Draft: WIP: Implement processing of al NX entries in NX file in nabu.reconstruct.py

This is a draft PR for the integration of fluo-tomo within the nabu framework. This is a first step of the integration, since the target integration would likely take the form of a new pipeline.

What needs to be done:

• Convert fluo tomom data in NX files. See nxtomomill and tomoscan PRs. Had to fix sth in branch fluo_fix (see PR.
• Make corrct/MLEM algo available. WIP...
• Allow "all" HDF5 entries in NX file to be reconstructed by the fullfield pipeline.
• Add unit tests.
• Add an nabu-fluo app that takes shifts as input and reconstructs all fluo lines.

For tests:

• Dataset boxK4 to be used for tests. Conversion to nx OK.
• Compute shifts and prepare translations.txt file.
• ...