[ENH]: DICOM to NIFTI BIDS conversion with Heudiconv

[alexpron]

Adding datalad support within shanoir2bids.py script relying on heudiconv software.

Tried to reproduce as far as possible the current usage of the shanoir2bids.py script from configuration.json file

• DICOM download only from Shanoir instance (regression)
• removed automri format output (regression)
• BIDS mapping generated from the result of the solr search to handle case where regex like terms were passed into .json file
• BIDS conversion is performed at the subject level (session if longitudinal)
• Modified a bit dcm2niix options so that it matches the --dcmconfig option .json file
  (keys are attributes to DCM2NIIX nipype class)
• DICOM to NIFTI BIDS conversion, both DICOM and NIFTI are stored into datalad datasets (sourcedata, rawdata)

TODO:

• check bids metadata generated
• add studyname in dl_dir using locator ?
• better handle temporary directory
• add conversion directly from zip files
• write Heudiconv heuristic to handle conversion into automri format
• define what to do when chosen outputs are not BIDS compliant (e.g. sub-\d\d--\d\d)

[alexpron]

some bids compliance errors with default dcm2niix options used:

[alexpron]

• bids compliance problem was solved using -ba option of dcm2niix if -ba should not be applyied check field by field
• bids validator was run on the obtained files and current filenames used should be modified in the configuration file
• some characters are forbidden in subject_id such as - hence XX-XX is not a BIDS compliant name
• in order to keep this name use the non bids formatting and manually recreats ontology within the heuristic file ?

[alexpron]

tar or tar.gz files only are correctly handled by heudiconv (.zip files currently not supported)

[alexpron]

• include debug mode with directory kept i.e. switch to manually specified directories and manual delete if required
• remove commit 5064edb
• remove {subject_session_dir} prefix in case of automri heuristic
• remove {subject_session_prefix} in case of automri heuristic
• if not working implement simple BIDS to automri mappping

Optional

to be done in a dedicated PR

• solr query enhancement : concatenate all queries into a single one

[alexpron]

• include debug mode with directory kept i.e. switch to manually specified directories and manual delete if required
  by default temporary directory are deleted after context execution.
  The delete keyword allowing to keep the directory is only available in python 3.12 which is too restrictive currently
  Hence temporary directory mechanism will be replaced by directories and conditional manual deletion

See https://docs.python.org/3.12/library/tempfile.html#tempfile.TemporaryDirectory and https://docs.python.org/3.11/library/tempfile.html#tempfile.TemporaryDirectory