Files
lmxopcua/src
Joseph Doherty 6c5b202910 TwinCAT follow-up — Native ADS notifications for ISubscribable. Closes task #189 — upgrades TwinCATDriver's subscription path from polling (shared PollGroupEngine) to native AdsClient.AddDeviceNotificationExAsync so the PLC pushes changes on its own cycle rather than the driver polling. Strictly better for latency + CPU — TC2 and TC3 runtimes notify on value change with sub-millisecond latency from the PLC cycle. ITwinCATClient gains AddNotificationAsync — takes symbolPath + TwinCATDataType + optional bitIndex + cycleTime + onChange callback + CancellationToken; returns an ITwinCATNotificationHandle whose Dispose tears the notification down on the wire. Bit-within-word reads supported — the parent word value arrives via the notification, driver extracts the bit before invoking the callback (same ExtractBit path as the read surface from PR 2). AdsTwinCATClient — subscribes to AdsClient.AdsNotificationEx in the ctor, maintains a ConcurrentDictionary<uint, NotificationRegistration> keyed on the server-side notification handle. AddDeviceNotificationExAsync returns Task<ResultHandle> with Handle + ErrorCode; non-NoError throws InvalidOperationException so the driver can catch + retry. Notification event args carry Handle + Value + DataType; lookup in _notifications dict routes the value through any bit-extraction + calls the consumer callback. Consumer-side exceptions are swallowed so a misbehaving callback can't crash the ADS notification thread. Dispose unsubscribes from AdsNotificationEx + clears the dict + disposes AdsClient. NotificationRegistration is ITwinCATNotificationHandle — Dispose fires DeleteDeviceNotificationAsync as fire-and-forget with CancellationToken.None (caller has already committed to teardown; blocking would slow shutdown). TwinCATDriverOptions.UseNativeNotifications — new bool, default true. When true the driver uses native notifications; when false it falls through to the shared PollGroupEngine (same semantics as other libplctag-backed drivers, also a safety valve for targets with notification limits). TwinCATDriver.SubscribeAsync dual-path — if UseNativeNotifications false delegate into _poll.Subscribe (unchanged behavior from PR 3). If true, iterate fullReferences, resolve each to its device's client via EnsureConnectedAsync (reuses PR 2's per-device connection cache), parse the SymbolPath via TwinCATSymbolPath (preserves bit-in-word support), call ITwinCATClient.AddNotificationAsync with a closure over the FullReference (not the ADS symbol — OPC UA subscribers addressed the driver-side name). Per-registration callback bridges (_, value) → OnDataChange event with a fresh DataValueSnapshot (Good status, current UtcNow timestamps). Any mid-registration failure triggers a try/catch that disposes every already-registered handle before rethrowing, keeping the driver in a clean never-existed state rather than half-registered. UnsubscribeAsync dispatches on handle type — NativeSubscriptionHandle disposes all its cached ITwinCATNotificationHandles; anything else delegates to _poll.Unsubscribe for the poll fallback. ShutdownAsync tears down native subs first (so AdsClient-level cleanup happens before the client itself disposes), then PollGroupEngine, then per-device probe CTS + client. NativeSubscriptionHandle DiagnosticId prefixes with twincat-native-sub- so Admin UI + logs can distinguish the paths. 9 new unit tests in TwinCATNativeNotificationTests — native subscribe registers one notification per tag, pushed value via FireNotification fires OnDataChange with the right FullReference (driver-side, not ADS symbol), unsubscribe disposes all notifications, unsubscribe halts future notifications, partial-failure cleanup via FailAfterNAddsFake (first succeeds, second throws → first gets torn down + Notifications count returns to 0 + AddCallCount=2 proving the test actually exercised both calls), shutdown disposes subscriptions, poll fallback works when UseNativeNotifications=false (no native handles created + initial-data push still fires), handle DiagnosticId distinguishes native vs poll. Existing poll-mode ISubscribable tests in TwinCATCapabilityTests updated with UseNativeNotifications=false so they continue testing the poll path specifically — both poll + native paths have test coverage now. TwinCATDriverTests got Probe.Enabled=false added because the default factory creates a real AdsClient which was flakily affected by parallel test execution sharing AMS router state. Total TwinCAT unit tests now 93/93 passing (+8 from PR 3's 85 counting the new native tests + 2 existing tests that got options tweaks). Full solution builds 0 errors; Modbus / AbCip / AbLegacy / other drivers untouched. TwinCAT driver is now feature-complete end-to-end — read / write / discover / native-subscribe / probe / host-resolve, with poll-mode as a safety valve. Unblocks closing task #120 for TwinCAT; remaining sub-task: FOCAS + task #188 (symbol-browsing — lower priority than FOCAS since real config flows still use pre-declared tags).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-19 18:49:48 -04:00
..
AB CIP PR 1 — extract shared PollGroupEngine into Core.Abstractions so the AB CIP driver (and any other poll-based driver — S7, FOCAS, AB Legacy) can reuse the subscription loop instead of reimplementing it. Behaviour-preserving refactor of ModbusDriver: SubscriptionState + PollLoopAsync + PollOnceAsync + ModbusSubscriptionHandle lifted verbatim into a new PollGroupEngine class, ModbusDriver's ISubscribable surface now delegates Subscribe/Unsubscribe into the engine and ShutdownAsync calls engine DisposeAsync. Interval floor (100 ms default) becomes a PollGroupEngine constructor knob so per-driver tuning is possible without re-shipping the loop. Initial-data push semantics preserved via forceRaise=true on the first poll. Exception-tolerant loop preserved — reader throws are swallowed, loop continues, driver's health surface remains the single reporting path. Placement in Core.Abstractions (not Core) because driver projects only reference Core.Abstractions by convention (matches OpcUaClient / Modbus / S7 csproj shape); putting the engine in Core would drag EF Core + Serilog + Polly into every driver. Module has no new dependencies beyond System.Collections.Concurrent + System.Threading, so Core.Abstractions stays lightweight. Modbus ctor converted from primary to explicit so the engine field can capture this for the reader + on-change bridge. All 177 ModbusDriver.Tests pass unmodified (Modbus subscription suite, probe suite, cap suite, exception mapper, reconnect, TCP). 10 new direct engine tests in Core.Abstractions.Tests covering: initial force-raise, unchanged-value single-raise, change-between-polls, unsubscribe halts loop, interval-floor clamp, independent subscriptions, reader-exception tolerance, unknown-handle returns false, ActiveSubscriptionCount lifecycle, DisposeAsync cancels all. No changes to driver-specs.md nor to the server Hosting layer — engine is a pure internal building block at this stage. Unblocks AB CIP PR 7 (ISubscribable consumes the engine); also sets up S7 + FOCAS to drop their own poll loops when they re-base.
2026-04-19 15:34:44 -04:00
AB CIP PR 9-12 — Per-PLC-family profile tests + GuardLogix safety-tag support. Consolidates PRs 9/10/11/12 from the plan (ControlLogix / CompactLogix / Micro800 / GuardLogix integration suites) into a single PR because the per-family work that actually ships without a live ab_server binary is profile-metadata assertion + unit-level driver-option binding. Per-family integration tests that require a running simulator are deferred to the ab_server-CI follow-up already tracked from PR 3 (download prebuilt Windows binary as GitHub release asset). ControlLogix — baseline profile asserted (controllogix attribute, 4002 LFO ConnectionSize, 1,0 default path, request-packing + connected-messaging, 4000B max fragment). CompactLogix — narrower 504 ConnectionSize for 5069-L3x safety, 500B max fragment, lib attribute compactlogix which libplctag maps to the ControlLogix family internally but via our profile chain we surface it as a distinct knob so future quirk handling (5069 narrow-window regression cases) hangs off the compactlogix attribute. Micro800 — empty CIP path for no-backplane routing, 488B ConnectionSize, 484B fragment cap, request packing + connected messaging both disabled (most models reject Forward_Open), micro800 lib attribute. Test asserts the driver correctly parses an ab://192.168.1.20/ host address with empty path + forwards the empty path through AbCipTagCreateParams so libplctag sees the unconnected-only configuration. GuardLogix — wire protocol identical to ControlLogix (safety partition is a per-tag concern, not a wire-layer distinction) so profile defaults match ControlLogix. New AbCipTagDefinition.SafetyTag field — when true, the driver forces SecurityClassification.ViewOnly in discovery regardless of the Writable flag, and IWritable rejects the write upfront with BadNotWritable. Matches the Rockwell safety-partition isolation model where non-safety-task writes to safety tags would be rejected by the PLC anyway — surfacing the intent at the driver surface prevents wasted wire round-trips + gives Admin UI users a correct ViewOnly rendering. 14 new unit tests in AbCipPlcFamilyTests covering — ControlLogix profile defaults + correct profile selection at Initialize, CompactLogix narrower-than-ControlLogix ConnectionSize + fragment cap, Micro800 empty path parses + SupportsConnectedMessaging=false + SupportsRequestPacking=false + read forwards empty path + micro800 attribute through to libplctag, GuardLogix wire-protocol parity with ControlLogix, GuardLogix safety tag surfaces as ViewOnly in discovery even when Writable=true, GuardLogix safety-tag write rejected with BadNotWritable even when Writable=true, ForFamily theory (4 families → correct libplctag attribute). Total AbCip unit tests now 161/161 passing (+14 from PR 8's 147). Modbus + other drivers untouched; full solution builds 0 errors. PR 13 (IAlarmSource via tag-projected ALMA/ALMD blocks) remains deferred per the plan — feature-flagged pattern not needed before go-live.
2026-04-19 17:18:51 -04:00
AB Legacy PR 3 — ITagDiscovery + ISubscribable + IHostConnectivityProbe + IPerCallHostResolver. Fills out the AbLegacy capability surface — the driver now implements the same 7-interface set as AbCip (IDriver + IReadable + IWritable + ITagDiscovery + ISubscribable + IHostConnectivityProbe + IPerCallHostResolver). ITagDiscovery emits pre-declared tags under an AbLegacy root folder with a per-device sub-folder keyed on HostAddress (DeviceName fallback to HostAddress when null). Writable tags surface as SecurityClassification.Operate, non-writable as ViewOnly. No controller-side enumeration — PCCC has no @tags equivalent on SLC / MicroLogix / PLC-5 (symbol table isn't exposed the way Logix exposes it), so the pre-declared path is the only discovery mechanism. ISubscribable consumes the shared PollGroupEngine extracted in AB CIP PR 1 — reader delegate points at ReadAsync (already handles lazy runtime init + caching), onChange bridges into the driver's OnDataChange event. 100ms interval floor. Initial-data push on first poll. Makes AbLegacy the third consumer of PollGroupEngine (after Modbus and AbCip). IHostConnectivityProbe — per-device probe loop when ProbeOptions.Enabled + ProbeAddress configured (defaults to S:0 status file word 0). Lazy-init on first tick, re-init on wire failure (destroyed native handle gets recreated rather than silently staying broken). Success transitions device to Running, exception to Stopped, same-state spurious event guard under per-device lock. GetHostStatuses returns one entry per device with current state + last-change timestamp for Admin /hosts surfacing. IPerCallHostResolver maps tag full-ref → DeviceHostAddress for the Phase 6.1 (DriverInstanceId, ResolvedHostName) bulkhead/breaker keying per plan decision #144. Unknown refs fall back to first device's address (invoker handles at capability level as BadNodeIdUnknown); no devices → DriverInstanceId. ShutdownAsync cancels + disposes each probe CTS, disposes PollGroupEngine cancelling active subscriptions, disposes every cached runtime. DeviceState gains ProbeLock / HostState / HostStateChangedUtc / ProbeCts / ProbeInitialized matching AbCip's DeviceState shape. 10 new unit tests in AbLegacyCapabilityTests covering — pre-declared tags emit under AbLegacy/device folder with correct SecurityClassification, subscription initial poll raises OnDataChange with correct value, unsubscribe halts polling (value change post-unsub produces no further events), GetHostStatuses returns one entry per device, probe Running transition on successful read, probe Stopped transition on read exception, probe disabled when ProbeAddress null, ResolveHost returns declared device for known tag, falls back to first device for unknown, falls back to DriverInstanceId when no devices. Total AbLegacy unit tests now 92/92 passing (+10 from PR 2's 82); full solution builds 0 errors; AbCip + Modbus + other drivers untouched. AB Legacy driver now complete end-to-end — SLC 500 / MicroLogix / PLC-5 / LogixPccc all shippable with read / write / discovery / subscribe / probe / host-resolve, feature-parity with AbCip minus IAlarmSource (same deferral per plan).
2026-04-19 18:02:52 -04:00
Phase 2 PR 14 — alarm subsystem wire-up. Per IsAlarm=true attribute (PR 9 added the discovery flag), GalaxyAlarmTracker in Backend/Alarms/ advises the four Galaxy alarm-state attributes: .InAlarm (boolean alarm active), .Priority (int severity), .DescAttrName (human-readable description), .Acked (boolean acknowledged). Runs the OPC UA Part 9 alarm lifecycle state machine simplified for the Galaxy AlarmExtension model and raises AlarmTransition events on transitions operators must react to — Active (InAlarm false→true, default Unacknowledged), Acknowledged (Acked false→true while InAlarm still true), Inactive (InAlarm true→false). MxAccessGalaxyBackend instantiates the tracker in its constructor with delegate-based subscribe/unsubscribe/write pointers to MxAccessClient, hooks TransitionRaised to forward each transition through the existing OnAlarmEvent IPC event that PR 4 ConnectionSink wires into MessageKind.AlarmEvent frames — no new contract messages required since GalaxyAlarmEvent already exists in Shared.Contracts. Field mapping: EventId = fresh Guid.ToString('N') per transition, ObjectTagName = alarm attribute full reference, AlarmName = alarm attribute full reference, Severity = tracked Priority, StateTransition = 'Active'|'Acknowledged'|'Inactive', Message = DescAttrName or tag fallback, UtcUnixMs = transition time. DiscoverAsync caches every IsAlarm=true attribute's full reference (tag.attribute) into _discoveredAlarmTags (ConcurrentBag cleared-then-filled on every re-Discover to track Galaxy redeploys). SubscribeAlarmsAsync iterates the cache and advises each via GalaxyAlarmTracker.TrackAsync; best-effort per-alarm — a subscribe failure on one alarm doesn't abort the whole call since operators prefer partial alarm coverage to none. Tracker is internally idempotent on repeat Track calls (second invocation for same alarm tag is a no-op; already-subscribed check short-circuits before the 4 MXAccess sub calls). Subscribe-failure rollback inside TrackAsync removes the alarm state + unadvises any of the 4 that did succeed so a partial advise can't leak a phantom tracking entry. AcknowledgeAlarmAsync routes to tracker.AcknowledgeAsync which writes the operator comment to <tag>.AckMsg via MxAccessClient.WriteAsync — writes use the existing MXAccess OnWriteComplete TCS-by-handle path (PR 4 Medium 4) so a runtime-refused ack bubbles up as Success=false rather than false-positive. State-machine quirks preserved from v1: (1) initial Acked=true on subscribe does NOT fire Acknowledged (alarm at rest, pre-acknowledged — default state is Acked=true so the first subscribe callback is a no-op transition), (2) Acked false→true only fires Acknowledged when InAlarm is currently true (acking a latched-inactive alarm is not a user-visible transition), (3) Active transition clears the Acked flag in-state so the next Acked callback correctly fires Acknowledged (v1 had this buried in the ConditionState logic; we track it on the AlarmState struct directly). Priority value handled as int/short/long via type pattern match with int.MaxValue guard — Galaxy attribute category returns varying CLR types (Int32 is canonical but some older templates use Int16), and a long overflow cast to int would silently corrupt the severity. Dispose cascade in MxAccessGalaxyBackend.Dispose: alarm-tracker unsubscribe→dispose, probe-manager unsubscribe→dispose, mx.ConnectionStateChanged detach, historian dispose — same discipline PR 6 / PR 8 / PR 13 established so dangling invocation-list refs don't survive a backend recycle. #pragma warning disable CS0067 around OnAlarmEvent removed since the event is now raised. Tests (9 new, GalaxyAlarmTrackerTests): four-attribute subscribe per alarm, idempotent repeat-track, InAlarm false→true fires Active with Priority + Desc, InAlarm true→false fires Inactive, Acked false→true while InAlarm fires Acknowledged, Acked transition while InAlarm=false does not fire, AckMsg write path carries the comment, snapshot reports latest four fields, foreign probe callback for a non-tracked tag is silently dropped. Full Galaxy.Host.Tests Unit suite 84 pass / 0 fail (9 new alarm + 12 PR 13 probe + 21 PR 12 quality + 42 pre-existing). Galaxy.Host builds clean (0/0). Branches off phase-2-pr13-runtime-probe so the MxAccessGalaxyBackend constructor/Dispose chain gets the probe-manager + alarm-tracker wire-up in a coherent order; fast-forwards if PR 13 merges first.
2026-04-18 07:34:13 -04:00
Phase 2 PR 11 — HistoryReadEvents IPC (alarm history). New Shared.Contracts messages HistoryReadEventsRequest/Response + GalaxyHistoricalEvent DTO (MessageKind 0x66/0x67). IGalaxyBackend gains HistoryReadEventsAsync, Stub/DbBacked return canonical pending error, MxAccessGalaxyBackend delegates to _historian.ReadEventsAsync (ported in PR 5) and maps HistorianEventDto → GalaxyHistoricalEvent — Guid.ToString() for EventId wire shape, DateTime → Unix ms for both EventTime (when the event fired in the process) and ReceivedTime (when the Historian persisted it), DisplayText + Severity pass through. SourceName is string? — null means 'all sources' (passed straight through to HistorianDataSource.ReadEventsAsync which adds the AddEventFilter('Source', Equal, ...) only when non-null). Distinct from the live GalaxyAlarmEvent type because historical rows carry both timestamps and lack StateTransition (Historian logs instantaneous events, not the OPC UA Part 9 alarm lifecycle; translating to OPC UA event lifecycle is the alarm-subsystem's job). Guards: null historian → Historian-disabled error; SDK exception → Success=false with message chained. Tests (3 new): disabled-error when historian null, maps HistorianEventDto with full field set (Id/Source/EventTime/ReceivedTime/DisplayText/Severity=900) to GalaxyHistoricalEvent, null SourceName passes through unchanged (verifies the 'all sources' contract). Galaxy.Host.Tests Unit suite 34 pass / 0 fail. Galaxy.Host builds clean. Branches off phase-2-pr10-history-attime since both extend the MessageKind enum; fast-forwards if PR 10 merges first.
2026-04-18 07:08:16 -04:00
AB CIP PR 1 — extract shared PollGroupEngine into Core.Abstractions so the AB CIP driver (and any other poll-based driver — S7, FOCAS, AB Legacy) can reuse the subscription loop instead of reimplementing it. Behaviour-preserving refactor of ModbusDriver: SubscriptionState + PollLoopAsync + PollOnceAsync + ModbusSubscriptionHandle lifted verbatim into a new PollGroupEngine class, ModbusDriver's ISubscribable surface now delegates Subscribe/Unsubscribe into the engine and ShutdownAsync calls engine DisposeAsync. Interval floor (100 ms default) becomes a PollGroupEngine constructor knob so per-driver tuning is possible without re-shipping the loop. Initial-data push semantics preserved via forceRaise=true on the first poll. Exception-tolerant loop preserved — reader throws are swallowed, loop continues, driver's health surface remains the single reporting path. Placement in Core.Abstractions (not Core) because driver projects only reference Core.Abstractions by convention (matches OpcUaClient / Modbus / S7 csproj shape); putting the engine in Core would drag EF Core + Serilog + Polly into every driver. Module has no new dependencies beyond System.Collections.Concurrent + System.Threading, so Core.Abstractions stays lightweight. Modbus ctor converted from primary to explicit so the engine field can capture this for the reader + on-change bridge. All 177 ModbusDriver.Tests pass unmodified (Modbus subscription suite, probe suite, cap suite, exception mapper, reconnect, TCP). 10 new direct engine tests in Core.Abstractions.Tests covering: initial force-raise, unchanged-value single-raise, change-between-polls, unsubscribe halts loop, interval-floor clamp, independent subscriptions, reader-exception tolerance, unknown-handle returns false, ActiveSubscriptionCount lifecycle, DisposeAsync cancels all. No changes to driver-specs.md nor to the server Hosting layer — engine is a pure internal building block at this stage. Unblocks AB CIP PR 7 (ISubscribable consumes the engine); also sets up S7 + FOCAS to drop their own poll loops when they re-base.
2026-04-19 15:34:44 -04:00
Phase 3 PR 76 -- OPC UA Client IHistoryProvider (HistoryRead passthrough). Driver now implements IHistoryProvider (Raw + Processed + AtTime); ReadEventsAsync deliberately inherits the interface default that throws NotSupportedException. ExecuteHistoryReadAsync is the shared wire path: parses the fullReference to NodeId, builds a HistoryReadValueIdCollection with one entry, calls Session.HistoryReadAsync(RequestHeader, ExtensionObject<details>, TimestampsToReturn.Both, releaseContinuationPoints:false, nodesToRead, ct), unwraps r.HistoryData ExtensionObject into the samples list, passes ContinuationPoint through. Each DataValue's upstream StatusCode + SourceTimestamp + ServerTimestamp preserved verbatim per driver-specs.md \u00A78 cascading-quality rule -- this matters especially for historical data where an interpolated / uncertain-quality sample must surface its true severity downstream, not a sanitized Good. SourceTimestamp=DateTime.MinValue guards map to null so downstream clients see 'source unknown' rather than an epoch-zero misread. ReadRawAsync builds ReadRawModifiedDetails with IsReadModified=false (raw, not modified-history), StartTime/EndTime, NumValuesPerNode=maxValuesPerNode, ReturnBounds=false (clients that want bounds request them via continuation handling). ReadProcessedAsync builds ReadProcessedDetails with ProcessingInterval in ms + AggregateType wrapping a single NodeId from MapAggregateToNodeId. MapAggregateToNodeId switches on HistoryAggregateType {Average, Minimum, Maximum, Total, Count} to the standard Part 13 ObjectIds.AggregateFunction_* NodeId -- future aggregate-type additions fail the switch with ArgumentOutOfRangeException so they can't silently slip through with a null NodeId and an opaque server-side BadAggregateNotSupported. ReadAtTimeAsync builds ReadAtTimeDetails with ReqTimes + UseSimpleBounds=true (returns boundary samples when an exact timestamp has no value -- the OPC UA Part 11 default). Malformed NodeId short-circuits to empty result without touching the wire, matching the ReadAsync / WriteAsync pattern. ReadEventsAsync stays at the interface-default NotSupportedException: the OPC UA call path (HistoryReadAsync with ReadEventDetails + EventFilter) needs an EventFilter SelectClauses spec which the current IHistoryProvider.ReadEventsAsync signature doesn't carry. Adding that would be an IHistoryProvider interface widening; out of scope for PR 76. Callers see BadHistoryOperationUnsupported on the OPC UA client which is the documented fallback. Name disambiguation: Core.Abstractions.HistoryReadResult and Opc.Ua.HistoryReadResult both exist; used fully-qualified Core.Abstractions.HistoryReadResult in return types + factory expressions. Shutdown unchanged -- history reads don't create persistent server-side resources, so no cleanup needed beyond the existing Session.CloseAsync. Unit tests (OpcUaClientHistoryTests, 7 facts): MapAggregateToNodeId theory covers all 5 aggregates; MapAggregateToNodeId_rejects_invalid_enum (defense against future enum addition silently passing through); Read{Raw,Processed,AtTime}Async_without_initialize_throws (RequireSession path); ReadEventsAsync_throws_NotSupportedException (locks in the intentional inheritance of the default). 78/78 OpcUaClient.Tests pass (67 prior + 11 new, -4 on the alarm suite moved into the events count). dotnet build clean. Final OPC UA Client capability surface: IDriver + ITagDiscovery + IReadable + IWritable + ISubscribable + IHostConnectivityProbe + IAlarmSource + IHistoryProvider -- 8 of 8 possible capabilities. Driver is feature-complete per driver-specs.md \u00A78.
2026-04-19 02:13:22 -04:00
TwinCAT follow-up — Native ADS notifications for ISubscribable. Closes task #189 — upgrades TwinCATDriver's subscription path from polling (shared PollGroupEngine) to native AdsClient.AddDeviceNotificationExAsync so the PLC pushes changes on its own cycle rather than the driver polling. Strictly better for latency + CPU — TC2 and TC3 runtimes notify on value change with sub-millisecond latency from the PLC cycle. ITwinCATClient gains AddNotificationAsync — takes symbolPath + TwinCATDataType + optional bitIndex + cycleTime + onChange callback + CancellationToken; returns an ITwinCATNotificationHandle whose Dispose tears the notification down on the wire. Bit-within-word reads supported — the parent word value arrives via the notification, driver extracts the bit before invoking the callback (same ExtractBit path as the read surface from PR 2). AdsTwinCATClient — subscribes to AdsClient.AdsNotificationEx in the ctor, maintains a ConcurrentDictionary<uint, NotificationRegistration> keyed on the server-side notification handle. AddDeviceNotificationExAsync returns Task<ResultHandle> with Handle + ErrorCode; non-NoError throws InvalidOperationException so the driver can catch + retry. Notification event args carry Handle + Value + DataType; lookup in _notifications dict routes the value through any bit-extraction + calls the consumer callback. Consumer-side exceptions are swallowed so a misbehaving callback can't crash the ADS notification thread. Dispose unsubscribes from AdsNotificationEx + clears the dict + disposes AdsClient. NotificationRegistration is ITwinCATNotificationHandle — Dispose fires DeleteDeviceNotificationAsync as fire-and-forget with CancellationToken.None (caller has already committed to teardown; blocking would slow shutdown). TwinCATDriverOptions.UseNativeNotifications — new bool, default true. When true the driver uses native notifications; when false it falls through to the shared PollGroupEngine (same semantics as other libplctag-backed drivers, also a safety valve for targets with notification limits). TwinCATDriver.SubscribeAsync dual-path — if UseNativeNotifications false delegate into _poll.Subscribe (unchanged behavior from PR 3). If true, iterate fullReferences, resolve each to its device's client via EnsureConnectedAsync (reuses PR 2's per-device connection cache), parse the SymbolPath via TwinCATSymbolPath (preserves bit-in-word support), call ITwinCATClient.AddNotificationAsync with a closure over the FullReference (not the ADS symbol — OPC UA subscribers addressed the driver-side name). Per-registration callback bridges (_, value) → OnDataChange event with a fresh DataValueSnapshot (Good status, current UtcNow timestamps). Any mid-registration failure triggers a try/catch that disposes every already-registered handle before rethrowing, keeping the driver in a clean never-existed state rather than half-registered. UnsubscribeAsync dispatches on handle type — NativeSubscriptionHandle disposes all its cached ITwinCATNotificationHandles; anything else delegates to _poll.Unsubscribe for the poll fallback. ShutdownAsync tears down native subs first (so AdsClient-level cleanup happens before the client itself disposes), then PollGroupEngine, then per-device probe CTS + client. NativeSubscriptionHandle DiagnosticId prefixes with twincat-native-sub- so Admin UI + logs can distinguish the paths. 9 new unit tests in TwinCATNativeNotificationTests — native subscribe registers one notification per tag, pushed value via FireNotification fires OnDataChange with the right FullReference (driver-side, not ADS symbol), unsubscribe disposes all notifications, unsubscribe halts future notifications, partial-failure cleanup via FailAfterNAddsFake (first succeeds, second throws → first gets torn down + Notifications count returns to 0 + AddCallCount=2 proving the test actually exercised both calls), shutdown disposes subscriptions, poll fallback works when UseNativeNotifications=false (no native handles created + initial-data push still fires), handle DiagnosticId distinguishes native vs poll. Existing poll-mode ISubscribable tests in TwinCATCapabilityTests updated with UseNativeNotifications=false so they continue testing the poll path specifically — both poll + native paths have test coverage now. TwinCATDriverTests got Probe.Enabled=false added because the default factory creates a real AdsClient which was flakily affected by parallel test execution sharing AMS router state. Total TwinCAT unit tests now 93/93 passing (+8 from PR 3's 85 counting the new native tests + 2 existing tests that got options tweaks). Full solution builds 0 errors; Modbus / AbCip / AbLegacy / other drivers untouched. TwinCAT driver is now feature-complete end-to-end — read / write / discover / native-subscribe / probe / host-resolve, with poll-mode as a safety valve. Unblocks closing task #120 for TwinCAT; remaining sub-task: FOCAS + task #188 (symbol-browsing — lower priority than FOCAS since real config flows still use pre-declared tags).
2026-04-19 18:49:48 -04:00