Re: [sage-devel] Re: PY_NEW(Integer) and Integer.__new__(Integer)

[Erik Bray]

On Mon, Apr 9, 2018 at 10:27 AM, Jean-Pierre Flori  wrote:
> I thought PY_NEW was to be avoided (except for classes overriding tp_new).
> See https://trac.sagemath.org/ticket/17668 and
> https://github.com/cython/cython/wiki/FAQ#can-cython-create-objects-or-apply-operators-to-locally-created-objects-as-pure-c-code

Which is exactly what we're doing for Integer.  In the ticket you cited:

"Change this everywhere except for Integer and RealDoubleElement
classes, which use a custom hooked tp_new that Cython isn't aware of."


> On Sunday, April 8, 2018 at 9:41:17 AM UTC+2, vdelecroix wrote:
>>
>> Dear all,
>>
>> What is the status of Integer.__new__(Integer) vs PY_NEW(Integer). One
>> can find both versions in the source code
>>
>> $ grep -R "PY_NEW(Integer)" | wc -l
>> 152
>> $ grep -R "Integer.__new__(Integer)" | wc -l
>> 94
>>
>> But the doc of PY_NEW (ext/stdsage.pxd) says that we should avoid the
>> second one
>>
>>  Return ``t.__new__(t)``.  This works even for types like
>>  :class:`Integer` where we change ``tp_new`` at runtime (Cython
>>  optimizations assume that ``tp_new`` doesn't change).
>>
>> Should I change the 94 bad usage?
>>
>> Vincent
>>
>> PS: this happen to be a question from
>> https://trac.sagemath.org/ticket/24717
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: PY_NEW(Integer) and Integer.__new__(Integer)

[Jean-Pierre Flori]

I thought PY_NEW was to be avoided (except for classes overriding tp_new).
See https://trac.sagemath.org/ticket/17668 and 
https://github.com/cython/cython/wiki/FAQ#can-cython-create-objects-or-apply-operators-to-locally-created-objects-as-pure-c-code

On Sunday, April 8, 2018 at 9:41:17 AM UTC+2, vdelecroix wrote:
>
> Dear all, 
>
> What is the status of Integer.__new__(Integer) vs PY_NEW(Integer). One 
> can find both versions in the source code 
>
> $ grep -R "PY_NEW(Integer)" | wc -l 
> 152 
> $ grep -R "Integer.__new__(Integer)" | wc -l 
> 94 
>
> But the doc of PY_NEW (ext/stdsage.pxd) says that we should avoid the 
> second one 
>
>  Return ``t.__new__(t)``.  This works even for types like 
>  :class:`Integer` where we change ``tp_new`` at runtime (Cython 
>  optimizations assume that ``tp_new`` doesn't change). 
>
> Should I change the 94 bad usage? 
>
> Vincent 
>
> PS: this happen to be a question from 
> https://trac.sagemath.org/ticket/24717 
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.