Hello Fabien,
This issue is indeed due to the tool design and is not related to the tool
shed. I'm not primarily a tool developer, so I'm not the best person to help
you with this issue.
However, one recommendation would be to place your scripts and binaries that a
specific tool wrapper uses in the same directory as the tool wrapper so that
paths are not necessary. Some of the binaries may be more appropriately added
to the environment path used by the user that starts up your Galaxy instance (
the R binaries are a good example ). This is the approach that the tools
included in the Galaxy distribution use.
The tool shed will not do anything special with tool wrapper that include
hard-coded paths like this. The tool shed generally only makes sure that the
tool loads properly in Galaxy. However, recent enhancements to the tool shed
will handle tool dependencies that are properly defined in the tool config file
within the <requirements> tag set. These new tool shed features will soon be
documented in the tool shed wiki.
Thanks,
Greg Von Kuster
On Jul 3, 2012, at 12:27 PM, Fabien Mareuil wrote:
> Hi,
>
> Thank for your answer, I copy you a part of the shell code
>
> LOCAL_DIR= #{GALAXY_DIR}/galaxy-dist/tools/AnnotateGenes
>
> R_DIR= #{R_PATH}/bin
>
> echo "ChIP:" >$LOG
>
> if [ -r $REG ]; then
>
> echo "1: perl $LOCAL_DIR/geneAnnotation.pl -g
> $LOCAL_DIR/$GENOME.noIdenticalTransc.txt -tf $CHIPFILE -selG $REG -o
> $OUTSTAT -lp $LEFTPROM -rightp $RIGHTPROM -enh $ENH -dg $DOWNGENE" >>
> $LOGTMP
>
> and a part of the xml
>
> <tool id="annotateGenes" name="Annotation of genes with Chip-Seq peaks"
> version="1.0">
> <description> </description>
> <command interpreter="bash"> #if $use_reg.use_reg_selector == "no" and
> $use_control.use_control_selector == "no" #annotateGenes_wrapper.sh -f
> $inputfile -y $log -l $left -o $outputPNG -r $right -d $DownGene -h
> $EnhLeft -u $stats -v $input_organism.version #elif
> $use_reg.use_reg_selector == "no" and $use_control.use_control_selector
> == "yes" # annotateGene_wrapper.sh -f $inputfile -y $log -c $controlfile
> -x $statsControl -o $outputPNG -l $left -r $right -d $DownGene -h
> $EnhLeft -u $stats -v $input_organism.version #elif
> $use_reg.use_reg_selector == "yes" and $use_control.use_control_selector
> == "no" # annotateGenes_wrapper.sh -y $log -f $inputfile -e $regfile -l
> $left -o $outputPNG -r $right -d $DownGene -h $EnhLeft -u $stats -v
> $input_organism.version #else # annotateGenes_wrapper.sh -f $inputfile
> -c $controlfile -x $statsControl -l $left -y $log -o $outputPNG -r $right
> -d $DownGene -h $EnhLeft -u $stats -v $input_organism.version -e $regfile
> #end if
>
>
> This tools are avaible in : http://nebula.curie.fr/
>
> You can see that the variable LOCAL_DIR is the PATH of the tool so I would
> like to know if it's possible to obtain this information without
> hard-coded this?
>
> Thank you for your answer.
> Best Regards,
>
> Fabien Mareuil
>
>> Hello Fabien,
>> I don't understand the issue - can you provide a sample tool config that
> includes these hard-coded paths? This initially sounds like an issue
> with
>> the tool configs, not the tool shed, but I may see the problem with your
> clarification.
>> Thanks,
>> Greg Von Kuster
>> On Jul 3, 2012, at 9:58 AM, Fabien Mareuil wrote:
>>> Hi,
>>> I have read the exchange betwen you and Florent Angly about "Problem with
>>> new tool shed" and I have a problem with Nebula Tools:
>>> http://nebula.curie.fr/.
>>> At the Pasteur Institute, we have 4 galaxy instances and I would like
> to
>>> use a local tool shed instance for Nebula installation.
>>> However, the nebula has tools with hard-coded path tool but I don't
> want
>>> hard-coded this so do you have a solution to add a thing like this
> ${tool.install_dir} in the xml?
>>> Thank you for your answer.
>>> Best Regards,
>>> Fabien Mareuil
>>> ___________________________________________________________
>>> The Galaxy User list should be used for the discussion of
>>> Galaxy analysis and other features on the public server
>>> at usegalaxy.org. Please keep all replies on the list by
>>> using "reply all" in your mail client. For discussion of
>>> local Galaxy instances and the Galaxy source code, please
>>> use the Galaxy Development list:
>>> http://lists.bx.psu.edu/listinfo/galaxy-dev
>>> To manage your subscriptions to this and other Galaxy lists,
>>> please use the interface at:
>>> http://lists.bx.psu.edu/
>
>
>
>
___________________________________________________________
The Galaxy User list should be used for the discussion of
Galaxy analysis and other features on the public server
at usegalaxy.org. Please keep all replies on the list by
using "reply all" in your mail client. For discussion of
local Galaxy instances and the Galaxy source code, please
use the Galaxy Development list:
http://lists.bx.psu.edu/listinfo/galaxy-dev
To manage your subscriptions to this and other Galaxy lists,
please use the interface at:
http://lists.bx.psu.edu/
