Sphinx question: automethod and autofunction in integrals.rst

[Joachim Durchholz]

Hi all,

I need some feedback about whether what I see is intentional (and I'm
missing something), or just a bug.

The file in question is doc/src/modules/integrals/integrals.rst .

First an example of a section that seems correct to me:
========================================
Integral Transforms
-------------------

.. module:: sympy.integrals.transforms

SymPy has special support for definite integrals, and integral transforms.

.. autofunction:: mellin_transform
.. autofunction:: inverse_mellin_transform
========================================
It generates
http://docs.sympy.org/latest/modules/integrals/integrals.html#module-sympy.integrals.transforms
and lists e.g. mellin_transform as
"sympy.integrals.transforms.mellin_transform" .

Now the section that I do not understand:
========================================
API reference
-------------

.. automethod:: sympy.integrals.integrate

.. automethod:: sympy.integrals.line_integrate
========================================
Output is on
http://docs.sympy.org/latest/modules/integrals/integrals.html#api-reference
and lists e.g. integrate as "static integrals.integrate".
What I find dubious here is that integrate() is not a method, it is a
global function in module sympy.integrals.integrals(!).

I suspect that Sphinx thinks that "integrals" is a class and "integrate"
a method in it, but happens to get it right that "integrate" is indeed
static.
Is that correct? If yes, I'd replace those "automethod" entries with
appropriate "autofunction" ones.

[Aaron Meurer]

That's wrong. It works because technically functions are methods of
their module, but it really should use autofunction or autoclass. I
remember fixing a lot of this a long time ago but I guess some of them
are still there. It's disappointing that Sphinx doesn't issue any
warnings in this case.

Aaron Meurer

> --
> 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 sympy+un...@googlegroups.com.
> To post to this group, send email to sy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/sympy.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sympy/54F5FC43.6080205%40durchholz.org.
> For more options, visit https://groups.google.com/d/optout.

[Joachim Durchholz]

Am 03.03.2015 um 20:48 schrieb Aaron Meurer:
> That's wrong.

Good to know. I'm going to include the fix for this specifically as a
"windfall fix" in my next remove-C update (I need it done to get a
Sphinx build failure out of Travis).

> It works because technically functions are methods of
> their module, but it really should use autofunction or autoclass.

Thanks, that confirms my assumptions about what's going on.

> I
> remember fixing a lot of this a long time ago but I guess some of them
> are still there. It's disappointing that Sphinx doesn't issue any
> warnings in this case.

I sympathize.