nmrmlconv should be prepared to receive multi-fid bruker zipped folder

[pcm32]

Current implementation expects to find a single acquisition file inside the zipped file, and produces likewise a single nmrml file:

export indir=`dirname /opt/galaxy_data/files/062/dataset_62500.dat`
export cpfile=mesa.zip; export samplename=`basename $cpfile .zip`
echo "vendor: bruker"; echo "infile: /opt/galaxy_data/files/062/dataset_62500.dat"; echo "cpfile: $cpfile"; echo "samplename: $samplename"; 
cp -f /opt/galaxy_data/files/062/dataset_62500.dat /tmp/$cpfile
mkdir -p /tmp/$samplename
unzip -q -d /tmp/$samplename /tmp/$cpfile

# here this find will return more than one entry if found, but downwards this is not handled
export acqusfile=`find /tmp/$samplename -name "acqus" -print`
echo "acquisitionfile: $acqusfile"
mv `dirname $acqusfile`/* /tmp/$samplename/

java -jar /usr/local/share/nmrML/converter.jar -l create -b -z -t bruker -a $samplename -i /tmp/$samplename -o /opt/galaxy_data/files/062/dataset_62501.dat;

Besides, files shouldn't be copied (and specially outside of the working directory).

What should happen is that a collection of NMRml files are generated, if only one acquisition file is found, then a collection with a single file would be generated. Issue as well here is that this would need to be distributed if the bruker zipped directory is too large.

@korseby @jianlianggao

This functionality seems to be supported by nmrml.org site, so we should probably be able to support it as well through our galaxy module.