temporal_logic_flutter 0.1.1

Flutter integration (TraceRecorder, Matchers) for temporal logic.

temporal_logic_flutter

This package bridges the gap between temporal logic specifications (LTL and MTL from temporal_logic_core and temporal_logic_mtl) and live Flutter application state.

It provides tools to monitor streams of application state changes and evaluate temporal properties against them in real-time, facilitating runtime verification and debugging of complex temporal behaviors.

Features

• Stream Checkers:
  • StreamLtlChecker: Evaluates standard LTL formulas against a stream of state updates.
  • StreamMtlChecker: Evaluates MTL formulas (with time constraints) against a stream of timed state updates.
  • StreamSustainedStateChecker: Checks if a specific state predicate holds continuously for a required duration within a stream.
• Trace Recording: TraceRecorder widget (or utility) to capture a sequence of state changes from a stream into a Trace object for later analysis or evaluation.
• Visualization Widgets (Examples/Utilities):
  • LtlCheckerWidget, MtlCheckerWidget, SustainedStateCheckerWidget: Example widgets that likely consume a stream checker and display its current status (CheckStatus: unknown, success, failure).
• State Matching: Utility functions or classes (in matchers.dart) likely used to define the atomic propositions based on application state.
• Status Reporting: CheckStatus enum to represent the outcome of a check.

Getting Started

(Note: This package is under active development and API details might change.)

Add this package along with its core dependencies to your pubspec.yaml:

dependencies:
  flutter:
    sdk: flutter
  temporal_logic_core: ^0.1.0 # Or latest
  temporal_logic_mtl: ^0.1.0 # Or latest
  temporal_logic_flutter: ^0.0.1-dev # Use the latest version from pub.dev or Git

Then run flutter pub get.

Usage (Conceptual)

While the exact API might evolve, the general usage pattern would involve:

1. Defining State Propositions: Use helpers (likely from matchers.dart or temporal_logic_core.builder) to create AtomicProposition instances based on your application state (e.g., from a BLoC, Riverpod provider, ValueNotifier, etc.).

   // Example using a hypothetical state class `CounterState`
   final isCounting = state<CounterState>((s) => s.isCounting);
   final countIsTen = state<CounterState>((s) => s.count == 10);
   

2. Defining Temporal Formulas: Construct LTL or MTL formulas using builders or constructors from the core/mtl packages.

   // LTL: It should eventually reach 10
   final ltlSpec = eventually(countIsTen);
   
   // MTL: Counting must start within 500ms
   final mtlSpec = EventuallyTimed(isCounting, TimeInterval.upTo(Duration(milliseconds: 500)));
   

3. Setting up a Stream Checker: Instantiate a checker (StreamLtlChecker, StreamMtlChecker, etc.) providing the formula and the stream of application state.

   // Assuming `myBloc.stream` provides a Stream<CounterState>
   final ltlChecker = StreamLtlChecker<CounterState>(
       formula: ltlSpec,
       stream: myBloc.stream,
   );
   
   // For MTL, the stream needs to provide timed values or be wrapped
   // final mtlChecker = StreamMtlChecker<CounterState>(...);
   

4. Observing Check Status: Listen to the checker's status stream (which likely emits CheckStatus) or use one of the provided *CheckerWidgets to display the status in the UI.

   // Using a widget
   MtlCheckerWidget(
     checker: mtlChecker,
     // Builder to display based on CheckStatus (success, failure, unknown)
   )
   
   // Or listening to the stream directly
   mtlChecker.statusStream.listen((status) {
     print('MTL Check Status: $status');
     if (status == CheckStatus.failure) {
       // Handle failure...
     }
   });
   

(Please refer to specific class documentation and examples once available for precise usage.)

Additional Information

• Dependencies: Relies heavily on temporal_logic_core for formula structure and temporal_logic_mtl for timed logic.
• State Management: Designed to work with streams, making it potentially compatible with various state management solutions (BLoC, Riverpod, etc.) that expose state via streams.
• Performance: Runtime checking can have performance implications. The efficiency of the checkers, especially for complex formulas and high-frequency streams, should be considered.

See the example/ directory in the main repository for potential usage scenarios.