Added imread, imsave and synchronize functionality

[thawn]

You can now just use imread(filename) and the library uses a local file on the cluster if it finds a matching one, or it downloads the file to a temporary location from the fileserver and accesses that one.

Similarly, imsave(filename, data), saves the file if the path is accessible. Otherwise, it saves to a temp file and then copies that to the final location (on the fileserver or project space) using the datamover tools.

The best way to test is to install the test singularity image for biapol-taurus:

git clone https://gitlab.mn.tu-dresden.de/bia-pol/singularity-devbio-napari.git
cd singularity-devbio-napari
./install.sh biapol_taurus

and then open a notebook with the devbio-napari_biapol_taurus kernel.

inside the notebook run:

from biapol_taurus import ProjectFileTransfer
from skimage.io import imshow

# Define the source directory. /grp/g_biapol/ is the mount point at the export node. HPCsupport can tell us this:
source_dir = "/grp/g_biapol/"

# Define the target directory. /projects/p_bioimage/ is the project space. HPCsupport told us that when we applied for the project:
target_dir = "/projects/p_bioimage/"

# We setup a connection:
pft = ProjectFileTransfer(source_dir, target_dir)

image = pft.imread('data/test/blobs19.tif')
# show image
imshow(image)

pft.imsave('/projects/p_bioimage/till/blobs_saved.tif', image)