[ome-users] Fwd: Errors when loading Cellomics .DIB with non-standard filename

[Melissa Linkert]

Hi Lee,

> We packaged a Cellomics dataset as part of the Broad Bioimage Benchmark
> Collection, BBBC018. It's available for download here:
> http://www.broadinstitute.org/bbbc/BBBC018/BBBC018_v1_images.zip
> 
> We renamed the files when we packaged it up - unfortunately. It's not the
> first time that something like this has happened, not under our control.
> When I unpack the zip and use showinf to read the image, I get the
> following stack trace:
*snip*
> I'm going to submit a pull request for a patch to this - I have both the
> directory of zip files and a representative directory of Cellomics files
> with their correct file names and I want to make sure what I do works in
> CellProfiler (https://github.com/CellProfiler/CellProfiler/issues/1043).

My apologies for not following up here sooner, and thank you for
submitting the pull request:

https://github.com/openmicroscopy/bioformats/pull/923

> I do have one question as I'm implementing. This line:
> https://github.com/openmicroscopy/bioformats/blob/develop/components/bio-formats/src/loci/formats/in/CellomicsReader.java#L195
> 
> prevents the reader from including all of a plate's files unless there is a
> full plate map. It seems a little cruel... just want to make sure it
> belongs there, I have one directory which is a full plate and another which
> is just short of full.

The main reason for that being in place was to prevent the reader from
being confused when multiple files copied from different plates were
present in the same directory.  It's not strictly necessary, and could
definitely be replaced with something a little smarter (which I can
certainly do in an upcoming pull request).

Regards,
-Melissa

On Thu, Feb 13, 2014 at 10:33:03AM -0500, Lee Kamentsky wrote:
> Hi all,
> We packaged a Cellomics dataset as part of the Broad Bioimage Benchmark
> Collection, BBBC018. It's available for download here:
> http://www.broadinstitute.org/bbbc/BBBC018/BBBC018_v1_images.zip
> 
> We renamed the files when we packaged it up - unfortunately. It's not the
> first time that something like this has happened, not under our control.
> When I unpack the zip and use showinf to read the image, I get the
> following stack trace:
> 
> c:\Temp\bioformats_5.0.0_rc2>showinf -no-upgrade
> c:\Temp\BBBC018_v1_images\00733-actin.DIB
> Checking file format [Cellomics C01]
> Initializing reader
> CellomicsReader initializing c:\Temp\BBBC018_v1_images\00733-actin.DIB
> Exception in thread "main" java.lang.StringIndexOutOfBoundsException:
> String index out of range: 1
>         at java.lang.String.charAt(Unknown Source)
>         at
> loci.formats.in.CellomicsReader.getWellName(CellomicsReader.java:371)
>         at
> loci.formats.in.CellomicsReader.getChannel(CellomicsReader.java:395)
>         at
> loci.formats.in.CellomicsReader.initFile(CellomicsReader.java:149)
>         at loci.formats.FormatReader.setId(FormatReader.java:1360)
>         at loci.formats.ImageReader.setId(ImageReader.java:781)
>         at loci.formats.ReaderWrapper.setId(ReaderWrapper.java:576)
>         at loci.formats.tools.ImageInfo.testRead(ImageInfo.java:993)
>         at loci.formats.tools.ImageInfo.main(ImageInfo.java:1075)
> c:\Temp\bioformats_5.0.0_rc2>
> 
> The problem here is that much of the code in the reader assumes that it
> will get a reasonable file name and doesn't catch the cases where the
> assumptions are violated. In this case, the code assumes that the well
> metadata appears after the last underbar: "images\00733-actin.DIB" (it
> should have a format like "A01f00o1.DIB").
> 
> I'm going to submit a pull request for a patch to this - I have both the
> directory of zip files and a representative directory of Cellomics files
> with their correct file names and I want to make sure what I do works in
> CellProfiler (https://github.com/CellProfiler/CellProfiler/issues/1043).
> 
> I do have one question as I'm implementing. This line:
> https://github.com/openmicroscopy/bioformats/blob/develop/components/bio-formats/src/loci/formats/in/CellomicsReader.java#L195
> 
> prevents the reader from including all of a plate's files unless there is a
> full plate map. It seems a little cruel... just want to make sure it
> belongs there, I have one directory which is a full plate and another which
> is just short of full.
> 
> --Lee

> _______________________________________________
> ome-users mailing list
> ome-users at lists.openmicroscopy.org.uk
> http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-users