[ITK] Tests with multi-file input data (Bruker2DSEQ)

Lowekamp, Bradley (NIH/NLM/LHC) [C] blowekamp at mail.nih.gov
Wed Jun 14 08:42:03 EDT 2017


Hello,

Thank you for your work on this.

Re 1:

The Bruker ImageIO should be moved out of the review directory an placed in own ITK module under "Modules/IO/Bruker2DSEQ”, which you appear to have already done in your branch. Outstanding! I would place the data in this modules under the “test/Input” subdirectory. If you need an entire folder, you can create it here. I am not sure what problem you encountered with this layout.


Re 2:
You can find the CMake documentation for that here:
https://cmake.org/cmake/help/v3.3/module/ExternalData.html#referencing-file-series

There are quite a number of options for referencing files series, file pair, and associated directories. I would expect this to meet you needs.

I glanced at the contents: https://github.com/spinicist/ITK/commit/541411449a9538b758507672583c88db949d90cb It’s not clear to me what files are needed when reading, if you need recommendation on using `DATA{}` please be specific with the files needed by the IO, and their locations.

HTH,
Brad



On Jun 14, 2017, at 6:19 AM, Wood, Tobias <tobias.wood at kcl.ac.uk<mailto:tobias.wood at kcl.ac.uk>> wrote:

Hello,

I have been slowly working on my updated Bruker2DSEQ file reader. It’s not quite ready for formal review, but it is available here: https://github.com/spinicist/ITK (On the Bruker branch)

I have reached a problem writing a suitable test, because the Bruker data format consists of multiple files. I previously discussed this issue with Jon Haitz, but despite his help I am stuck.

I have added a new example Bruker file in Testing/Data/Input/Bruker2DSEQ. There was already a Bruker dataset in Testing/Data/Input/DCB120604.t61, but this is from (I think) ParaVision 3. Bruker made a *lot* of changes when they released ParaVision 6, one of which was the introduction of a new header file called “visu_pars”. The actual image data is still in the “2dseq” file.

My problem is that the itk Test Driver only transfers the 2dseq file to build/ExternalData/Testing/Data/Input/Bruker2DSEQ/... , and not the “visu_pars” file, and hence trying to read the file fails.

So, questions:
1 - Is this the correct place to put the test data? Because of the Bruker directory structure, I assumed the recommended way of requesting an MD5 hash and putting the data in the module’s test directory won’t work?
2 - How do I indicate to the test driver that it needs to copy the whole directory structure, and not just the 2dseq file? Eventually, I would like the reader to read all the header files (there are several of them) because they contain useful meta-data that users might want to extract.

I was hoping to do a bit more polishing before putting this out for any kind of review, but I guess now is as a good a time as any. Let me know if you need more information. Note - I moved the module out of Nonunit/Review, I hope this was a good idea but I can always move it back if necessary.

Thanks,
Toby

_______________________________________________
Community mailing list
Community at itk.org<mailto:Community at itk.org>
http://public.kitware.com/mailman/listinfo/community

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/community/attachments/20170614/a792d608/attachment.html>


More information about the Community mailing list