Re: [myhdl-list] MEP-108 Bundle

[Jan D. <ja...@ja...>]

On 06/24/2012 06:46 PM, Christopher Felton wrote:
> On 6/15/12 4:07 AM, Jan Decaluwe wrote:
>> On 06/01/2012 03:35 AM, Christopher Felton wrote:
>>
>>>> I think we should have a go at MEP 108 to remove the
>>>> confusing "class attribute" terminology. Perhaps the
>>>> explanations make it less clear than just stating what
>>>> this MEP is about. Perhaps we should just
>>>> talk about a method at the top level - I understand
>>>> methods at lower levels already worked without this MEP.
>>>>
>>>
>>> Sounds good, I should have some time tomorrow to work
>>> through the modifications you suggested.
>>
>> About the "self" argument: the MEP says that the
>> implementation looks for this argument and assumes it
>> has the usual meaning. This is just fine with me,
>> provided it is the *first* argument as per convention.
>> Probably this is what the implemenation does? If not,
>> it should, and the documentation should be adapated
>> also.
>>
>>
>
> There are two locations in the code that were modified.  First, the
> top-level visitor.  These modifications didn't require the use of the
> "self".
>
> The second in the extract hierarchy , the original code uses a different
> mechanism (not ast) to determine if user defined code is present.  The
> modifications here do not enforce the "self" as the first argument.  The
> code where the changes were added uses the python profile extraction to
> identify a function on a return event.  For a method, the "func" from
> the global space would be None, if that was the case then self was
> retrieved from the local space and funcname was checked as an attribute.
>
> What does this all mean, it means it doesn't enforce "self" to be the
> first position but it will only use self if it has the /function/ that
> it is looking for (funcname).  If the code fails to find the funcname in
> the global space or as an attribute (method) of "self" in the localspace.

Perhaps that is an indirect check of the conventional use of
self, including the position?
  
> I will investigate if there is a mechanism to also verify "self" is the
> first argument.  I tried to find a means to not use "self" but with the
> profile extraction I was not successful.

Mm, in any case it looks like a very minor issue (if any) and I don't
want to cause too much extra effort on it.

-- 
Jan Decaluwe - Resources bvba - http://www.jandecaluwe.com
     Python as a HDL: http://www.myhdl.org
     VHDL development, the modern way: http://www.sigasi.com
     World-class digital design: http://www.easics.com