int const pixels = img.depth() > 8 ?
img.width() * img.height() : img.numColors();
- // FIXME this code used to be like this:
- //
- //unsigned int * const data = img.depth() > 8 ?
- // (unsigned int *)img.bits() :
- // (unsigned int *)img.jumpTable();
- //
- // But Qt doc just say use bits...
- unsigned int * const data = (unsigned int *)img.bits();
+ unsigned int *data = img.depth() > 8 ?
+ reinterpret_cast<unsigned int *>(img.bits()) :
+ reinterpret_cast<unsigned int *>(&img.colorTable()[0]);
for(int i = 0; i < pixels; ++i){
int const val = qGray(data[i]);
QMatrix m;
m.rotate(-params.angle);
- // FIXME: alpha chanel detection is automautic for monochrome
- // and 8-bit images. For 32 bit, is something like still necessary?
- //transformed_.setAlphaBuffer(true);
transformed_ = transformed_.transformed(m);
}