I had a bit of trouble with codegen here:
https://github.com/sympy/sympy/pull/18116

The CodegenArrayContraction class is non-Expr but is used in Mul.

On Tue, 24 Dec 2019 at 17:46, Aaron Meurer <[email protected]> wrote:
>
> What sort of troubles?
>
> Aaron Meurer
>
> On Tue, Dec 24, 2019 at 5:44 AM Francesco Bonazzi <[email protected]> 
> wrote:
>>
>> It looks like the codegen module is really causing troubles to mypy. The 
>> module is a real mess, we'll have to review it one day.
>>
>>
>> On Monday, 23 December 2019 01:53:58 UTC+1, Aaron Meurer wrote:
>>>
>>> We already dropped 3.4 support after the 1.4 release. See
>>> https://github.com/sympy/sympy/wiki/Release-Notes-for-1.4 and
>>> https://github.com/sympy/sympy/wiki/Python-version-support-policy.
>>>
>>> I agree that mypy will not be easy. I also suspect there would be
>>> issues getting mypy to work with the SymPy type system. I'm not sure
>>> how well they support custom types. It would be a good GSoC project or
>>> if an existing dev wants to work on it, a good proposal for a NumFOCUS
>>> small development grant.
>>>
>>> Aaron Meurer
>>>
>>> On Sun, Dec 22, 2019 at 5:32 PM Oscar Benjamin
>>> <[email protected]> wrote:
>>> >
>>> > I think it would be good to use mypy but it will be a lot of work just
>>> > to get working on a small part of the codebase because so much of
>>> > SymPy is inconsistent about types right now. Ultimately it's not just
>>> > a case of "using mypy" but actually fixing the inconsistencies which
>>> > is a big task. Perhaps this could be suggested as an idea for a GSOC
>>> > proposal...
>>> >
>>> > On Sun, 22 Dec 2019 at 16:16, Francesco Bonazzi <[email protected]> 
>>> > wrote:
>>> > >
>>> > > Python 3.4 adds support for singledispatch
>>> > > Python 3.5 adds support for typing: 
>>> > > https://docs.python.org/3.5/library/typing.html
>>> > > Add mypy to Travis tests.
>>> > >
>>> > > The typing module adds support for type annotations of various kinds to 
>>> > > Python. Using mypy in Travis would enforce type-checking.
>>> > >
>>> > > Type declarations would be beneficial to:
>>> > >
>>> > > analyze the code and find potential bugs
>>> > > serve as a first step before translating the code to C++
>>> > >
>>> > >
>>> > > --
>>> > > You received this message because you are subscribed to the Google 
>>> > > Groups "sympy" group.
>>>
>>> > > To unsubscribe from this group and stop receiving emails from it, send 
>>> > > an email to [email protected].
>>> > > To view this discussion on the web visit 
>>> > > https://groups.google.com/d/msgid/sympy/c7640d2a-802a-4ec9-890f-66a18d202395%40googlegroups.com.
>>> >
>>> > --
>>> > You received this message because you are subscribed to the Google Groups 
>>> > "sympy" group.
>>> > To unsubscribe from this group and stop receiving emails from it, send an 
>>> > email to [email protected].
>>> > To view this discussion on the web visit 
>>> > https://groups.google.com/d/msgid/sympy/CAHVvXxTH0JiGDmemOmC-nEFzJTKdPkkAC5d06rRVQydg%3DgyZtg%40mail.gmail.com.
>>
>> --
>> You received this message because you are subscribed to the Google Groups 
>> "sympy" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to [email protected].
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/sympy/54220440-0069-4481-89f5-99a5a48728c5%40googlegroups.com.
>
> --
> You received this message because you are subscribed to the Google Groups 
> "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to [email protected].
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sympy/CAKgW%3D6Jcf%3Dh0kJmLgAfsM3X7RDpghb3XLbicj6utzK3My-Fm5g%40mail.gmail.com.

-- 
You received this message because you are subscribed to the Google Groups 
"sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sympy/CAHVvXxSGHa9QWh4KKSKVDvNpkmiE7nae4nCkyxcZggZM5ggUzQ%40mail.gmail.com.

Reply via email to