On Mon, Oct 1, 2012 at 6:41 PM, wang yunbo <[email protected]> wrote:
> Hi Dr. Wheeler,
>
> 1. lsmlib.pxd and lsmlibpyx are in the same directory.
>
> 2. When I removed /home/ba01/u142/wang1320/.pyxbld and re-ran the test, I got 
> exactly the same error (compared by diff command).
>
> The system generated another .pyxbld under my home directory automatically 
> when I imported pylsmlib. That's weird.

Pyximport stores its intermediate files in .pyxbld so it is good that
it is being created. I just wanted you to delete it to see if there
was any old crud interfering with the compiling and linking in some
way. It was just shot in the dark. Going forward here are some ideas
for trying to figure out what is happening:

  * Make a simple hello_world() function in c (compile it and link
against it) and use a .pxd file to define it for cython and then call
it in a .pyx file (don't import it directly in the .pyx file, but only
define it in the .pxd file). Basically try an recreate the situation,
but without using LSMLIB, You can then ask on the Cython mailing list
with a very simple example if it doesn't work.

  * Try building using setup.py rather than using pyximport just to
change things up a bit. It could give you some insight into the
cimport errror.

  * Use my latest reincarnation of Pylsmlib
<https://github.com/wd15/LSMLIB/tree/fipy/pylsmlib>. This is still
under construction, but it will give you some ideas. I'm using
setup.py rather than pyximport.

Cheers

-- 
Daniel Wheeler
_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
  [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]

Reply via email to