apify 2.5.0b3__tar.gz → 2.5.0b5__tar.gz

{apify-2.5.0b3 → apify-2.5.0b5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: apify
-Version: 2.5.0b3
+Version: 2.5.0b5
 Summary: Apify SDK for Python
 Project-URL: Homepage, https://docs.apify.com/sdk/python/
 Project-URL: Apify homepage, https://apify.com

{apify-2.5.0b3 → apify-2.5.0b5}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "apify"
-version = "2.5.0b3"
+version = "2.5.0b5"
 description = "Apify SDK for Python"
 authors = [{ name = "Apify Technologies s.r.o.", email = "support@apify.com" }]
 license = { file = "LICENSE" }
@@ -63,14 +63,14 @@ dev = [
     "mypy~=1.15.0",
     "pre-commit~=4.2.0",
     "pydoc-markdown~=4.8.0",
-    "pytest-asyncio~=0.25.0",
+    "pytest-asyncio~=0.26.0",
     "pytest-cov~=6.0.0",
     "pytest-only~=2.1.0",
     "pytest-xdist~=3.6.0",
     "pytest~=8.3.0",
     "respx~=0.22.0",
     "ruff~=0.11.0",
-    "setuptools~=77.0.1",     # setuptools are used by pytest but not explicitly required
+    "setuptools~=78.1.0",     # setuptools are used by pytest but not explicitly required
 ]
 
 [tool.hatch.build.targets.wheel]
@@ -91,6 +91,9 @@ ignore = [
     "D100",     # Missing docstring in public module
     "D104",     # Missing docstring in public package
     "D107",     # Missing docstring in `__init__`
+    "D203",     # One blank line required before class docstring
+    "D213",     # Multi-line docstring summary should start at the second line
+    "D413",     # Missing blank line after last section
     "EM",       # flake8-errmsg
     "G004",     # Logging statement uses f-string
     "ISC001",   # This rule may cause conflicts when used with the formatter
@@ -162,9 +165,6 @@ runtime-evaluated-base-classes = [
 [tool.ruff.lint.flake8-builtins]
 builtins-ignorelist = ["id"]
 
-[tool.ruff.lint.pydocstyle]
-convention = "google"
-
 [tool.ruff.lint.isort]
 known-local-folder = ["apify"]
 known-first-party = ["apify_client", "apify_shared", "crawlee"]

{apify-2.5.0b3 → apify-2.5.0b5}/src/apify/_actor.py

@@ -1138,7 +1138,7 @@ class _ActorType:
         return proxy_configuration
 
     def _get_default_exit_process(self) -> bool:
-        """Returns False for IPython, Pytest, and Scrapy environments, True otherwise."""
+        """Return 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

{apify-2.5.0b3 → apify-2.5.0b5}/src/apify/_configuration.py

@@ -5,8 +5,8 @@ from decimal import Decimal
 from logging import getLogger
 from typing import Annotated, Any
 
-from pydantic import AliasChoices, BeforeValidator, Field
-from typing_extensions import deprecated
+from pydantic import AliasChoices, BeforeValidator, Field, model_validator
+from typing_extensions import Self, deprecated
 
 from crawlee._utils.models import timedelta_ms
 from crawlee._utils.urls import validate_http_url
@@ -365,6 +365,19 @@ class Configuration(CrawleeConfiguration):
         ),
     ] = None
 
+    @model_validator(mode='after')
+    def disable_browser_sandbox_on_platform(self) -> Self:
+        """Disable the browser sandbox mode when running on the Apify platform.
+
+        Running in environment where `is_at_home` is True does not benefit from browser sandbox as it is already running
+        in a container. It can be on the contrary undesired as the process in the container might be running as root and
+        this will crash chromium that was started with browser sandbox mode.
+        """
+        if self.is_at_home and not self.disable_browser_sandbox:
+            self.disable_browser_sandbox = True
+            logger.warning('Actor is running on the Apify platform, `disable_browser_sandbox` was changed to True.')
+        return self
+
     @classmethod
     def get_global_configuration(cls) -> Configuration:
         """Retrieve the global instance of the configuration.

{apify-2.5.0b3 → apify-2.5.0b5}/src/apify/_platform_event_manager.py

@@ -48,11 +48,11 @@ class SystemInfoEventData(BaseModel):
     is_cpu_overloaded: Annotated[bool, Field(alias='isCpuOverloaded')]
     created_at: Annotated[datetime, Field(alias='createdAt')]
 
-    def to_crawlee_format(self) -> EventSystemInfoData:
+    def to_crawlee_format(self, dedicated_cpus: float) -> EventSystemInfoData:
         return EventSystemInfoData.model_validate(
             {
                 'cpu_info': {
-                    'used_ratio': self.cpu_current_usage / 100,
+                    'used_ratio': (self.cpu_current_usage / 100) / dedicated_cpus,
                     'created_at': self.created_at,
                 },
                 'memory_info': {
@@ -218,7 +218,7 @@ class PlatformEventManager(EventManager):
                             event=parsed_message.name,
                             event_data=parsed_message.data
                             if not isinstance(parsed_message.data, SystemInfoEventData)
-                            else parsed_message.data.to_crawlee_format(),
+                            else parsed_message.data.to_crawlee_format(self._config.dedicated_cpus or 1),
                         )
 
                         if parsed_message.name == Event.MIGRATING:

{apify-2.5.0b3 → apify-2.5.0b5}/src/apify/_utils.py

@@ -31,10 +31,16 @@ GroupName = Literal['Classes', 'Abstract classes', 'Interfaces', 'Data structure
 
 
 def docs_group(group_name: GroupName) -> Callable:  # noqa: ARG001
-    """Decorator to mark symbols for rendering and grouping in documentation.
+    """Mark a symbol for rendering and grouping in documentation.
 
-    This decorator is used purely for documentation purposes and does not alter the behavior
+    This decorator is used solely for documentation purposes and does not modify the behavior
     of the decorated callable.
+
+    Args:
+        group_name: The documentation group to which the symbol belongs.
+
+    Returns:
+        The original callable without modification.
     """
 
     def wrapper(func: Callable) -> Callable:
@@ -44,12 +50,16 @@ def docs_group(group_name: GroupName) -> Callable:  # noqa: ARG001
 
 
 def docs_name(symbol_name: str) -> Callable:  # noqa: ARG001
-    """Decorator for renaming symbols in documentation.
+    """Rename a symbol for documentation rendering.
 
-    This changes the rendered name of the symbol only in the rendered web documentation.
+    This decorator modifies only the displayed name of the symbol in the generated documentation
+    and does not affect its runtime behavior.
 
-    This decorator is used purely for documentation purposes and does not alter the behavior
-    of the decorated callable.
+    Args:
+        symbol_name: The name to be used in the documentation.
+
+    Returns:
+        The original callable without modification.
     """
 
     def wrapper(func: Callable) -> Callable:

{apify-2.5.0b3 → apify-2.5.0b5}/src/apify/scrapy/_async_thread.py

@@ -113,7 +113,7 @@ class AsyncThread:
         await asyncio.gather(*tasks, return_exceptions=True)
 
     def _force_exit_event_loop(self) -> None:
-        """Forcefully shut down the event loop and its thread."""
+        """Shut down the event loop and its thread forcefully."""
         try:
             logger.info('Forced shutdown of the event loop and its thread...')
             self._eventloop.call_soon_threadsafe(self._eventloop.stop)

{apify-2.5.0b3 → apify-2.5.0b5}/src/apify/storages/_request_list.py

@@ -51,7 +51,7 @@ class RequestList(CrawleeRequestList):
         request_list_sources_input: list[dict[str, Any]] | None = None,
         http_client: HttpClient | None = None,
     ) -> RequestList:
-        """Creates RequestList from Actor input requestListSources.
+        """Initialize a new instance from request list source input.
 
         Args:
             name: Name of the returned RequestList.
@@ -108,9 +108,10 @@ class RequestList(CrawleeRequestList):
 
     @staticmethod
     async def _fetch_requests_from_url(
-        remote_url_requests_inputs: list[_RequestsFromUrlInput], http_client: HttpClient
+        remote_url_requests_inputs: list[_RequestsFromUrlInput],
+        http_client: HttpClient,
     ) -> list[Request]:
-        """Crete list of requests from url.
+        """Create list of requests from url.
 
         Send GET requests to urls defined in each requests_from_url of remote_url_requests_inputs. Run extracting
         callback on each response body and use URL_NO_COMMAS_REGEX regex to find all links. Create list of Requests from
@@ -119,7 +120,11 @@ class RequestList(CrawleeRequestList):
         created_requests: list[Request] = []
 
         def create_requests_from_response(request_input: _RequestsFromUrlInput, task: Task) -> None:
-            """Callback to scrape response body with regexp and create Requests from matches."""
+            """Extract links from response body and use them to create `Request` objects.
+
+            Use the regular expression to find all matching links in the response body, then create `Request`
+            objects from these links and the provided input attributes.
+            """
             matches = re.finditer(URL_NO_COMMAS_REGEX, task.result().read().decode('utf-8'))
             created_requests.extend(
                 [

{apify-2.5.0b3 → apify-2.5.0b5}/tests/integration/test_actor_scrapy.py

@@ -41,7 +41,8 @@ async def test_actor_scrapy_title_spider(
 
     items = await actor.last_run().dataset().list_items()
 
-    assert items.count >= 10
+    # CLOSESPIDER_PAGECOUNT is set to 10 in the spider settings.
+    assert items.count >= 9
 
     for item in items.items:
         assert 'url' in item

apify-2.5.0b5/tests/unit/actor/test_configuration.py

@@ -0,0 +1,21 @@
+import pytest
+
+from apify import Configuration
+
+
+@pytest.mark.parametrize(
+    ('is_at_home', 'disable_browser_sandbox_in', 'disable_browser_sandbox_out'),
+    [
+        (False, False, False),
+        (False, True, True),
+        (True, False, True),
+        (True, True, True),
+    ],
+)
+def test_disable_browser_sandbox(
+    *, is_at_home: bool, disable_browser_sandbox_in: bool, disable_browser_sandbox_out: bool
+) -> None:
+    assert (
+        Configuration(is_at_home=is_at_home, disable_browser_sandbox=disable_browser_sandbox_in).disable_browser_sandbox
+        == disable_browser_sandbox_out
+    )