[Dcmlib] [Fwd: RE: [Insight-developers] Update of GDCM library]

[Mathieu Malaterre]

Ok au prochain chgmt d'API c'est sur gdcm vire d'ITK. Je sais meme pas 
si on va garder 1.0 dans ITK.

Faut dire le changement
gdcm::Header -> gdcm::File
gdcm::File   -> gdcm::FileHelper

c'est du pure vice...

-------- Original Message --------
Subject: RE: [Insight-developers] Update of GDCM library
Date: Thu, 24 Feb 2005 14:51:43 -0500
From: Lorensen, William E (Research) <>
To: Mathieu Malaterre <>
CC: Insight Developers <>

In the future, we shoulod discuss any API changes. We have assured our 
users that itk API will be stable. If this does not match the philosopy 
of GDCM, then we may have to keep our own version.

Bill


-----Original Message-----
From: insight-developers-bounces at itk.org
[mailto:insight-developers-bounces at itk.org]On Behalf Of Mathieu
Malaterre
Sent: Thursday, February 24, 2005 2:18 PM
To: Mathieu Malaterre
Cc: Insight Developers
Subject: Re: [Insight-developers] Update of GDCM library


One thing I forgot to mention. Is that the API was seriously changed. So
if you were using

gdcm::Header::GetEntryByNumber

you should not use:

gdcm::File::GetEntry

since the GetEntryByName was removed (dangerous) there was no need to
maintain the function with a longer name.

Also

- gdcm::File is now gdcm::FileHelper

- gdcm::SerieHeader is now gdcm::SerieHelper

And I believe these are the mose common 'public' API changes.

Let me know if you have any more troubles updating your code.

Thanks for your patience
Mathieu

Mathieu Malaterre wrote:
> Hi,
> 
>     I have just updated the GDCM library from 0.6 branch to 1.0 branch. 
> I also removed the GDCM option so from now on GDCM is always build as 
> part of ITK.
> 
>     Changes include:
> 
> - Better big endian support (in particular for jpeg image)
> 
> - Full support of fragmented jpeg images (where jpeg stream was split in 
> small chunks)
> 
> - Lots of API cleanup (remove the redundant gdcm at begining of functions)
> 
> - Complete removal of the accessors by DICOM name. This was considered 
> dangerous since DICOM name is not guarantee to be unique, and the name 
> could change from one revision to another. Not a notion of generelized 
> key is used intead. Ex. insteasd of "Patient's Name", now you should use 
>  "0010|0010"
> 
> - dicomDirV3.dic was update to also take into account the VM (value 
> multiplicity).
> 
> - Enhance support for DICOMDIR file
> 
> - Partial support for Papyrus image (very close to Dicom V3 specification)
> 
> - Huge number of memory leaks were fixed.
> 
> 
> For more information:
> http://www.creatis.insa-lyon.fr/Public/Gdcm/
> and
> http://www.creatis.insa-lyon.fr/Public/Gdcm/News.html
> 
> Mathieu
> _______________________________________________
> Insight-developers mailing list
> Insight-developers at itk.org
> http://www.itk.org/mailman/listinfo/insight-developers
> 

_______________________________________________
Insight-developers mailing list
Insight-developers at itk.org
http://www.itk.org/mailman/listinfo/insight-developers
_______________________________________________
Insight-developers mailing list
Insight-developers at itk.org
http://www.itk.org/mailman/listinfo/insight-developers