked_to_result)
         except Exception as e:
             return SubCrawlerRun(exception=e)
 
@@ -375,7 +370,8 @@ class AdaptivePlaywrightCrawler(
                 self.track_http_only_request_handler_runs()
 
                 static_run = await self._crawl_one(rendering_type='static', context=context)
-                if static_run.result and self.result_checker(static_run.result):
+                if static_run.result and static_run.run_context and self.result_checker(static_run.result):
+                    self._update_context_from_copy(context, static_run.run_context)
                     self._context_result_map[context] = static_run.result
                     return
                 if static_run.exception:
@@ -406,13 +402,10 @@ class AdaptivePlaywrightCrawler(
         if pw_run.exception is not None:
             raise pw_run.exception
 
-        if pw_run.result:
-            self._context_result_map[context] = pw_run.result
-
+        if pw_run.result and pw_run.run_context:
             if should_detect_rendering_type:
                 detection_result: RenderingType
                 static_run = await self._crawl_one('static', context=context, state=old_state_copy)
-
                 if static_run.result and self.result_comparator(static_run.result, pw_run.result):
                     detection_result = 'static'
                 else:
@@ -421,6 +414,9 @@ class AdaptivePlaywrightCrawler(
                 context.log.debug(f'Detected rendering type {detection_result} for {context.request.url}')
                 self.rendering_type_predictor.store_result(context.request, detection_result)
 
+            self._update_context_from_copy(context, pw_run.run_context)
+            self._context_result_map[context] = pw_run.result
+
     def pre_navigation_hook(
         self,
         hook: Callable[[AdaptivePlaywrightPreNavCrawlingContext], Awaitable[None]] | None = None,
@@ -455,8 +451,32 @@ class AdaptivePlaywrightCrawler(
     def track_rendering_type_mispredictions(self) -> None:
         self.statistics.state.rendering_type_mispredictions += 1
 
+    def _update_context_from_copy(self, context: BasicCrawlingContext, context_copy: BasicCrawlingContext) -> None:
+        """Update mutable fields of `context` from `context_copy`.
+
+        Uses object.__setattr__ to bypass frozen dataclass restrictions,
+        allowing state synchronization after isolated crawler execution.
+        """
+        updating_attributes = {
+            'request': ('headers', 'user_data'),
+            'session': ('_user_data', '_usage_count', '_error_score', '_cookies'),
+        }
+
+        for attr, sub_attrs in updating_attributes.items():
+            original_sub_obj = getattr(context, attr)
+            copy_sub_obj = getattr(context_copy, attr)
+
+            # Check that both sub objects are not None
+            if original_sub_obj is None or copy_sub_obj is None:
+                continue
+
+            for sub_attr in sub_attrs:
+                new_value = getattr(copy_sub_obj, sub_attr)
+                object.__setattr__(original_sub_obj, sub_attr, new_value)
+
 
 @dataclass(frozen=True)
 class SubCrawlerRun:
     result: RequestHandlerRunResult | None = None
     exception: Exception | None = None
+    run_context: BasicCrawlingContext | None = None

crawlee/crawlers/_basic/_basic_crawler.py

@@ -56,7 +56,7 @@ from crawlee.errors import (
     SessionError,
     UserDefinedErrorHandlerError,
 )
-from crawlee.events._types import Event, EventCrawlerStatusData
+from crawlee.events._types import Event, EventCrawlerStatusData, EventPersistStateData
 from crawlee.http_clients import ImpitHttpClient
 from crawlee.router import Router
 from crawlee.sessions import SessionPool
@@ -437,14 +437,23 @@ class BasicCrawler(Generic[TCrawlingContext, TStatisticsState]):
         self._statistics_log_format = statistics_log_format
 
         # Statistics
-        self._statistics = statistics or cast(
-            'Statistics[TStatisticsState]',
-            Statistics.with_default_state(
-                periodic_message_logger=self._logger,
-                statistics_log_format=self._statistics_log_format,
-                log_message='Current request statistics:',
-            ),
-        )
+        if statistics:
+            self._statistics = statistics
+        else:
+
+            async def persist_state_factory() -> KeyValueStore:
+                return await self.get_key_value_store()
+
+            self._statistics = cast(
+                'Statistics[TStatisticsState]',
+                Statistics.with_default_state(
+                    persistence_enabled=True,
+                    periodic_message_logger=self._logger,
+                    statistics_log_format=self._statistics_log_format,
+                    log_message='Current request statistics:',
+                    persist_state_kvs_factory=persist_state_factory,
+                ),
+            )
 
         # Additional context managers to enter and exit
         self._additional_context_managers = _additional_context_managers or []
@@ -689,7 +698,6 @@ class BasicCrawler(Generic[TCrawlingContext, TStatisticsState]):
         except CancelledError:
             pass
         finally:
-            await self._crawler_state_rec_task.stop()
             if threading.current_thread() is threading.main_thread():
                 with suppress(NotImplementedError):
                     asyncio.get_running_loop().remove_signal_handler(signal.SIGINT)
@@ -721,8 +729,6 @@ class BasicCrawler(Generic[TCrawlingContext, TStatisticsState]):
     async def _run_crawler(self) -> None:
         event_manager = self._service_locator.get_event_manager()
 
-        self._crawler_state_rec_task.start()
-
         # Collect the context managers to be entered. Context managers that are already active are excluded,
         # as they were likely entered by the caller, who will also be responsible for exiting them.
         contexts_to_enter = [
@@ -733,6 +739,7 @@ class BasicCrawler(Generic[TCrawlingContext, TStatisticsState]):
                 self._statistics,
                 self._session_pool if self._use_session_pool else None,
                 self._http_client,
+                self._crawler_state_rec_task,
                 *self._additional_context_managers,
             )
             if cm and getattr(cm, 'active', False) is False
@@ -744,6 +751,9 @@ class BasicCrawler(Generic[TCrawlingContext, TStatisticsState]):
 
             await self._autoscaled_pool.run()
 
+            # Emit PERSIST_STATE event when crawler is finishing to allow listeners to persist their state if needed
+            event_manager.emit(event=Event.PERSIST_STATE, event_data=EventPersistStateData(is_migrating=False))
+
     async def add_requests(
         self,
         requests: Sequence[str | Request],
@@ -972,6 +982,7 @@ class BasicCrawler(Generic[TCrawlingContext, TStatisticsState]):
                         label=label,
                         user_data=user_data,
                         transform_request_function=transform_request_function,
+                        **kwargs,
                     ),
                     rq_id=rq_id,
                     rq_name=rq_name,