Identifying States #

Syntax of State Identifiers #

In describing a transition, we need a way to represent the target state.

A state is uniquely identified by the vector of its path from the root, with the first element being a special keyword :>.

For instance, if we have a state :s1 and it has a child state :s1.1, then they could be identified by [:> :s1] and [:> :s1 :s1.1] respectively.

There are other ways to represent a state in transition:

• If the first element of the vector of keywords is not :>, it represents a relative path.
• A keyword :foo is short for [:foo]
• If the first element of the vector of keywords is special keyword :., it represents a child state of the current state.
• A nil target represents a internal self-transition

Some examples:

{:states {:s1
          {:on {:event1_2 :s2 ;; (1)
                :event_1_1.1 [:. :s1.1]} ;; (2)
           :states {:s1.1
                    {:on {:event1.1_1.2 :s1.2
                          :event1.1_2 [:> :s2] ;; (3)
                          }}}}
          :s2
          {:on {:event_2_2 {:actions some-action}}}}} ;; (4)

(1) We want to represent state :s2 in the context of :s1, so we can simply write :s2. We could of course use the absolute syntax [:> :s2].

(2) :s1.1 is a child state of :s1, and we want to represent it in the context of :s1, which means we can use either the relative syntax [:. :s1.1] or the absolute syntax [:> :s1 :s1.1].

(3) We want to represent the state :s2 in the context of :s1.1, which means we had to use the absolute syntax here.

(4) When the target is not given (or nil), it means an internal self-transition.