[ome-devel] adding woolz to bioformats
Melissa Linkert
melissa at glencoesoftware.com
Fri Jul 12 14:57:57 BST 2013
Hi Bill,
> I've pulled a fresh development (the default) branch of
> bioformats from github and added (reduced git status output):
>
>
> # new file:
> components/bio-formats/src/loci/formats/in/WlzReader.java
>
> # modified: components/scifio/src/loci/formats/readers.txt
> # modified: scan-deps.pl
> #
> # Untracked files:
> # (use "git add <file>..." to include in what will be committed)
> #
> # 00-bills-notes.txt
> # jar/JWlz.jar
> # lib/libJWlz-1.3.2.so
> # lib/libJWlz.so
>
> *Q* How do I get the build system to see/use my JWlz jar file?
To have your reader build with jar/JWlz.jar, you
will need to update the component.classpath property in
components/bio-formats/build.properties to include ${lib.dir}/JWlz.jar.
You will also need to add a block similar to this in the "dependencies"
section of components/bio-formats/pom.xml:
<dependency>
<groupId>woolz</groupId>
<artifactId>JWlz</artifactId>
<version>1.3.2</version>
<scope>system</scope>
<systemPath>${basedir}/jar/JWlz.jar</systemPath>
</dependency>
The JWlz.jar and any of the *.so files that it needs should be committed
on the same branch as WlzReader.java.
> I modified scan-deps.pl:
>
> +my %woolz = (
> + NAME => "woolz",
> + TITLE => "Woolz",
> + JAR => "JWlz.jar",
> + PACKAGE => "uk.ac.mrc.hgu.Wlz",
> + LICENSE => "GPL",
> + URL => "https://github.com/ma-tech/Woolz",
> + NOTES => <<ZZ,
> +used for reading/writing Woolz objects
> +ZZ
> + VERSION => "1.3.2"
> +);
>
> although I'm not sure if I needed to do this (*Q*)? Trying
> to run scan-deps.pl just gives complaints about non-existant
> jar files, eg
>
> Component bf-autogen.jar does not exist at ./scan-deps.pl line 623.
>
> *Q* Is this normal?
Updating scan-deps.pl is not something you need to worry about; it won't
affect building the reader. If needed, we can always update it later.
> Previously I just hacked in a symbolic link to my java-woolz
> sources.
>
> I know I've still to write the service that will isolate
> bioformats from Woolz, but I'll write it once I've a sensible
> build going with read/write of Woolz objects. That way I'll
> know what's needed.
Makes complete sense. If any of the above is unclear, just let us know.
Regards,
-Melissa
On Fri, Jul 12, 2013 at 02:06:53PM +0100, Bill Hill wrote:
> I'm revisiting the Woolz file reader. This now works for most
> Woolz objects that would probably be considered images (2D,
> 3D, 1/2 a dozen pixel/voxel types, no pixels/voxels give an
> unsigned byte 0/255 image corresponding to the object domain.
>
> When I wrote the reader I just hacked the build system to get
> it going - I now want to add it the right way.
>
> I've pulled a fresh development (the default) branch of
> bioformats from github and added (reduced git status output):
>
>
> # new file:
> components/bio-formats/src/loci/formats/in/WlzReader.java
>
> # modified: components/scifio/src/loci/formats/readers.txt
> # modified: scan-deps.pl
> #
> # Untracked files:
> # (use "git add <file>..." to include in what will be committed)
> #
> # 00-bills-notes.txt
> # jar/JWlz.jar
> # lib/libJWlz-1.3.2.so
> # lib/libJWlz.so
>
> *Q* How do I get the build system to see/use my JWlz jar file?
>
> I modified scan-deps.pl:
>
> +my %woolz = (
> + NAME => "woolz",
> + TITLE => "Woolz",
> + JAR => "JWlz.jar",
> + PACKAGE => "uk.ac.mrc.hgu.Wlz",
> + LICENSE => "GPL",
> + URL => "https://github.com/ma-tech/Woolz",
> + NOTES => <<ZZ,
> +used for reading/writing Woolz objects
> +ZZ
> + VERSION => "1.3.2"
> +);
>
> although I'm not sure if I needed to do this (*Q*)? Trying
> to run scan-deps.pl just gives complaints about non-existant
> jar files, eg
>
> Component bf-autogen.jar does not exist at ./scan-deps.pl line 623.
>
> *Q* Is this normal?
>
> Previously I just hacked in a symbolic link to my java-woolz
> sources.
>
> I know I've still to write the service that will isolate
> bioformats from Woolz, but I'll write it once I've a sensible
> build going with read/write of Woolz objects. That way I'll
> know what's needed.
>
> Thanks for the help so far.
>
> Cheers,
> Bill
>
> >After speaking to Melissa at the Paris meeting I've written a
> >Woolz reader for bioformats using the JNI. There are still a
> >few bits to complete and I haven't started the writer, but
> >the basics are there.
> >
> >*Q* At the meeting we talked about some way of encoding the
> >origin (Woolz objects have an arbitrary, possibly -ve origin).
> >I can't remember, did it involve stage coordinates meta-data?
> >
> >Test run:
> >
> >bill at bill-lt% setenv LD_LIBRARY_PATH \
> > /home/bill/src/Bioformats/bioformats/lib
> >bill at bill-lt% ./tools/showinf ~/pics/artic_s4_02_06_06.wlz
> >Checking file format [Woolz]
> >Initializing reader
> >WlzReader initializing /home/bill/pics/artic_s4_02_06_06.wlz
> >Initialization took 0.038s
> >
> >Reading core metadata
> >Filename = /home/bill/pics/artic_s4_02_06_06.wlz
> >Series count = 1
> >Series #0 :
> > Image count = 1
> > RGB = false (1)
> > Interleaved = false
> > Indexed = false (false color)
> > Width = 424
> > Height = 199
> > SizeZ = 1
> > SizeT = 1
> > SizeC = 1
> > Thumbnail size = 128 x 60
> > Endianness = motorola (big)
> > Dimension order = XYZCT (certain)
> > Pixel type = uint8
> > Valid bits per pixel = 8
> > Metadata complete = false
> > Thumbnail series = false
> > -----
> > Plane #0 <=> Z 0, C 0, T 0
> >
> >
> >Reading pixel data (0-0)
> > Read 1/1 planes (100%)
> >[done]
> >0.027s elapsed (27.0ms per plane)
> >
> >Launching image viewer
> >
> >Reading global metadata
> >
> >Reading metadata
> >
> >Cheers,
> >Bill
>
> --
> Bill Hill : http://www.hgu.mrc.ac.uk
> MRC Human Genetics Unit : http://www.emouseatlas.org
> MRC IGMM, University of Edinburgh : Bill.Hill at igmm.ed.ac.uk
> Western General Hospital : +44-131-3322471x2119
> Crewe Road, Edinburgh EH4 2XU, UK. : +44-131-4678456
>
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
>
> _______________________________________________
> ome-devel mailing list
> ome-devel at lists.openmicroscopy.org.uk
> http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-devel
More information about the ome-devel
mailing list