[gt-users] extending python bindings

[Brent Pedersen]

On Fri, Nov 21, 2008 at 3:52 AM, Gordon Gremme <gremme at gmail.com> wrote:
>> i'm not sure how to handle the strand enum with ctypes. and i'm not
>> sure of the relation between the _api.h files and
>> the corresponding .h files...
>> any help on that or comments on the code are welcome.
>
> Some additional remarks on the _api.h files in general and the
> GtFeatureNode API in particular.
>
> With the _api.h files we try to solve the conflict between the need of
> stable APIs for libgenometools users and the need to evolve the
> implementation of libgenometools for GenomeTools developers.
>
> All interfaces which can be found in _api.h files define the exposed
> API of libgenometools.
> They should be documented and the documentation is used to generate
> the C API documentation at http://genometools.org/libgenometools.html
> We try to keep this public API fairly stable and all bindings should
> only use this API.
>
> In contrast, all interfaces which are not defined in the _api.h files
> are more likely to change, because we can be sure that there are no
> external dependencies which would break when changing them. This
> allows us to improve upon the APIs, since often a better API becomes
> clear after using the first implementation.
>
> We move APIs from the ``regular'' header files to the API headers if
> interfaces stabilize, but pretty much on an ``as needed'' basis.
> That is, if you see something in regular headers which would be nice
> to have in the public API, just let me know and we discuss moving it.
>
> Concerning the GtFeatureNode API: I will make many APIs public today
> to allow the manual construction of GtFeatureNodes on the binding
> level.
>
> Gordon
> _______________________________________________
> gt-users mailing list
> gt-users at genometools.org
> http://genometools.org/mailman/listinfo/gt-users
>

that's exactly the part i didnt follow. i was thinking of them as
fixed. makes perfect sense now.
thanks.
-brent