Re: [pygccxml-development] C type normalization utility function?

[Gustavo C. <gjc...@gm...>]

2008/6/13 Gustavo Carneiro <gjc...@gm...>:

> 2008/6/11 Gustavo Carneiro <gjc...@gm...>:
>
>> 2008/6/10 Roman Yakovenko <rom...@gm...>:
>>
>>> On Mon, Jun 9, 2008 at 1:36 PM, Gustavo Carneiro <gjc...@gm...>
>>> wrote:
>>> > As seen in
>>> >
>>> >     https://bugs.launchpad.net/bugs/238455
>>> >
>>> > I am getting problems in PyBindGen related to differences in types like
>>> > "char*" vs "char *", also the position of the 'const' keyword may
>>> change,
>>> > etc.
>>> >
>>> > I am looking for some robust, pure python code to parse C types and
>>> extract
>>> > features from it, or produce a canonical type string, so that I can use
>>> it
>>> > PyBindGen.  Does something like this exist in pygccxml
>>>
>>> No, but we can discuss such addition to pygccxml. I think this is a
>>> nice & useful feature.
>>>
>>> > or elsewhere that I can steal? :-)
>>>
>>> May be from here: http://code.google.com/p/cppclean/wiki/InternalDesign
>>
>>
>> Thanks for the link!  This code looks nice, if excessive for what I want.
>> Using and adapting this code to parse C types is now on my medium priority
>> TODO list.  If I get something useful out of it, I'll post here the result.
>>
>
> OK, I have results :-)
>

Due to the mailing list rejecting ZIP files, the zip file I sent can be
found here:

http://telecom.inescporto.pt/~gjc/ctypeparser.zip

Regards.


>
>
> gjc@dark-tower:ctypeparser$ python ctypeparser.py
> 'char *' => 'char *'
> 'const foo::bar<const char*, zbr&>*' => 'foo::bar< char const *, zbr & >
> const *'
> 'const ::bar*' => '::bar const *'
> 'const char*const' => 'char const * const'
> 'const char*const*const' => 'char const * const * const'
>
>
> This code uses the tokenizer from cppclean project, with minor
> modifications.  On top of it it I wrote a minimal type parser which reorders
> 'const' modifiers to _a_ canonical position (to the right, in this case),
> and finally converts the result back to a string with deterministic spacing
> applied to tokens.  It also handles templated types recursively.
>
> Let me know what you think.
>
> If it is suitable, next step is add some unit tests and then you can copy
> it to pygccxml, if you want.  I'll add another copy to pybindgen.
>
> One thing I am not completely sure yet is how to handle namespaced
> identififers.  I am tempted to assume a context where no 'using namespace
> xxx' statment would have been used and assume all identifiers are relative
> to the root namespace.  On that assumption I would like to remove any
> leading :: in identifiers.  Any thoughts?
>
> Regards.
>
>
> --
> Gustavo J. A. M. Carneiro
> INESC Porto, Telecommunications and Multimedia Unit
> "The universe is always one step beyond logic." -- Frank Herbert
>



-- 
Gustavo J. A. M. Carneiro
INESC Porto, Telecommunications and Multimedia Unit
"The universe is always one step beyond logic." -- Frank Herbert