On Sep 3, 1:37 pm, Georg Brandl <[email protected]> wrote:
> John H Palmieri schrieb:
>
>
>
>
>
>
>
> > On Sep 3, 1:02 pm, Georg Brandl <[email protected]> wrote:
> >> John H Palmieri schrieb:
>
> >> > Another question about the interaction between Sphinx and Sage. In
> >> > Sage, we use Python files, but also "Cython" files -- Cython is a
> >> > hybrid between Python and C which allows Python-like syntax and which
> >> > can also produce very fast code. Anyway, Cython files have suffice
> >> > ".pyx" and when they are compiled they produce ".so" files. In Sphinx
> >> > 0.5.1, the file "matrix0.so" was recognized as a dependency of
> >> > "matrix0.pyx", so when "matrix0.pyx" was changed, the autodoc
> >> > extension knew to rebuild that portion of the reference manual. In
> >> > Sphinx 0.6.2, "matrix0.so" is no longer recognized as a dependency.
> >> > (However, a file called "latex.py" has "latex.pyc" as a dependency,
> >> > and autodoc in either version of Sphinx recognizes (by comparing
> >> > modification times between the .py file and the .pyc file) when
> >> > latex.py has been changed.)
>
> >> > So: what changes were made in the dependency code between 0.5.1 and
> >> > 0.6.2, and what can I do to remedy this situation?
>
> >> Easy: this is a bug in Sphinx 0.6.x -- for C modules I forgot to add
> >> the underlying module.__file__ as a dependency.
>
> >> Should be fixed now in the 0.6 branch and trunk.
>
> > Thanks a lot.
>
> Least I can do. I also promise a 0.6.3 as soon as the memory usage issue
> is worked out.
>
> >> cheers,
> >> Georg
>
> >> PS: I've downloaded a binary Sage package and tried to build the docs,
> >> but I'm getting "Illegal instruction" errors when importing some sage
> >> modules. Even ``./sage -ba`` did not help :|
>
> > You could try
>
> > sage -f mpir-1.2.p4
> > sage -f atlas-3.8.3.p7
>
> > These commands rebuild these parts of Sage, and these seem to often
> > cause this error message when they're built on a processor which
> > doesn't quite match the one you're trying to run them on.
>
> Rebuilding mpir seems to have done the trick! Thanks very much.
>
> Georg
While we're thanking each other :), I've now had a chance to test your
patch, and it works just the way it used to. So thanks: we're one
step closer to including Sphinx 0.6.2 in Sage. We may use the
"weakref" fix mentioned in the other thread, since it seems to fix our
problem without causing any issues that we've noticed. and I think
it's our only remaining obstacle. But if you come up with another
idea, we'll be happy to try it out. If you're interested in seeing
some of the discussion from our end of things, visit the web page
<http://trac.sagemath.org/sage_trac/ticket/6586>
Regards,
John
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"sphinx-dev" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/sphinx-dev?hl=en
-~----------~----~----~----~------~----~------~--~---
