[Dcmlib] Re: gdcm::IsReadable()

[Jean-Pierre Roux]

Mathieu Malaterre wrote:

[...]

>> But some mandatory Tags (such as Pixel Representation, Sample Per 
>> Pixels, Hight Bit, etc) are missing.
>> Now gdcmFile::IsReadable() doesn't say 'false' is *only one* is missing.
>> (The further heuristics try to manage with them).
>
>
> Just a quick note, it looks to me that now IsReadable() can return 
> true as long as -let say- b0028_0103 is true, are you sure this is 
> enough ?

I do agree with you : the stuff looks strange, but it works :
A DICOMDIR, a Dicom RT, a Dicom SR (or a .zip, .exe, or whatever you 
want) will be flagged as 'non gdcm::File readable'.
A rotten header ACR-NEMA will be flagged as gdcm::File readable, since 
he has at least *one* of the few mandatory tags we check.
(they all have a reasonable default value, in our heuristics)

gdcm::File::IsReadable() is just an information on whether or not it's 
worthy to try to read it.
JPEG-LS are flagged as gdcm::File readable, even if actually there 
aren't (till now).
So would be a  buggly compressed file...
The only accurate way of check if a file *is* readable is to read it.
gdcm::File::IsReadable() == false just tells us : don't even try !

JP

>> And there is no more DicomWorks compatible image that hang gdcm.
>> ;-)
>
>
> GDCM ROCKS DA WORLD !
>
> Mathieu
> Ps: Does this means there are still some DicomWorks *non*-compatible 
> that hang gdcm ?


Some  e-film hanger ( + DicomWorks hanger) DICOMDIR -comming from 
Philips ViewForum station- cause gdcm inifinite loop.
(Marlene cannot read the CD under Benoit's CreaTools)

I'm trying to locate the pb, and solve it.