Features¶

This document provides a comprehensive overview of routilux features.

Core Features¶

Routine Base Class¶

✅ Support for defining 0-N slots (input slots)
✅ Support for defining 0-N events (output events)
✅ Provides stats() method returning state dictionary
✅ Supports serialization
✅ Automatically records emitted events to execution history

Slot (Input Slot)¶

✅ Can connect to multiple events (many-to-many relationship)
✅ Supports data reception and processing
✅ Supports merge strategies (override, append, custom)
✅ Intelligent parameter matching (supports various handler signatures)
✅ Exception handling (handler exceptions don’t interrupt flow)

Event (Output Event)¶

✅ Can connect to multiple slots (many-to-many relationship)
✅ Supports event triggering and data passing
✅ Passes data through Connection with parameter mapping
✅ Automatically recorded to execution history

Connection¶

✅ Connects event to slot
✅ Supports parameter mapping
✅ Automatically applies parameter mapping

Flow (Workflow Manager)¶

✅ Manages multiple Routine nodes
✅ Manages connections between nodes
✅ Executes workflows (execute)
✅ Resumes execution (resume)
✅ Supports serialization/deserialization
✅ Execution tracking (ExecutionTracker)

JobState (Job State)¶

✅ Records flow execution state
✅ Records state of each routine
✅ Records execution history (ExecutionRecord)
✅ Supports serialization/deserialization

ExecutionTracker¶

✅ Tracks routine execution
✅ Tracks event flow
✅ Performance metrics (execution time, success rate, etc.)
✅ Supports getting routine and flow performance metrics

Advanced Features¶

Event-Driven Execution¶

✅ Event-driven execution mechanism
✅ Flow context automatically passed
✅ Data passed through Connection with parameter mapping

Intelligent Parameter Matching¶

✅ Automatic handler parameter matching
✅ Supports **kwargs, single parameter, multiple parameters
✅ Automatically extracts values from dictionaries

Execution Tracking¶

✅ Automatically records all emitted events
✅ Records execution time
✅ Records start/end times
✅ Performance metrics statistics

Error Handling¶

✅ Automatic error capture and recording
✅ Error information recorded to JobState
✅ Errors recorded to execution history
✅ Multiple error handling strategies (STOP, CONTINUE, RETRY, SKIP)

Persistence¶

✅ Complete JobState persistence (serialize/deserialize)
✅ Flow structure persistence (serialize/deserialize)
✅ Supports JSON format
✅ Full serialization support for all core classes

Error Handling Strategies¶

STOP Strategy¶

Stop execution immediately when an error occurs.

CONTINUE Strategy¶

Continue execution despite errors, logging them for review.

RETRY Strategy¶

Retry the failed routine with configurable:

Maximum retry attempts
Retry delay
Retry backoff multiplier
Retryable exception types

SKIP Strategy¶

Skip the failed routine and continue with the next routine.

Pause and Resume¶

✅ Pause Functionality: Flow.pause() supports pausing execution
✅ Resume Functionality: Flow.resume() supports resuming execution
✅ Cancel Functionality: Flow.cancel() supports cancelling execution
✅ Pause Point Recording: Automatically records pause points and checkpoint information to JobState
✅ Design Optimization: Execution control (pause/resume/cancel) handled by Flow, JobState only responsible for state recording

Concurrent Execution¶

✅ Unified Event Queue: Both sequential and concurrent modes use the same queue-based mechanism
✅ Non-blocking emit(): Event emission returns immediately, tasks execute asynchronously
✅ Fair Scheduling: Tasks are processed fairly, preventing long chains from blocking shorter ones
✅ Concurrent Execution Strategy: Support for parallel execution using thread pools (controlled by max_workers)
✅ Event Loop: Background thread processes tasks from the queue
✅ Thread Pool Management: Configurable thread pool size (max_workers)
✅ Dependency Handling: Automatic dependency detection and waiting
✅ Thread Safety: All state updates are thread-safe
✅ Performance Optimization: Significant speedup for I/O-bound operations
✅ Strategy Switching: Dynamic switching between sequential and concurrent execution
✅ Serialization Support: Concurrent flows can be serialized and deserialized
✅ Task Completion Tracking: Automatic tracking of active concurrent tasks
✅ Wait for Completion: wait_for_completion() method to wait for all tasks with optional timeout
✅ Resource Cleanup: shutdown() method for proper thread pool cleanup

Concurrent Execution Details¶

Execution Strategies: - sequential: Routines execute one after another (default) - concurrent: Routines execute in parallel using thread pools

Key Features: - Automatic detection of parallelizable routines - Thread pool management with configurable size - Thread-safe state management - Error handling in concurrent scenarios - Performance monitoring for concurrent execution

Use Cases: - Multiple API calls that can run in parallel - Concurrent data fetching from multiple sources - Parallel processing of independent tasks - I/O-bound operations that benefit from concurrency