Now thats a really nice explanation of union versus intersection in termes 
of types. 

Thank you very much for that,
Sven

Am Donnerstag, 16. April 2015 23:25:42 UTC+2 schrieb Ambrose 
Bonnaire-Sergeant:
>
> It might help thinking in terms of Java interfaces, Foo and Bar.
>
> (definterface Foo
>   (foo []))
> (definterface Bar
>   (bar []))
>
> (I Foo Bar) is a value that extends both Foo and Bar.
>
> (deftype IImp []
>   Foo
>   (foo [this])
>   Bar
>   (bar [this]))
>
> (->IImp) is of type Foo, Bar, (I Foo Bar) and (U Foo Bar).
>
> Assuming we assign (->IImp) the type (I Foo Bar), we can call these safely:
>
> (let [i :- (I Foo Bar), (->IImp)]
>  (.foo i)
>  (.bar i))
>
> A type that just implements Foo is not a Bar, so we can't claim it's a Foo 
> *and* a Bar.
>
> (deftype UImp []
>   Foo
>   (foo [this]))
>
> (->UImp) is of type Foo, and (U Foo Bar).
>
> Assuming we assign (->UImp) the type (U Foo Bar), the same operations now 
> must cast at runtime.
>
> (let [i :- (U Foo Bar), (->UImp)]
>  (if (instance? Foo)
>   (.foo i)
>   (.bar i))
>
>
> Thanks,
> Ambrose
>
> On Thu, Apr 16, 2015 at 5:15 PM, Sven Richter <sve...@googlemail.com 
> <javascript:>> wrote:
>
>> I meant when I change:
>>
>> (t/HVec [Keyword (t/U Keyword (t/HVec [Keyword Number])) t/Any t/Any *])  
>> to  (t/HVec [Keyword (t/I Keyword (t/HVec [Keyword Number])) t/Any t/Any *])
>>                                 
>> ^^                                                                           
>>               
>> to                                  ^^
>> Still I think that making an intersection of it is wrong conceptually.
>>
>>
>> Thanks,
>> Sven
>>
>> Am Donnerstag, 16. April 2015 23:08:30 UTC+2 schrieb Ambrose 
>> Bonnaire-Sergeant:
>>>
>>> I don't see an intersection, what do you mean?
>>>
>>> Thanks,
>>> Ambrose
>>>
>>> On Thu, Apr 16, 2015 at 4:43 PM, Sven Richter <sve...@googlemail.com> 
>>> wrote:
>>>
>>>> Hi,
>>>>
>>>> I tried both destructuring and the nth form instead of second and 
>>>> first. None of which worked.
>>>>
>>>> However, if I change the Union to Intersection in 
>>>>
>>>> (t/HVec [Keyword (t/U Keyword (t/HVec [Keyword Number])) t/Any t/Any *])
>>>>
>>>> it works for the definition of the multimethod. Does that make sense? I 
>>>> thought Union was either one type or the other.
>>>>
>>>> Thanks,
>>>> Sven
>>>>
>>>>
>>>> Am Donnerstag, 16. April 2015 21:44:25 UTC+2 schrieb Ambrose 
>>>> Bonnaire-Sergeant:
>>>>>
>>>>> I don't think second's type is is smart enough.
>>>>>
>>>>> Try using nth or destructuring instead:
>>>>>
>>>>> (let [[f s] v]
>>>>>   (if (vector? s) (first s) s))
>>>>>
>>>>> or
>>>>>
>>>>> (if (vector? (nth v 1)) (first (nth v 1)) s)
>>>>>
>>>>> Thanks,
>>>>> Ambrose
>>>>>
>>>>> On Thu, Apr 16, 2015 at 3:39 PM, Sven Richter <sve...@googlemail.com> 
>>>>> wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> I have this code:
>>>>>>
>>>>>> (defalias html-label (t/HVec [Keyword (t/HMap :mandatory {:for String
>>>>>> }) String]))
>>>>>> (defalias html-form (t/HVec [Keyword (t/HMap :mandatory {:id String}) 
>>>>>> t/Any *]))
>>>>>> (defalias html-form-group (t/HVec [html-label html-form]))
>>>>>>  
>>>>>> (t/ann dt->hiccup [(HVec [Keyword (U Keyword (HVec [Keyword Number])) 
>>>>>> t/Any t/Any *]) -> 
>>>>>>                                html-form-group])
>>>>>> (defmulti dt->hiccup (t/fn [col :- pt/et-column]
>>>>>>                        (if (vector? (second col))
>>>>>>                          (first (second col))
>>>>>>                          (second col))))
>>>>>>
>>>>>> And here comes the error message when checking this function:
>>>>>>
>>>>>> Type Error (leiningen/td_to_hiccup.clj:25:34) Polymorphic function 
>>>>>> first could not be applied to arguments:
>>>>>> Polymorphic Variables:
>>>>>>         x
>>>>>>  
>>>>>> Domains:
>>>>>>         (t/HSequential [x t/Any *])
>>>>>>         (t/Option (t/EmptySeqable x))
>>>>>>         (t/NonEmptySeqable x)
>>>>>>         (t/Option (clojure.lang.Seqable x))
>>>>>>  
>>>>>> Arguments:
>>>>>>         (t/U Keyword (t/HVec [clojure.lang.Keyword java.lang.Number])
>>>>>> )
>>>>>>  
>>>>>> Ranges:
>>>>>>         x :object {:path [(Nth 0)], :id 0}
>>>>>>         nil
>>>>>>         x
>>>>>>         (t/Option x)
>>>>>>  
>>>>>> in: (first (second col))
>>>>>> in: (first (second col))
>>>>>>  
>>>>>>  
>>>>>> ExceptionInfo Type Checker: Found 1 error  clojure.core/ex-info (core
>>>>>> .clj:4403)
>>>>>>
>>>>>> My assumption is that the check (if (vector? (second col will return 
>>>>>> only true if it is this type: (HVec [Keyword Number]). However, I 
>>>>>> have the impression that core.typed does not resolve the if expression 
>>>>>> properly.
>>>>>> Or maybe I am missing something completely.
>>>>>>
>>>>>> Any hints or recommendations?
>>>>>>
>>>>>> Thanks,
>>>>>> Sven
>>>>>>
>>>>>>
>>>>>>  -- 
>>>>>> You received this message because you are subscribed to the Google
>>>>>> Groups "Clojure" group.
>>>>>> To post to this group, send email to clo...@googlegroups.com
>>>>>> Note that posts from new members are moderated - please be patient 
>>>>>> with your first post.
>>>>>> To unsubscribe from this group, send email to
>>>>>> clojure+u...@googlegroups.com
>>>>>> For more options, visit this group at
>>>>>> http://groups.google.com/group/clojure?hl=en
>>>>>> --- 
>>>>>> You received this message because you are subscribed to the Google 
>>>>>> Groups "Clojure" group.
>>>>>> To unsubscribe from this group and stop receiving emails from it, 
>>>>>> send an email to clojure+u...@googlegroups.com.
>>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>>
>>>>>
>>>>>  -- 
>>>> You received this message because you are subscribed to the Google
>>>> Groups "Clojure" group.
>>>> To post to this group, send email to clo...@googlegroups.com
>>>> Note that posts from new members are moderated - please be patient with 
>>>> your first post.
>>>> To unsubscribe from this group, send email to
>>>> clojure+u...@googlegroups.com
>>>> For more options, visit this group at
>>>> http://groups.google.com/group/clojure?hl=en
>>>> --- 
>>>> You received this message because you are subscribed to the Google 
>>>> Groups "Clojure" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>> an email to clojure+u...@googlegroups.com.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>
>>>  -- 
>> You received this message because you are subscribed to the Google
>> Groups "Clojure" group.
>> To post to this group, send email to clo...@googlegroups.com 
>> <javascript:>
>> Note that posts from new members are moderated - please be patient with 
>> your first post.
>> To unsubscribe from this group, send email to
>> clojure+u...@googlegroups.com <javascript:>
>> For more options, visit this group at
>> http://groups.google.com/group/clojure?hl=en
>> --- 
>> You received this message because you are subscribed to the Google Groups 
>> "Clojure" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to clojure+u...@googlegroups.com <javascript:>.
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your 
first post.
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"Clojure" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to clojure+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to