wxruby users - Sep 2006 - wxImage patches

These patches allow the RMagick sample to work with wxRuby2.  They also 
open up some functions that were %ignored that probably shouldn''t be
(at
least with the typemap adjustment).

I think we may need to visit (or document carefully) what happens when 
you pass in a string of data to the constructor or to SetAlpha.  We need 
to decide which setting of static_data is correct for the library.

Does anyone know of any reason we would or wouldn''t want to just set
the
unsigned char * typemap in typemap.i?

Roy



_______________________________________________
wxruby-users mailing list
wxruby-users@rubyforge.org
http://rubyforge.org/mailman/listinfo/wxruby-users

Roy Sutton wrote:
> These patches allow the RMagick sample to work with wxRuby2. 
Committed, thanks. I get several warnings now when compiling Image.cpp 
of the type:

src/Image.cpp: In function `VALUE _wrap_new_wxImage__SWIG_3(int, VALUE*, 
long
   unsigned int)'':
src/Image.cpp:2560: warning: invalid conversion from `char*'' to 
`unsigned char*
   ''
> I think we may need to visit (or document carefully) what happens when 
> you pass in a string of data to the constructor or to SetAlpha.
We should have aliases that disambiguate the different forms of new - 
maybe Image.from_data or .new_from_data.
> Does anyone know of any reason we would or wouldn''t want to just
set
> the unsigned char * typemap in typemap.i? 
Nope, but I probably wouldn''t if there was one. Would be nice to fix 
Icon.new(data) as well.  There is Icon.new(void*data) and 
Icon.new(char**bits) ...

thanks
alex

Alex Fenton wrote:
> Committed, thanks. I get several warnings now when compiling Image.cpp 
> of the type:
>
> src/Image.cpp: In function `VALUE _wrap_new_wxImage__SWIG_3(int, VALUE*, 
> long
>    unsigned int)'':
> src/Image.cpp:2560: warning: invalid conversion from `char*'' to 
> `unsigned char*
>    ''
>   
Hmm...  Yeah, that''s not ideal.  I may have to visit that to see if 
there''s a good workaround.  I also was noticing that SWIG already 
handles unsigned char (not the pointer) so I may patch to remove that 
particular one.
>> I think we may need to visit (or document carefully) what happens when 
>> you pass in a string of data to the constructor or to SetAlpha.
>>     
> We should have aliases that disambiguate the different forms of new - 
> maybe Image.from_data or .new_from_data.
>   
Could be good.  I think wxPython does this.  Most of the time we''ve
been
able to disambiguate without the use of new method names.

Roy