throttle

Next to handling concurrencies you can also throttle the the execution of a rule. But that also means, that your execution gets delayed by the numer of millisecond you set. Normally you could handle concurrency way better by the concurrency key. But there can be situations where it might be usefull to not execute the consequence on every action. E.g you want to create a autocomplete feature:

addRule({
  id: 'search/AUTOCOMPLETE',
  target: 'search/SET_SEARCH',
  throttle: 300,
  concurrency: 'SWITCH',
  consequence: action => api.searchSuggestions(action.payload).then(
    suggestions => actions.setSuggestions(suggestions),
    error => actions.setSuggestions([])
  )
})

dispatch({type:'search/SET_SEARCH', payload: 'r'}) 
await wait(200)
dispatch({type:'search/SET_SEARCH', payload: 're'})
await wait(200) // consequence gets called with payload 're'
dispatch({type:'search/SET_SEARCH', payload: 'red'})
await wait(200) 
dispatch({type:'search/SET_SEARCH', payload: 'redu'})
await wait(200) // consequence gets called with payload 'redu'
dispatch({type:'search/SET_SEARCH', payload: 'redux'})
await wait(400) // consequence gets called with payload 'redux'

Throttle works as follows: Whenever the target gets dispatched nothing happens, but the rule will wait until the throttle time is over. When the the same target gets dispatched again the old consequence won't get called and the timer remains untouched (timer won't reset to zero). Whenever the timer reaches its limit the last consequence gets executed