//////////////// Step 1 (see above description):
- gdcmHeader *header = new gdcmHeader( filename, false );
+ gdcm::Header *header = new gdcm::Header( filename );
if( !header->IsReadable() )
{
std::cerr << "Test::TestReadWriteReadCompare: Image not gdcm compatible:"
//////////////// Step 2:
- gdcmFile* file = new gdcmFile( header );
+ gdcm::File* file = new gdcm::File( header );
int dataSize = file->GetImageDataSize();
- void* imageData = file->GetImageData(); //EXTREMELY IMPORTANT
+ uint8_t* imageData = file->GetImageData(); //EXTREMELY IMPORTANT
+ // Sure, it is : It's up to the user to decide if he wants to
+ // GetImageData or if he wants to GetImageDataRaw
+ // (even if we do it by setting a flag, he will have to decide)
/// \todo Following line commented out because gdcmFile::SetImageData() is
/// brain dead: it sets ImageDataSize to its argument and PixelRead to a.
- /// Later on, when writing gdcmFile::WriteBase() and because PixelRead == 1
- /// we call
+ /// Later on, when writing gdcmFile::WriteBase()
+ /// and because PixelRead == 1 we call
/// PixelElement->SetLength( ImageDataSizeRaw );
/// where we use ImageDataSizeRAW instead of ImageDataSize !
/// But when the original image made the transformation LUT -> RGB,
/// images 8BitsUncompressedColor.dcm, OT-PAL-8-face.dcm and
/// US-PAL-8-10x-echo.dcm...
/// In conclusion fix gdcmFile, and then uncomment the following line.
- /// file->SetImageData(imageData, dataSize);
+
+ // --> I did. ctest doesn't break. But ... is it enought to say it's OK ?
+
+ file->SetImageData(imageData, dataSize);
+
file->WriteDcmExplVR( "TestReadWriteReadCompare.dcm" );
std::cout << "2...";
//////////////// Step 3:
- gdcmFile* reread = new gdcmFile( "TestReadWriteReadCompare.dcm", false );
+ gdcm::File* reread = new gdcm::File( "TestReadWriteReadCompare.dcm" );
if( !reread->GetHeader()->IsReadable() )
{
std::cerr << "Test::TestReadWriteReadCompare: Could not reread image "