Hi Andrew,
I have just put together the various changes suggested in this thread,
updating daeRMaterials.cpp to better handle texture paths. These
changes are now checked into CVS.
These changes work fine on the models I have, but would appreciate
others having a glance over the changes and testing them out.
Cheers,
Robert.
On 10/16/06, [EMAIL PROTECTED]
<[EMAIL PROTECTED]> wrote:
Robert Osfield wrote:
> Hi Andrew,
>
> I just reviewed you changes but am curious about a couple of parts to it:
>
> const unsigned int bufSize = 261;
> char path[bufSize];
> if ( !dImg->getInit_from()->getValue().getPath( path, bufSize ) )
> {
> osg::notify( osg::WARN ) << "Unable to get path from URI." <<
> std::endl;
> return NULL;
> }
>
>
> Is there a reason for fixed size of 261?
I believe Windows has a max filepath length of 260 characters. I don't
know if *nix machines have similar restrictions. We could dynamically
allocate the buffer if you think it would be better.
unsigned int bufSize = 1; //for the null char
if ( dImg->getInit_from()->getValue().getFilepath() != NULL )
{
bufSize += strlen( dImg->getInit_from()->getValue().getFilepath() );
}
if ( dImg->getInit_from()->getValue().getFile() != NULL )
{
bufSize += strlen( dImg->getInit_from()->getValue().getFile() );
}
char *path = new char[bufSize+1];
if ( !dImg->getInit_from()->getValue().getPath( path, bufSize ) )
{
osg::notify( osg::WARN ) << "Unable to get path from URI." << std::endl;
return NULL;
}
Just remember to delete it afterwards.
>
> #ifdef WIN32
> std::string filename( path+1 );
> #else
> std::string filename( path );
> #endif
>
>
> Why the +1 in the case of WIN32? Is there a Win32 specific character
> appended on the front on the Win32 version of the COLLADA DOM?
There isn't any platform specific handling of URI's in the DOM. The
reason for the +1 is that we get back a leading '/' that needs to be
removed on Win32 systems, ie. /c:/path/file
>
>
> img = osgDB::readImageFile( filename.c_str() );
>
>
>
> Since readImageFile takes a c char* pointer perhaps we could just pass
> in the path rather than converting to std::string, then back to c_str()
> as above.
Thats an oversight on my part. Yeah it would be better to pass in the
path directly without converting to std::string.
>
> Thoughts?
> Robert.
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> osg-users mailing list
> [email protected]
> http://openscenegraph.net/mailman/listinfo/osg-users
> http://www.openscenegraph.org/
_______________________________________________
osg-users mailing list
[email protected]
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
_______________________________________________
osg-users mailing list
[email protected]
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
