CS2002 - Week 9 - State Machine Diagrams

Purpose of a state machine diagram

To model:

- possible states of a system/object

- transitions between states triggered by events

- behaviour in each state

Activities within a state

entry/Activity: runs on entering the state

exit/Activity: runs on exiting the state

do/Activity: runs while in the state

Components of a transition

Event (trigger)

Guard (boolean condition)

Effect (actions during transition)

Internal vs External transitions

Internal: No state change (e.g. event1 triggers Activity3 in state1) External: State change (e.g. event1 moves from state1 to state2)

Types of events

Signal event (e.g. rightMouseDown)

Call event (e.g. register(exam))

Time event (e.g. after(5 secs))

Initial state

A pseudo-state with:

- no incoming edges

- immediate transition to next state

- guards on outgoing edges must be mutually exclusive

Final state vs Terminate node

Final state: object remains here (end of lifecycle)

Terminate node: object is deleted

Decision node

Splits transitions into alternative paths with guards (e.g. [x > 0])

Parallelisation node

Splits control flow into concurrent sub-flows (1 input -> N outputs)

Synchronisation node

Merges concurrent sub-flows into one (N inputs -> 1 output)

Change events vs Guards

Change event: continuously checked (e.g. when(x > y))

Guard: checked only when event occurs

Model a LectureHall

[Free] -(occupy)-> [Occupied] [Occupied] -(release)-> [Free]

States

Represents conditions or situations during the lifecycle of an object

Transitions

Represents the movement from one state to another, triggered by events, guards or activities

Events

Triggers that cause a transition

Guards

Boolean conditions that must be true from a transition to occur

Activities

Actions executed when entering, exiting or during a state

When is a do-activity aborted?

When an event triggers a transition with no guard condition

Time event

An event like when(date=31.12.2001) that triggers at a specific time