breily / states

DFA
===

* DFA lets you easily create deterministic finite automata in Ruby.
* You create a machine by giving it a list of states and then feeding
  it input.


Example
-------

* A DFA to represent open and closed doors:

    > require 'dfa'
    >
    > machine = automaton "Opened", {'close_door' => "Closed"},
    >                     "Closed", {'open_door' => "Opened"}
    >
    => <DFA: [Opened, Closed]>
    > machine.give "close_door"
    - exit state[Opened] >> input[close_door]
    + input[close_door] >> enter state[Closed]
    => <State: Closed :: {"open_door"=>"Opened"}>
    > machine.give "open_door"
    - exit state[Closed] >> input[open_door]
    + input[open_door] >> enter state[Opened]
    => <State: Opened :: {"close_door"=>"Closed"}>

* Further examples are in machines.rb.


Notes
-----

* The start state of the DFA is automatically the first state given.

* The DFA will continue to transition until an input doesn't match, at which
  point it returns nil instead of a State.

* The special symbol :dot will match any input, but would only be matched
  if no other input matches.

* The special state :error will raise an Exception if it is ever entered.

* See the 'nice_machine' for an example of both.

Actions
-------

* Each state has @onentry and @onexit procs that fire accordingly when you
  enter or exit the state.  Each gets the input causing the transition as
  a parameter.

* By default, the @onentry proc prints the lines from the above example that
  say "enter state[...]".  The @onexit proc prints the "exit state[...]" lines.

* To define these easily, use the 'state' function, which takes a state name
  and a block:
    
    state 1 do
        transition "c" => 2, :dot => :error
        on_exit { |i| puts "input causing exit is: #{i}" }
    end