undetermine type

[Marcus.CM]

def hello (rcptto) :
    assert(isinstance(rcptto, list))          
    listrcpt = rcptto[:]  

The listrcpt is also marked as undetermine type in the sample above. The
ide needs to detect this kind of usage.


_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide

[Lawrence Oluyede-2]

On Fri, Jul 4, 2008 at 9:56 AM, Marcus Low <marcus@...> wrote:
> def hello (rcptto) :
>   assert(isinstance(rcptto, list))             listrcpt = rcptto[:]
> The listrcpt is also marked as undetermine type in the sample above. The ide
> needs to detect this kind of usage.
>

How it should? Asserting something is a istance of list doesn't mean
that in the line below you will use the variable as one.
By the way this code breaks if rcptto is not a list but something with
slicing support.

--
Lawrence, stacktrace.it - oluyede.org - neropercaso.it
"It is difficult to get a man to understand
something when his salary depends on not
understanding it" - Upton Sinclair
_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide

[Marcus.CM]

 listrcpt = rcptto[:]

This is valid list copy. The statement "assert(isinstance(rcptto,
list))" is only added there to HELP wing-ide to recognize that rcptto is
a list. Else i wont be bother putting it.
Now if wing-ide recognize now that rcptto is a LIST then my next
statement "listrcpt = rcptto[:]"  should be recognize as well.
If i am not using listrcpt as a list here, i dont know what other
possible use you are talking about given just those codes.

>By the way this code breaks if rcptto is not a list but something with
>licing support.

I already said the "assert" is for the sake of the IDE and the user of this function gurantees its a list. I hope you are in the right forum.



Lawrence Oluyede wrote:

> On Fri, Jul 4, 2008 at 9:56 AM, Marcus Low <marcus@...> wrote:
>  
>> def hello (rcptto) :
>>   assert(isinstance(rcptto, list))             listrcpt = rcptto[:]
>> The listrcpt is also marked as undetermine type in the sample above. The ide
>> needs to detect this kind of usage.
>>
>>    
>
> How it should? Asserting something is a istance of list doesn't mean
> that in the line below you will use the variable as one.
> By the way this code breaks if rcptto is not a list but something with
> slicing support.
>
>  

_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide

[Lawrence Oluyede-2]

On Fri, Jul 4, 2008 at 10:33 AM, Marcus Low <marcus@...> wrote:

> listrcpt = rcptto[:]
>
> This is valid list copy. The statement "assert(isinstance(rcptto, list))" is
> only added there to HELP wing-ide to recognize that rcptto is a list. Else i
> wont be bother putting it.
> Now if wing-ide recognize now that rcptto is a LIST then my next statement
> "listrcpt = rcptto[:]"  should be recognize as well.
> If i am not using listrcpt as a list here, i dont know what other possible
> use you are talking about given just those codes.
>

I think I did not explain myself correctly. I think WingIDE can infer
if a variable is "something" by how you initialize it. I doubt that
asserting something will help hi.
I'd try with

listrcpt = list(rcptto)

without the assert


--
Lawrence, stacktrace.it - oluyede.org - neropercaso.it
"It is difficult to get a man to understand
something when his salary depends on not
understanding it" - Upton Sinclair
_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide

[Wingware Support]

Marcus Low wrote:
> def hello (rcptto) :
>    assert(isinstance(rcptto, list))              listrcpt = rcptto[:]
> The listrcpt is also marked as undetermine type in the sample above. The
> ide needs to detect this kind of usage.

Yes, I believe Wing is missing x = y[:] as an inferencing opportunity.  We hope
to improve this in the future.  Note that x = list(y) would work in this case
but of course isn't helpful if you want something like x = y[:-3]

--

Stephan Deibel
Wingware | Python IDE
Advancing Software Development

www.wingware.com

_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide

[Lawrence Oluyede-2]

On Fri, Jul 4, 2008 at 3:21 PM, Wingware Support <support@...> wrote:
> Yes, I believe Wing is missing x = y[:] as an inferencing opportunity.  We
> hope
> to improve this in the future.  Note that x = list(y) would work in this
> case
> but of course isn't helpful if you want something like x = y[:-3]

Yep. I used list() because [:] does the same. Create a copy of the list.
Otherwise if the list is modified afterwards also the reference passed would.
Just assumed the behavior :)

--
Lawrence, stacktrace.it - oluyede.org - neropercaso.it
"It is difficult to get a man to understand
something when his salary depends on not
understanding it" - Upton Sinclair
_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide

[Marcus.CM]

Hi,

Is there a way to manually add this to the IDe, somekind of config file
or macro? I mean putting things like assert(isinstance, and  changing
coding style to fit the ide is kinda weird lingo to me, coming from VS
environment.
The are many styles in python i wish to use as is and which i really do
like the style for eg :

 l = o[:] instead of l = list(o)
 l = (1,2,3) instead of l = tuple((1,2,3))

- but of course isn't helpful if you want something like x = y[:-3]

exactly, the [] is used so naturally in python that i would rather stick to that
then changing for the ide sake. At the end of the day, its makes a difference when i am reading the whole code.


Marcus.



Lawrence Oluyede wrote:

> On Fri, Jul 4, 2008 at 3:21 PM, Wingware Support <support@...> wrote:
>  
>> Yes, I believe Wing is missing x = y[:] as an inferencing opportunity.  We
>> hope
>> to improve this in the future.  Note that x = list(y) would work in this
>> case
>> but of course isn't helpful if you want something like x = y[:-3]
>>    
>
> Yep. I used list() because [:] does the same. Create a copy of the list.
> Otherwise if the list is modified afterwards also the reference passed would.
> Just assumed the behavior :)
>
>  

_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide

[Wingware Support]

Marcus Low wrote:
> Is there a way to manually add this to the IDe, somekind of config file
> or macro? I mean putting things like assert(isinstance, and  changing
> coding style to fit the ide is kinda weird lingo to me, coming from VS
> environment.

This is something that requires modification to the source analyzer and
is not something that can be done through a config file or a script.  We
do know of this limitation and that you (& others) would like to see it
fixed; we'll try to fix it in a future release.

Thanks,

John
_________________________________________________
Wing IDE users list
http://wingware.com/lists/wingide