[C++-sig] Lost Again! :)

[Roman Yakovenko]

On 9/1/06, Nicolas Tessore <wz_ at gmx.net> wrote:
> Nicolas Tessore wrote:
> >> from pygccxml import declarations
> >>
> >> def rename_cast( cast ):
> >>     cls_to = declarations.class_traits.get_declaration( cast.return_type )
> >>     cast.rename( 'DynamicCastTo' + cls_to.name )
> >>
> >> casts = mb.free_functions ( lambda decl: decl.name.startswith(
> >> 'DynamicCast<' ) )
> >> map( rename_cast, casts )
> >
> > This always leaves me with a "query returned no matches" error :(
> Ok, I figured this out myself. GCCXML does not save the template part of
> the name, so all those functions are called "DynamicCast". I slightly
> modified your approach:
>
>    01:  #Find object type in Casts. Const T casts are ignored
>    02:  pattern = "::Mylib::Ref<Mylib::(.*)>"
>    03:  p = re.compile(pattern)
>    04:
>    05:  def rename_casts( cast ):
>    06:      m = p.match(cast.return_type.decl_string)
>    07:      if m:
>    08:          cast.rename( cast.name + 'To' + m.groups()[0] )
>    09:      else: #DynamicCast should be excluded
>    10:          cast.exclude()
>    11:
>    12:  casts = mb.free_functions ( lambda decl: ( decl.name ==
> 'DynamicCast' or decl.name == 'StaticCast' ) )
>    13:  map( rename_casts, casts )
>
> This works beautifully :) But is it a correct approach?

If it works, that the answer is yes.

> And it clearly
> shows I barely scratched the surface of the power that Py++/ pygccxml
> hold :/

Please take a look on
http://language-binding.net/pygccxml/apidocs/pygccxml.declarations.templates-module.html
The functionality provided by this module will do the job for you.

But you ignored my advice:

cls_to = declarations.class_traits.get_declaration( cast.return_type )
cast.rename( 'DynamicCastTo' + cls_to.name )

You don't have to parse, just extract class declaration from DynamicCast return
type.

> Now while still giggling at the wonderful result I saw, it happened that
> I noticed my typedefs à la
>
>      typedef Mylib::Ref<Mylib::MyClass> MyClassRef;
>      sizeof(MyClassRef);
>
> are not automatically detected and thus, the Mylib::Ref<Mylib::MyClass>
> is still named Mylib_scope_Ref_less_than_Mylib_scope_MyClass_grate etc
> in the generated binding source. Any clues why that could be? It is not
> a problem though, I will just apply the same method as above to them.

Read this: http://language-binding.net/pyplusplus/documentation/hints.html
It should explain why Py++ does not work as you expects.

-- 
Roman Yakovenko
C++ Python language binding
http://www.language-binding.net/