Improve handling of failed scan tasks and allow reconnect#332
Merged
Improve handling of failed scan tasks and allow reconnect#332
Conversation
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
📝 WalkthroughWalkthroughControllers now construct and expose a ControllerAPI tree and provide create_api_and_tasks() returning (ControllerAPI, periodic_coros, initial_coros). Transport-layer ControllerAPI was removed; modules now import ControllerAPI from fastcs.controllers. Added Changes
Sequence DiagramsequenceDiagram
participant CS as ControlSystem
participant C as Controller
participant API as ControllerAPI
participant Scheduler as Scheduler
participant Attr as AttrR
CS->>C: create_api_and_tasks()
C->>C: _build_api(path)
C->>API: construct ControllerAPI tree
C-->>CS: (ControllerAPI, periodic_coros, initial_coros)
CS->>CS: await initial_coros
CS->>Scheduler: start periodic_coros
loop periodic
Scheduler->>C: check _connected
alt connected
Scheduler->>C: run scan methods
C->>Attr: trigger attribute update callbacks (await coroutine)
else disconnected
Scheduler-->>Scheduler: sleep/backoff until reconnect
end
end
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (4)
tests/conftest.py (1)
50-52: Consider using the public API pathway if available.The fixture now calls the private
_build_api([])method directly. While acceptable for test infrastructure, relying on private methods can make tests fragile. If a publiccreate_api_and_tasks()or similar method is intended for external use, consider whether the fixture should use that instead for better alignment with the public contract.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@tests/conftest.py` around lines 50 - 52, The fixture controller_api currently calls the private helper controller._build_api([]); change it to use the controller's public factory method (e.g. create_api_and_tasks or another documented public API-building method) if one exists so tests exercise the public contract instead of private internals; locate the controller_api fixture and replace controller._build_api([]) with the appropriate public call (passing equivalent arguments) to construct the API and tasks.tests/test_controllers.py (1)
260-283: Consider usingconnect()instead of directly setting_connected, and improve task cleanup.
Line 271: Prefer calling
await controller.connect()instead of manually settingcontroller._connected = True. This better reflects the intended usage pattern and tests the public API.Lines 282-283: The task cancellation should handle
CancelledErrorto avoid potential issues:♻️ Proposed improvements
`@pytest.mark.asyncio` async def test_scan_exception_sets_disconnected_and_reconnect_resumes(): class MyTestController(Controller): `@scan`(0.01) async def failing_scan(self): raise RuntimeError("scan error") controller = MyTestController() controller.post_initialise() _, scan_coros, _ = controller.create_api_and_tasks() - controller._connected = True + await controller.connect() task = asyncio.create_task(scan_coros[0]()) # Wait long enough for the scan to run and raise, setting _connected = False await asyncio.sleep(0.1) assert not controller._connected # Trigger reconnect - _connected resumes scan tasks await controller.reconnect() assert controller._connected task.cancel() - await asyncio.sleep(0.01) + try: + await task + except asyncio.CancelledError: + pass🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@tests/test_controllers.py` around lines 260 - 283, The test sets controller state directly and doesn't properly await task cancellation; change the setup to call and await controller.connect() (instead of setting controller._connected = True) so the public API is exercised (see MyTestController, connect, reconnect, scan decorator and scan_coros usage), and when cancelling the background task created from scan_coros[0]() ensure you await it and handle asyncio.CancelledError to cleanly swallow cancellation (i.e., cancel the task, await it inside a try/except catching CancelledError) so the test teardown is robust.src/fastcs/controllers/controller.py (1)
55-57: Make basedisconnect()clear connection state.Base
disconnect()currently leaves_connectedunchanged. Resetting it toFalsemakes lifecycle semantics safer for subclasses that don’t overridedisconnect().🔧 Suggested change
async def disconnect(self) -> None: """Hook to tidy up resources before stopping the application""" - pass + self._connected = False🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/fastcs/controllers/controller.py` around lines 55 - 57, The base async disconnect() leaves the controller's connection flag unchanged; update Controller.disconnect() to set self._connected = False to reliably clear connection state for subclasses that do not override disconnect(), ensuring lifecycle semantics are correct (modify the async def disconnect(self) method in the Controller class to assign self._connected = False before returning).tests/assertable_controller.py (1)
79-79: Prefer public API construction in this test helper.Line 79 uses
controller._build_api([]), which binds tests to a private method. Usingcreate_api_and_tasks()keeps this helper aligned with supported API boundaries.♻️ Suggested change
- controller_api = controller._build_api([]) + controller_api, _, _ = controller.create_api_and_tasks()🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@tests/assertable_controller.py` at line 79, The test helper currently calls the private method controller._build_api([]); replace that with the public factory create_api_and_tasks() to avoid depending on a private symbol. Call create_api_and_tasks() (or controller.create_api_and_tasks() if it is a method) and unpack its return (e.g., api, tasks = create_api_and_tasks(...)) or assign the returned API to controller_api so the helper uses the supported public API surface instead of _build_api.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/fastcs/connections/ip_connection.py`:
- Around line 86-92: The current close sequence only clears self.__connection
when connection.close() succeeds or raises ConnectionResetError; move the
clearing into a finally block so self.__connection is set to None regardless of
what exception close() raises. Concretely, wrap the await connection.close()
call in try/except/finally (or use try/finally) inside the async with
self._connection context, catch/handle specific ConnectionResetError if desired,
and always assign self.__connection = None in the finally so the stale handle is
never retained; reference the symbols self._connection, connection.close(), and
self.__connection to locate and update the code.
In `@src/fastcs/control_system.py`:
- Around line 118-121: The startup currently calls self._run_initial_coros()
unconditionally after await self._controller.reconnect(), which means if
reconnect() returns without raising but leaves the connection flag
(self._connected or self._controller.connected) False those ONCE coroutines will
run while disconnected and never be retried; change the logic so
_run_initial_coros() only runs when a successful connection exists (check
self._connected or self._controller.connected) and otherwise mark the initial
coroutines as pending (e.g., set an attribute like self._initial_coros_pending =
True) and invoke _run_initial_coros() from the reconnect-success path (or the
reconnect callback/handler) when connection becomes True, leaving
_start_scan_tasks() behavior unchanged or similarly gated if needed.
In `@src/fastcs/demo/controllers.py`:
- Around line 98-102: reconnect currently calls self.connection.close() and
self.connection.connect(...) without handling exceptions, violating the
Controller.reconnect() contract; wrap the close/connect calls in a try/except
that catches Exception, log the failure with a descriptive message and exception
info (use a module logger via logging.getLogger(__name__)), and ensure
self._connected is only set True on successful connect and left False on failure
(do not re-raise).
---
Nitpick comments:
In `@src/fastcs/controllers/controller.py`:
- Around line 55-57: The base async disconnect() leaves the controller's
connection flag unchanged; update Controller.disconnect() to set self._connected
= False to reliably clear connection state for subclasses that do not override
disconnect(), ensuring lifecycle semantics are correct (modify the async def
disconnect(self) method in the Controller class to assign self._connected =
False before returning).
In `@tests/assertable_controller.py`:
- Line 79: The test helper currently calls the private method
controller._build_api([]); replace that with the public factory
create_api_and_tasks() to avoid depending on a private symbol. Call
create_api_and_tasks() (or controller.create_api_and_tasks() if it is a method)
and unpack its return (e.g., api, tasks = create_api_and_tasks(...)) or assign
the returned API to controller_api so the helper uses the supported public API
surface instead of _build_api.
In `@tests/conftest.py`:
- Around line 50-52: The fixture controller_api currently calls the private
helper controller._build_api([]); change it to use the controller's public
factory method (e.g. create_api_and_tasks or another documented public
API-building method) if one exists so tests exercise the public contract instead
of private internals; locate the controller_api fixture and replace
controller._build_api([]) with the appropriate public call (passing equivalent
arguments) to construct the API and tasks.
In `@tests/test_controllers.py`:
- Around line 260-283: The test sets controller state directly and doesn't
properly await task cancellation; change the setup to call and await
controller.connect() (instead of setting controller._connected = True) so the
public API is exercised (see MyTestController, connect, reconnect, scan
decorator and scan_coros usage), and when cancelling the background task created
from scan_coros[0]() ensure you await it and handle asyncio.CancelledError to
cleanly swallow cancellation (i.e., cancel the task, await it inside a
try/except catching CancelledError) so the test teardown is robust.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 942faf23-2977-4afa-afe4-a32b3fe23c64
📒 Files selected for processing (33)
src/fastcs/attributes/attr_r.pysrc/fastcs/connections/ip_connection.pysrc/fastcs/control_system.pysrc/fastcs/controllers/__init__.pysrc/fastcs/controllers/base_controller.pysrc/fastcs/controllers/controller.pysrc/fastcs/controllers/controller_api.pysrc/fastcs/demo/controllers.pysrc/fastcs/transports/__init__.pysrc/fastcs/transports/controller_api.pysrc/fastcs/transports/epics/ca/ioc.pysrc/fastcs/transports/epics/ca/transport.pysrc/fastcs/transports/epics/docs.pysrc/fastcs/transports/epics/gui.pysrc/fastcs/transports/epics/pva/ioc.pysrc/fastcs/transports/epics/pva/pvi.pysrc/fastcs/transports/epics/pva/transport.pysrc/fastcs/transports/epics/util.pysrc/fastcs/transports/graphql/graphql.pysrc/fastcs/transports/graphql/transport.pysrc/fastcs/transports/rest/rest.pysrc/fastcs/transports/rest/transport.pysrc/fastcs/transports/tango/dsr.pysrc/fastcs/transports/tango/transport.pysrc/fastcs/transports/transport.pytests/assertable_controller.pytests/conftest.pytests/test_control_system.pytests/test_controllers.pytests/transports/epics/ca/test_gui.pytests/transports/epics/ca/test_softioc.pytests/transports/epics/pva/test_pva_gui.pytests/transports/rest/test_rest.py
💤 Files with no reviewable changes (2)
- src/fastcs/transports/init.py
- src/fastcs/transports/controller_api.py
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #332 +/- ##
==========================================
+ Coverage 90.83% 90.90% +0.07%
==========================================
Files 70 70
Lines 2574 2574
==========================================
+ Hits 2338 2340 +2
+ Misses 236 234 -2 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
Pause all scan tasks when an exception occurs in one Add `Controller.reconnect` method to reocnnect and unpause scan tasks Move creation of initial tasks and scan tasks into Controller
GDYendell
force-pushed
the
reconnect-handling
branch
from
62d40ea to
07bcbc4
Compare
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (2)
src/fastcs/controllers/base_controller.py (1)
393-404: Clean API tree construction with recursive sub-controller handling.The
_build_apimethod correctly builds theControllerAPIhierarchy, including proper path propagation to nested sub-controllers.Consider using unpacking syntax for the path construction (per Ruff RUF005):
🔧 Minor style improvement
sub_apis={ - name: sub_controller._build_api(path + [name]) # noqa: SLF001 + name: sub_controller._build_api([*path, name]) # noqa: SLF001 for name, sub_controller in self.sub_controllers.items() },🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/fastcs/controllers/base_controller.py` around lines 393 - 404, In _build_api, replace the list concatenation used to build child paths (currently using path + [name] inside the sub_apis comprehension) with list unpacking so the child path is constructed as [*path, name]; update the comprehension where sub_controller._build_api is called (referencing sub_controllers, sub_controller, sub_apis) to use the unpacked path form to satisfy Ruff RUF005 and keep semantics identical.src/fastcs/controllers/controller.py (1)
85-93: Refactor to use the publicio_refproperty instead of directly accessing the private_io_refattribute.The
Attributeclass (inherited byAttrR) provides a publicio_refproperty andhas_io_ref()method. Replace the pattern match on the private_io_refwith logic that uses these public accessors:Example refactor
for attribute in api.attributes.values(): if isinstance(attribute, AttrR) and attribute.has_io_ref(): io_ref = attribute.io_ref if isinstance(io_ref, AttributeIORef): update_period = io_ref.update_period if update_period is ONCE: initial_coros.append(attribute.bind_update_callback()) elif update_period is not None: scan_dict[update_period].append( attribute.bind_update_callback() )🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/fastcs/controllers/controller.py` around lines 85 - 93, The pattern-match is directly accessing the private _io_ref; update the loop to use the public API: check isinstance(attribute, AttrR) and attribute.has_io_ref(), get io_ref = attribute.io_ref, ensure isinstance(io_ref, AttributeIORef), then read update_period = io_ref.update_period and branch as before (if update_period is ONCE -> initial_coros.append(attribute.bind_update_callback()); elif update_period is not None -> scan_dict[update_period].append(attribute.bind_update_callback())). Keep references to AttrR, has_io_ref(), io_ref, AttributeIORef, update_period, ONCE, bind_update_callback(), initial_coros and scan_dict to locate and update the code.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@tests/test_controllers.py`:
- Around line 282-283: After calling task.cancel(), replace the sleep-based
cleanup with explicitly awaiting the cancelled task to ensure proper teardown;
await the Task object named task and suppress/handle asyncio.CancelledError
(e.g., with a try/except around await task) so the cancelled coroutine is fully
awaited and avoids pending-task warnings in tests that use the task variable in
tests/test_controllers.py.
---
Nitpick comments:
In `@src/fastcs/controllers/base_controller.py`:
- Around line 393-404: In _build_api, replace the list concatenation used to
build child paths (currently using path + [name] inside the sub_apis
comprehension) with list unpacking so the child path is constructed as [*path,
name]; update the comprehension where sub_controller._build_api is called
(referencing sub_controllers, sub_controller, sub_apis) to use the unpacked path
form to satisfy Ruff RUF005 and keep semantics identical.
In `@src/fastcs/controllers/controller.py`:
- Around line 85-93: The pattern-match is directly accessing the private
_io_ref; update the loop to use the public API: check isinstance(attribute,
AttrR) and attribute.has_io_ref(), get io_ref = attribute.io_ref, ensure
isinstance(io_ref, AttributeIORef), then read update_period =
io_ref.update_period and branch as before (if update_period is ONCE ->
initial_coros.append(attribute.bind_update_callback()); elif update_period is
not None -> scan_dict[update_period].append(attribute.bind_update_callback())).
Keep references to AttrR, has_io_ref(), io_ref, AttributeIORef, update_period,
ONCE, bind_update_callback(), initial_coros and scan_dict to locate and update
the code.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 4018d294-4bb1-4a14-a4fb-b0d7bc0e8d1b
📒 Files selected for processing (33)
src/fastcs/attributes/attr_r.pysrc/fastcs/connections/ip_connection.pysrc/fastcs/control_system.pysrc/fastcs/controllers/__init__.pysrc/fastcs/controllers/base_controller.pysrc/fastcs/controllers/controller.pysrc/fastcs/controllers/controller_api.pysrc/fastcs/demo/controllers.pysrc/fastcs/transports/__init__.pysrc/fastcs/transports/controller_api.pysrc/fastcs/transports/epics/ca/ioc.pysrc/fastcs/transports/epics/ca/transport.pysrc/fastcs/transports/epics/docs.pysrc/fastcs/transports/epics/gui.pysrc/fastcs/transports/epics/pva/ioc.pysrc/fastcs/transports/epics/pva/pvi.pysrc/fastcs/transports/epics/pva/transport.pysrc/fastcs/transports/epics/util.pysrc/fastcs/transports/graphql/graphql.pysrc/fastcs/transports/graphql/transport.pysrc/fastcs/transports/rest/rest.pysrc/fastcs/transports/rest/transport.pysrc/fastcs/transports/tango/dsr.pysrc/fastcs/transports/tango/transport.pysrc/fastcs/transports/transport.pytests/assertable_controller.pytests/conftest.pytests/test_control_system.pytests/test_controllers.pytests/transports/epics/ca/test_gui.pytests/transports/epics/ca/test_softioc.pytests/transports/epics/pva/test_pva_gui.pytests/transports/rest/test_rest.py
💤 Files with no reviewable changes (2)
- src/fastcs/transports/init.py
- src/fastcs/transports/controller_api.py
🚧 Files skipped from review as they are similar to previous changes (15)
- src/fastcs/transports/graphql/graphql.py
- src/fastcs/connections/ip_connection.py
- tests/transports/epics/ca/test_softioc.py
- src/fastcs/transports/rest/transport.py
- tests/transports/rest/test_rest.py
- tests/conftest.py
- tests/transports/epics/pva/test_pva_gui.py
- src/fastcs/transports/epics/docs.py
- src/fastcs/demo/controllers.py
- src/fastcs/transports/epics/ca/transport.py
- src/fastcs/transports/epics/pva/transport.py
- src/fastcs/transports/graphql/transport.py
- src/fastcs/controllers/init.py
- tests/assertable_controller.py
- src/fastcs/transports/epics/pva/ioc.py
shihab-dls
approved these changes
Mar 6, 2026
Contributor
shihab-dls
left a comment
shihab-dls
left a comment
There was a problem hiding this comment.
I've pointed out one coderabbit suggestion I think we want, and a suggestion for scan coros, the latter is up for debate, so I'm approving as I'm happy with this current iteration anyways!
There was a problem hiding this comment.
🧹 Nitpick comments (1)
src/fastcs/controllers/controller.py (1)
119-135: Thesleep(1)for log ordering is fragile.Using
await asyncio.sleep(1)at line 133 to ensure the error message appears after the exception log is timing-dependent and could be unreliable under load. Consider combining these into a single structured log statement or removing the artificial delay.♻️ Suggested simplification
except Exception: - logger.exception("Exception in scan task", period=period) - self._connected = False - - await asyncio.sleep(1) # Wait so this message appears last - logger.error("Pausing scan tasks and waiting for reconnect") + logger.exception( + "Exception in scan task - pausing and waiting for reconnect", + period=period, + ) + self._connected = False🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/fastcs/controllers/controller.py` around lines 119 - 135, In scan_coro(), remove the fragile await asyncio.sleep(1) used to force log ordering and instead emit a single clear log (or back-to-back logs) that include the exception context; specifically, in the except block for the asyncio.gather call replace the sleep-and-error sequence with a single logger call (or immediate logger.exception followed by logger.error) that mentions period and reconnection state, keep setting self._connected = False, and ensure you pass the exception context (use logger.exception or include exc_info) so the stacktrace and the "Pausing scan tasks and waiting for reconnect" message are logged together without relying on time-based ordering.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@src/fastcs/controllers/controller.py`:
- Around line 119-135: In scan_coro(), remove the fragile await asyncio.sleep(1)
used to force log ordering and instead emit a single clear log (or back-to-back
logs) that include the exception context; specifically, in the except block for
the asyncio.gather call replace the sleep-and-error sequence with a single
logger call (or immediate logger.exception followed by logger.error) that
mentions period and reconnection state, keep setting self._connected = False,
and ensure you pass the exception context (use logger.exception or include
exc_info) so the stacktrace and the "Pausing scan tasks and waiting for
reconnect" message are logged together without relying on time-based ordering.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 212d87c3-c425-4096-8ca9-130126b6155b
📒 Files selected for processing (2)
src/fastcs/controllers/controller.pytests/test_controllers.py
GDYendell
force-pushed
the
reconnect-handling
branch
from
3cc47af to
02fb2f6
Compare
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@tests/test_controllers.py`:
- Around line 260-284: The test
test_scan_exception_sets_disconnected_and_reconnect_resumes currently reconnects
and immediately cancels the scan task so it never verifies a resumed scan;
modify the test to wait for a second scan attempt after calling
controller.reconnect() (i.e., sleep or await long enough to surpass the 1s
backoff in Controller.scan handling) before cancelling the task so that
failing_scan (the scan coroutine created via scan decorator) actually runs again
and you can assert that scanning resumed and _connected stayed True.
261931c to
29a6607
Compare
There was a problem hiding this comment.
♻️ Duplicate comments (1)
tests/test_controllers.py (1)
260-284:⚠️ Potential issue | 🟠 MajorThis still doesn't prove scans resumed after
reconnect().The scan loop backs off for 1 second after the first exception, but this test reconnects after
0.1sand cancels immediately. A regression wherereconnect()only flips_connectedwithout ever running another scan would still pass.✅ Tighten the assertion to wait for a post-reconnect scan attempt
`@pytest.mark.asyncio` async def test_scan_exception_sets_disconnected_and_reconnect_resumes(): class MyTestController(Controller): + def __init__(self): + super().__init__() + self.scan_attempts = 0 + self.resumed = asyncio.Event() + `@scan`(0.01) async def failing_scan(self): + self.scan_attempts += 1 + if self.scan_attempts >= 2: + self.resumed.set() raise RuntimeError("scan error") controller = MyTestController() controller.post_initialise() _, scan_coros, _ = controller.create_api_and_tasks() @@ await asyncio.sleep(0.1) assert not controller._connected - # Trigger reconnect - _connected resumes scan tasks + attempts_before_reconnect = controller.scan_attempts await controller.reconnect() - assert controller._connected + await asyncio.wait_for(controller.resumed.wait(), timeout=2) + assert controller._connected + assert controller.scan_attempts > attempts_before_reconnect task.cancel() with pytest.raises(asyncio.CancelledError): await task🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@tests/test_controllers.py` around lines 260 - 284, The test test_scan_exception_sets_disconnected_and_reconnect_resumes currently only flips _connected and cancels the task too soon; modify it to assert that a scan actually runs after reconnect by waiting for a post-reconnect scan attempt: instrument the failing_scan (or the test) to record a second invocation (e.g., increment a counter or set an asyncio.Event) from the scan coroutine, then after await controller.reconnect() wait for that counter/event (with a timeout) before asserting controller._connected and cancelling the task, ensuring reconnect() caused a new scan run; reference MyTestController, failing_scan, controller.reconnect(), and the scan_coros/task to locate where to add the counter/event and the post-reconnect wait.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Duplicate comments:
In `@tests/test_controllers.py`:
- Around line 260-284: The test
test_scan_exception_sets_disconnected_and_reconnect_resumes currently only flips
_connected and cancels the task too soon; modify it to assert that a scan
actually runs after reconnect by waiting for a post-reconnect scan attempt:
instrument the failing_scan (or the test) to record a second invocation (e.g.,
increment a counter or set an asyncio.Event) from the scan coroutine, then after
await controller.reconnect() wait for that counter/event (with a timeout) before
asserting controller._connected and cancelling the task, ensuring reconnect()
caused a new scan run; reference MyTestController, failing_scan,
controller.reconnect(), and the scan_coros/task to locate where to add the
counter/event and the post-reconnect wait.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: c9036862-d040-40d1-a455-3729622014f1
📒 Files selected for processing (3)
src/fastcs/connections/ip_connection.pytests/test_controllers.pytests/test_ip_connection.py
GDYendell
force-pushed
the
reconnect-handling
branch
from
29a6607 to
dd6944f
Compare
GDYendell
force-pushed
the
reconnect-handling
branch
from
dd6944f to
be338db
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary by CodeRabbit
Refactoring
Bug Fixes
New Features
Tests