[Pyrex] Questions about your approach

[Robert Bradshaw]

Greg Ewing would be able to give the authoritative response to these  
questions, but I'll try and answer them to the best of my ability.

On Apr 10, 2008, at 9:54 AM, Paul Biggar wrote:
> Hi,
>
> I'm writing a paper about our PHP compiler, phc. In our 'related work'
> section, I'd like to discuss pyrex, so I have a few questions. I've
> looked through your documentation and releases, but the answers
> haven't been apparent.
>
> Did pyrex originally compile standard python, or was it annotated with
> types from the start? If the former, why the change?

 From my understanding, it always had type annotations, because its  
original purpose was to make it easy to wrap other libraries (for  
which type annotations were necessary).

> Has you run benchmarks on the compiled applications? How about when
> you try to compile plain python? What's the performance like?

See the thread at http://www.mail-archive.com/cython- 
dev at codespeak.net/msg00104.html

> Are there any papers describing python? I've seen a lot of citations
> of the web page, which leads me to suspect there aren't.

Not that I am aware of.

> I wonder if you have any comments on your approach vs. that of the
> defunct python2c?

Pyrex works, python2c doesn't. The big difference in approach is that  
Pyrex calls the Python library to do pythonic things rather then try  
to emulate its behavior in a pure C file, and so integrates very well  
with and in a Python environment.

- Robert